UNIVERSITY OF CALIFORNIA DEPARTMENT OF CIVIL ENGINEER^ PERKELEY. CALIFORNIA Engineering Library UN.VERS.TY OF CAUFORN.A OF 'ERKELEY. CALIFORNIA * 4* AN INTRODUCTORY COURSE IN QUANTITATIVE CHEMICAL ANALYSIS THE MACMILLAN COMPANY NEW YORK BOSTON CHICAGO DALLAS ATLANTA SAN FRANCISCO MACMILLAN & CO., LIMITED LONDON BOMBAY CALCUTTA MELBOURNE THE MACMILLAN CO. OF CANADA, LTD. TORONTO AN INTRODUCTORY COURSE IN I" QUANTITATIVE CHEMICAL ANALYSIS WITH EXPLANATORY NOTES, STOICHIOMETRICAL PROBLEMS AND QUESTIONS BY GEORGE McPHAIL SMITH n Associate Professor of Chemistry in the t University of Illinois ' o . . , > >, . . , 5 ' ' JO J >'*"*" Ntfo gorfc THE MACMILLAN COMPANY 1919 All rights reserved S5" Engineering Library COPYRIGHT, 1919, BY THE MACMILLAN COMPANY. Set up and electrotyped. Published July, 1919. J. S. Gushing Co. Berwick & Smith Co. Norwood, Mass., U.S.A. PREFACE THIS introductory course in Quantitative Analysis is designed for use with classes consisting of students who have completed courses in Elementary Chemistry and Qualitative Analysis, and who are beginning work in Quantitative Analysis. On the laboratory side, its primary intent is to provide the student with directions sufficiently detailed to offer little opportunity for going astray, and thus to enable him to work successfully without an undue amount of personal supervision. The instructor is thereby placed in a position, in the laboratory as well as in the classroom, to exert his personal influence more especially towards the development of theoretical knowledge and independent thought on the part of the students. The use of the book in the laboratory should of course be supplemented by regular classroom instruction ; and, with this in mind, it has seemed desirable to include the stoichiometrical problems of Part IV, and the questions of Part V. The problems of Part IV are such as are constantly met with in analytical work, and their conscientious study will give the student an in- sight into the principles of a wide variety of processes; the answers to the problems have been intentionally omitted. It is the writer's practice to require, as a written exercise to be handed in at the beginning of a recitation, the solution of a definite number of problems each week throughout the course ; these are graded, and are returned at the end of the following recitation. The questions of Part V are for the most part answered in the notes or elsewhere in the book; but it has been the writer's experience that the beginner reacts more favorably to concrete questions assigned in advance for study, than to the same ques- tions when put to him for the first time just after he is supposed vi PREFACE to have mastered the principles and details of a specific analytical process. The general directions and discussions of Part I are intended to emphasize those matters, both of theory and practice, which should receive especial attention from the worker in analytical chemistry. It is of course realized that a mere reading of Part I will not go far towards familiarizing the student with its contents ; but it is sought to accomplish this end by referring later on in the text to special subjects as occasion presents. The analyses selected for practice, included in Parts II and III, are those which are comprised in the elementary courses of quantitative analysis at the University of Illinois. They have been chosen as being satisfactory types of gravimetric and volumetric analysis, and, after several years' experience, they are considered to afford to all classes of students a suitable foundation for more advanced work. It is believed that they furnish also a good insight into the methods of quantitative analysis, and hence are adapted to the needs of students who will not extend their study beyond the period of an introductory course. In addition to the help derived from other books and from journal articles, the writer wishes to acknowledge his especial indebtedness in the preparation of this manual to the following works on analytical chemistry: H. P. Talbot's Quantitative Chemical Analysis; J. W. Mellor's Treatise on Quantitative In- organic Analysis; W. F. Hillebrand's The Analysis of Silicate and Carbonate Rocks; F. P. Treadwell's Lehrbuch der analytischen Chemie; W. C. Blasdale's Principles of Quantitative Analysis; and A. Fischer's Elektroanalytische Schnellmethoden. G. McP. SMITH. UNIVERSITY OP ILLINOIS 1918 CONTENTS PART I PAGE A. INTRODUCTION . . . ; i Gravimetric and Volumetric Analysis. B. GENERAL REMARKS CONCERNING QUANTITATIVE WORK . . 3 Neatness ; Accuracy and Integrity ; Economy of Time ; Note- books; Reagents. C. THE OPERATIONS or ANALYTICAL CHEMISTRY I. Weighing - 7 The Balance ; The Use and Care of the Analytical Balance ; Determination of the Zero-point; Methods of Weighing; The Calibration of a Set of Weights; Errors Due to In- equalities in Length in the Beam Arms ; Errors Due to the Buoyancy of the Atmosphere. II. Precipitation 20 Qualities Desirable in Precipitates Which Are to Be Used in Gravimetric Determinations; Colloidal and Fine-grained Precipitates; The Contamination of Precipitates; The Theory of Precipitation. III. Filtration and the Washing of Precipitates . . . . 30 The Selection and Use of Paper Filters; Wash Bottles; Gooch's Filtration Crucible; The Theory of Washing Precipitates. IV. The Drying and Ignition of Precipitates . ... 37 Drying Ovens; Desiccators; Crucibles. V. The Evaporation of Liquids . .... . 41 VI. The Volumetric Measurement of Liquids . . . -43 Volumetric Apparatus ; Necessary Precautions in the Use of Volumetric Apparatus; The Calibration of Volumetric Apparatus. D. THE PREPARATION OF SAMPLES FOR ANALYSIS . . . . 51 vii viii CONTENTS PART II GRAVIMETRIC ANALYSIS PAGE EXERCISES WITH THE BALANCE . . . ..-. . . . . 53 THE DETERMINATION or CHLORINE IN A SOLUBLE CHLORIDE . 54 THE DETERMINATION or IRON AND OF SULPHUR IN A SOLUBLE SUL- PHATE OF IRON . . ... . .", . . ... 59 THE DETERMINATION OF SULPHUR IN AN ORE v . . . 65 THE DETERMINATION OF PHOSPHORIC ANHYDRIDE IN PHOSPHATE ROCK 66 THE DETERMINATION OF CALCIUM AND MAGNESIUM OXIDES IN LIME- STONE . . . . . * . ... . . 70 THE DETERMINATION OF CARBON DIOXIDE IN LIMESTONE . . 76 THE DETERMINATION OF SILICA IN A REFRACTORY SILICATE . . So THE DETERMINATION OF POTASH IN SOLUBLE SALTS '. . . . 84 THE ELECTROLYTIC DETERMINATION OF COPPER .... 86 PART III VOLUMETRIC ANALYSIS GENERAL DISCUSSION . . . . . * ..... . 97 Fundamental Principles ; Reactions Suitable for Volumetric Pro- cesses; Determination of the End-point; General Theory of Indicators ; The Advantages of the Volumetric System ; General Directions. A. NEUTRALIZATION METHODS: ALKALIMETRY AND ACIDIMETRY . 105 Standard Acid Solutions ; Standard Alkali Solutions ; Indicators for Use hi Alkalimetry and Acidimetry. THE PREPARATION AND STANDARDIZATION OF APPROXIMATELY HALF-NORMAL HYDROCHLORIC ACID AND SODIUM HYDROXIDE 109 THE DETERMINATION OF THE TOTAL ALKALINE VALUE OF SODA ASH . . ,,. 113 THE DETERMINATION OF THE NEUTRALIZATION VALUE OF AN Aero 1 1 5 THE DETERMINATION OF PROTEIN NITROGEN BY THE KJELDAHL METHOD . 116 CONTENTS ix PAGE B. METHODS OF OXIDATION AND REDUCTION . * . . . 119 Standard Solutions ; Indicators. 1. BICHROMATE PROCESSES . ^ . ..... . 120 Fundamental Principles. THE PREPARATION AND STANDARDIZATION or APPROXIMATELY TENTH-NORMAL BICHROMATE AND FERROUS IRON SOLU- TIONS . . . . . i 121 THE DETERMINATION OF IRON IN SIDERITE . . . .124 THE DETERMINATION OF CHROMIUM IN CHROME IRON ORE . 125 2. PERMANGANATE PROCESSES 127 Fundamental Principles. THE PREPARATION AND STANDARDIZATION OF AN APPROXI- MATELY TENTH-NORMAL SOLUTION OF POTASSIUM PER- MANGANATE .'.:...'.- 128 THE DETERMINATION OF IRON IN HEMATITE . . .130 THE DETERMINATION OF CALCIUM IN LIMESTONE . . .133 THE DETERMINATION OF THE OXIDIZING VALUE OF PYRO- LUSITE 134 THE DETERMINATION OF PHOSPHORUS IN STEEL . . .135 THE DETERMINATION OF MANGANESE IN AN ORE . . .139 3. IODOMETRIC PROCESSES 142 Fundamental Considerations. THE PREPARATION AND STANDARDIZATION OF APPROXIMATELY TENTH-NORMAL SOLUTIONS OF IODINE AND SODIUM Tmo- SULPHATE . ..... J 46 THE DETERMINATION OF ANTIMONY IN STTBNITE . . .148 THE DETERMINATION OF LEAD IN AN ORE . . . .150 THE DETERMINATION OF COPPER IN AN ORE . . .152 C. PRECIPITATION METHODS . . ... . . . . 155 General Discussion. THE PREPARATION AND STANDARDIZATION OF APPROXIMATELY TENTH-NORMAL SOLUTIONS OF SILVER NITRATE AND POTAS- SIUM THIOCYANATE . .,'... . . . 156 THE DETERMINATION OF CHLORINE IN A SOLUBLE CHLORIDE . 158 X CONTENTS PART IV STOICHIOMETRY PAGE PRELIMINARY DISCUSSION: THE SOLUTION OF TYPICAL PROBLEMS 159 PROBLEMS . . . . . , , . ... .166 PART V QUESTIONS . . . . . . . . . . 178 APPENDIX PREPARATION OF THE REAGENTS 193 SULPHURIC ACID-DICHROMATE CLEANING SOLUTION . . .196 ANALYTICAL SAMPLES FOR THE USE OF STUDENTS . . . .196 APPARATUS IN THE STUDENT'S DESK 197 LOGARITHMS . . .." . 198 ANTILOGARITHMS . 200 INTERNATIONAL ATOMIC WEIGHTS, 1917 . . . Back Cover Sheet AN INTRODUCTORY COURSE IN QUANTITATIVE CHEMICAL ANALYSIS PART I INTRODUCTION A. GRAVIMETRIC AND VOLUMETRIC ANALYSIS QUANTITATIVE analysis has for its object the determination of the quantities of the elements or compounds which are present in particular samples of material. The results are usually ex- pressed in terms of percentage, ordinarily by weight ; but some- times, as in the analysis of gases, by volume. The procedure to be employed in a specific case will often de- pend upon the qualitative composition of the sample. A quali- tative analysis, therefore, should always precede a quantitative, unless the composition of the sample is sufficiently well known. In the performance of quantitative determinations there are two principal methods of procedure, according to which the sub- ject is subdivided into gravimetric and volumetric analysis. In addition, there are gasometric methods, and various physical methods, of analysis ; but these will not be described in this book. In a gravimetric analysis, a weighed sample is taken, and the substances to be determined are separated, one after another, either in the free state, or in the form of suitable compounds. Each final product is weighed, and, from its weight, the weight, and therefore the percentage, of the corresponding substance in the sample can be calculated. The substance to be weighed is in most cases separated from solution by precipitation, though in many instances it is deposited upon a weighed cathode or anode by electrolysis. Sometimes it is separated from other substances by extraction with a solvent, and sometimes in the form of a gas, the weight of the gas being determined either by absorbing it in a weighed quantity of some substance and noting the increase, or by noting the decrease in weight due to the removal of the gas alone. IN QUANTITATIVE ANALYSIS In a volumetric analysis a weighed sample is also taken, but the quantity of the substance to be determined is arrived at by causing some well-defined reaction to take place, the reagent being added from a burette, in the form of a solution of known concentration. This operation is called titration. From the volume of the solution added, it is easy to calculate the weight of the substance present in the sample. In many instances, it is necessary in volumetric analysis also to separate the substance to be determined from interfering substances present with it in the sample ; but, instead of mak- ing a final weighing, the substance is again brought into solu- tion, in suitable form, and its quantity estimated by titration. In order to illustrate the two methods, let us consider the determination of chlorine in sodium chloride. (a) Gravimetric Method. The weighed sample is dissolved in water, the solution acidified with nitric acid, and the chlorine converted into insoluble silver chloride by means of an excess of silver nitrate solution. The precipitate is filtered off, washed, dried, and weighed. From its weight, the weight of chlorine may be calculated, as follows: Cl - X wt. of precipitate = wt. of chlorine. And, of course, wt f chlori " e Xioo = % of chlorine. wt. of sample (b) Volumetric Method. The weighed sample is dissolved in water, the solution acidified with nitric acid, and the chlorine converted into silver chloride by the gradual addition, from a burette, of a silver nitrate solution of known concentration. As soon as, after stirring each time and allowing the precipitate to settle, the first drop is added which fails to produce a pre- cipitate, the reaction is known to be complete ; and the number of cubic centimeters required, multiplied by the chlorine equiva- lent of the silver nitrate contained in each cubic centimeter, gives directly the weight of chlorine in the sample. INTRODUCTION 3 B. GENERAL REMARKS CONCERNING QUANTITATIVE WORK Neatness. The drawers and* cupboards of the desk, and all apparatus, should at all times be ; neat and clean. A sponge or an old towel should always be at hand, and the desk top and filter-stands should be kept dry and clean. Vessels should be scrupulously clean, inside and out, and the outer surfaces of beakers, flasks, etc. should be wiped dry with a clean, lintless towel, before use. If the inner surfaces of funnels, flasks, etc. become contam- inated with a film of grease, they should be rinsed either with a strong solution of sodium hydroxide or with sulphuric acid- dichromate cleaning solution. (The latter may be prepared by pouring, cautiously and with stirring, 4 volumes of con- centrated sulphuric acid (sp. gr., 1.84) into 3 volumes of cold water, and saturating the resulting hot solution, without further heating, with powdered sodium or potassium dichromate.) In extreme cases it may be necessary to allow the apparatus to stand overnight in contact with this solution. Accuracy and Integrity. It is of fundamental importance in quantitative work to guard against loss of material or the in- troduction of foreign matter. All filters and solutions should be kept covered to protect them from dust, and in dissolving substances for analysis, the vessels should always be kept covered to prevent mechanical losses. Success in quantitative work demands first of all a certain amount of dexterity in the performance of the mechanical opera- tions involved. Certain individuals are able to acquire this skill with comparative ease, but the majority of persons can acquire it only through patient and persistent application. If the student finds himself unable to do as good work as his more experienced or more fortunate neighbor, he should rather devote his energies to increasing his proficiency than to trying to con- ceal his lack of it. Nothing less than absolute integrity can be 4 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS demanded of an analytical chemist, and any disregard of this principle is certain to be fatal to his success. Economy of Time. An economical use of laboratory hours is best secured by acquiring a thorough knowledge of the char- acter of the work to be done before undertaking it, and then arranging the work so that no time shall be wasted during the evaporation of liquids and other time-consuming operations. At least two determinations should be in progress at the same time, and confusion should be carefully guarded against by a free use of labels. In general, economy of time results from the evaporation or filtration of several solutions at once; four or more precipitates may often be washed in the time required for any one of them, if taken alone. Notebooks. Notebooks should contain, besides the analytical data, descriptive notes regarding any special difficulties en- countered in the analysis and the remedies applied, and also any incidents in the course of the analysis which might influence the results injuriously. All analytical data, such as records of weights and volumes, should be placed upon the right-hand page, while the left-hand page should be reserved for the descriptive notes, the calcula- tion of factors, of the amounts of reagents required, etc. All analyses should^be made in duplicate, and in general a close agreement in results should be expected. It should, how- ever, be realized that " check results " do not furnish con- clusive evidence of accuracy. Since check results depend almost entirely upon the prevalence of identical conditions throughout the course of the two analyses, they are apt to be obtained even when inaccurate methods of analysis are employed. A common fallacy is to the effect that no part of the work need be performed more carefully than that part which is necessarily least accurate. For example, it is said that if a certain step in a process involves an unavoidable error of 0.1%, it is a waste of time to attempt to avoid errors in other parts of the work amounting to 0.05% or even 0.09%. This unfortunate attitude would lead to the INTRODUCTION 5 conclusion that if a method cannot yield results involving an error of less than 0.10%, it should be given a chance to depart o.io%-hwxo.09%, if there are n other places where errors may occur. Of course these errors may, to a certain extent, counter- act one another in effect, but there is no assurance that they will do so. Nevertheless, when a certain minimum error is unavoid- able, e.g. 0.20%, it is not wise to expend an undue amount of time in trying to prevent other possible errors when the ratio of these errors to the larger error is very small ; because in such a method the percentage result has no significance whatever beyond the first decimal. It is a good rule always to report one decimal place further than the one that is considered to be certainly correct. All records should be dated, and all observations should be recorded at once in the notebook. Records should never be made upon loose sheets of paper. Since the neat and systematic arrangement of the analytical data in the notebook is a matter of the first importance, the following sample right-hand page is given as a suggestion of the manner in which such records should be kept. In the analysis here given, it is uncertain whether the first figure after the decimal should be 4, 3, or 2 ; of course, then, nothing is known concerning the value of the second figure, which therefore is not included in the mean result. DETERMINATION OF CHLORINE IN A SOLUBLE CHLORIDE I Sample Tube, etc. 8.4237 Tube minus Sample 8.2377 Wt. of Sample 0.1860 Wt. of Crucible 5-3588 Crucible+AgCl, ist time 5.7830 2d time 5.7828 Wt. of Crucible 5.3588 Wt. of AgCl 0.4240 Per cent of Chlorine 56.40 Mean Value =56.3% 6 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Reagents. Probably the greatest hindrance to good work in otherwise well-equipped laboratories is the difficulty of secur- ing satisfactory reagents. Also much of the glassware on the market is of an inferior grade and utterly unsuited for analytical work. The habit of carefully testing reagents, including distilled water, cannot be too early acquired ; the most ceaseless vigilance should at all times be practiced in guarding against the presence of impurities which would vitiate the analytical work under way. As is generally known, a " C. P." label is no guaranty whatever of the purity of a reagent, and the " guaranteed " or " analyzed " reagents, sold at high prices, are at times worse than products for which no claim to special purity has been made. Acids of a high degree of purity can be obtained commer- cially, and, although exceptions have been noted, these in most cases need no redistillation. But, owing to its basic nature, ammonia ought always to be redistilled at short intervals, after first shaking it up with slacked lime to remove any carbonic acid. Glass stock bottles may be coated inside with ceresin, to prevent contact between the glass and the ammoniacal solution. Owing to the solvent action on glass of many solutions of solid reagents, these should be made up at frequent intervals in limited quantities, or, preferably, the solid should be dis- solved as wanted. This is particularly called for with such reagents as ammonium oxalate and microcosmic salt, and alka- line " magnesia mixture " should not be kept in contact with glass. The stopper of a reagent bottle should never be laid upon the desk, but should always be held in the fingers until returned to the bottle. This will prevent contamination, whether due to an interchange of stoppers, or to some other cause. The necks and mouths of such bottles should of course be kept scrupulously clean. INTRODUCTION 7 C. THE OPERATIONS OF ANALYTICAL CHEMISTRY The chief operations involved in analytical work which can be profitably discussed at this point are weighing, precipitation, filtration, and the washing of precipitates, the drying and igni- tion of precipitates, the evaporation of liquids, and the volumet- ric measurement of liquids. These operations will be described in the following sections, which should be studied carefully by the beginner. It is of prime importance for success as an analyst to pay great atten- tion to details and scrupulously to avoid any conditions which may destroy the analysis, or lessen confidence in the accuracy of the data. The adoption of the suggestions given will do much to insure work of a high grade, while neglect of them will often lead to inaccurate results and loss of time. I. WEIGHING The purpose of weighing is to compare the quantity of matter in a specific object with the quantity of matter in a given stand- ard a gram or kilogram weight. The comparison is made on the balance by suspending the object to be weighed at one end of a beam, and the weights at the opposite end of the beam. The beam is virtually a kind of lever, and the mechanical theory of the balance is founded mainly on the properties of levers. The Balance. The beam of the balance is supported on a central knife-edge, usually of agate, which rests upon a plane agate plate; and two pans for supporting the masses to be compared are vertically suspended from stirrups, each of which has an agate bearing which rests on a knife-edge fixed at one extremity of the beam. The arms of the balance are so gradu- ated that a rider (of known weight) can be placed on the beam at any required distance from the central knife-edge. If the three knife-edges are allowed to press continually upon their agate bearings, they soon become blunted, and wear fur- 8 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS rows in the bearings. In order to prolong the life of the knife- edges and bearings, the balance is provided with a " release " which separates the knife-edges from their bearings when the balance is not in use. If the balance shows signs of stiffness in the motion of beam and pans, the fault should be investigated at once. The defect may be due to an accumulation of dust between the knife-edges and their bearings; to the blunting of the knife-edges; or to the wearing of furrows in the bearings. To prevent the accumulation of dust, and also to prevent the interference of air currents while weighing, the balance is in- closed in a glass case. In order to render small movements of the beam perceptible, there extends downwards from its center a long pointer which multiplies the rotational displacement. When equilibrium is established, the lower end of the pointer should come to rest in front of the zero of a scale which is located immediately behind this end. The conditions which must be satisfied by a good balance are : (i) The balance must be consistent. It must give the same result in successive weighings of the same body. This condition depends upon the trueness of the knife-edges. (2) The balance must be accurate. At rest the beam must be horizontal when the pans are empty, and when equal weights are placed upon the pans. This condition depends upon the equality of the two arms. (3) The balance must be stable. The beam after being displaced from its horizontal position must return to its horizontal position. This condition depends upon the adjust- ment of the center of gravity. (4) The balance must be sensi- tive. It must show even a very small inequality in the two masses on the scale pans. This condition depends largely upon the length of the arms. (5) The balance must oscillate with reasonable rapidity. Short beams oscillate more rapidly than long ones. The analytical balance will perform excellent service under the proper conditions, but great care in its use is essential if its INTRODUCTION 9 accuracy is to be relied upon. It should be located in a room that is free from dust and fumes, and should stand upon a sup- port that is free from shocks and vibrations. The Use and Care of the Analytical Balance. The following rules embody the main points to be observed in the use and care of a balance. (1) Each student must feel a personal responsibility for the proper use of his balance; carelessness on the part of any one is apt to render inaccurate not only his own work, but also that of all others who use the same balance. (2) The balance pans should be brushed off, if necessary, and the adjustment of the balance tested before use. The balance is properly adjusted only if the following con- ditions are fulfilled: (a) The spirit level or plumb bob inside the balance case should show that the balance is level ; (b) the mechanism for raising and lowering the beam should work smoothly ; (c) the pan arrests should just touch the pans when the beam is lowered; (d) the pointer should rest at zero when the beam is raised, and also when it is lowered so that the pan arrests touch the pans ; and (e) the pointer should swing equal distances on either side of the zero-point when the beam is set in motion without any load on the pans. In the latter case, if the variation does not exceed two divisions on the scale, it is hardly worth while to disturb the balance by an attempt at correction ; it is better to make a proper allowance for the small zero error. (3) The beginner should never attempt to make adjustments himself, but should always apply to the instructor in charge. (4) The beam should never be set in motion by lowering it upon its knife-edge, nor by touching the pans, but rather by means of the rider ; however, there is a " trick " in lowering the pan supports so that the oscillations of the pointer will have the required amplitude. The pans should be arrested and the beam raised before any change is made in the load or weights on the pans except in the case of the small fractional weights, when it is only necessary io INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS to arrest the pans. The object to be weighed and the heavy weights should be placed in- the middle of their respective pans, since a heavy load near the edge of a pan is apt to cause trouble- some oscillations. The beam and stirrups should never be left upon their knife- edges, and the motion of the beam should be arrested only by means of the pan arrests, and only when the pointer is passing the center of the scale ; otherwise the knife-edges become dull and their agate bearings furrowed. (5) The weights should be cared for not less than the balance, and should be standardized by the analyst unless they are known to be in satisfactory condition. The weights should be handled carefully, and only with the forceps provided for that purpose ; they should never be touched with the fingers. In weighing, the weights should always be placed upon the same pan, and they should be taken in the order in which they occur in the box, the larger ones first; and the weight of the object should be recorded by noting the vacant spaces in the box. The record so obtained should be checked as the weights are removed from the pan. In this way errors are not likely to occur. (6) No analytical sample should ever be placed directly upon the balance pan. Furthermore, the object to be weighed should neither be warmer nor colder than the air in the balance case. Currents of hot air may impinge on the arms of the balance and buoy up the beam, or cause one arm of the balance to expand unequally. 1 If the object is colder than the atmosphere of the balance case, moisture may condense on its surface. If the body to be weighed is likely to be electrified (e.g. a glass weigh- ing tube), it should be allowed to stand for some time after it has been wiped, before weighing. (7) The balance case should be closed while weighing with the rider, so as to avoid currents of air. 1 For instance, a platinum crucible which appeared to weigh 20.649 g. when warm, weighed 20.6920 g. when cold. INTRODUCTION n As soon as the object is apparently balanced by the weights, the beam should be raised and again lowered into place, and the observation repeated. This will assure the proper alignment of the beam and pans at the time when this is most important. (8) In using weighing bottles or tubes, the vessel should be weighed together with its contents. A quantity suitable for analysis should then be "removed without loss, and the vessel and contents again weighed. The difference in weight indicates the quantity of sample taken. Cork stoppers in weighing tubes are apt to change in weight, owing to varying amounts of moisture absorbed from the at- mosphere. It is therefore necessary, before weighing out a new sample from it, to confirm the recorded weight of a tube which has been unused for some time. (9) Errors in weighing should fall well within the limits of the experimental error due to the analytical operations. If, for ex- ample, an error of o.ooi g. were made in weighing out a gram sample of clay containing 0.20% of MgO, the resulting error in the determination of the magnesia could be no greater than 0.1% of its value ; the final result could not be affected by more than 0.1% of 0.20%, i.e. 0.0002%. This is negligibly small. Suppose, however, that an even smaller error of 0.0005 S- were made in weighing the 0.0055 g- f Mg 2 P 2 7 ; this would repre- sent an error of 9% of the magnesia value, which is inexcusable. (10) If any substance is spilled upon the pans, or if anything at all appears to be the matter with a balance, the fact should at once be reported to the instructor in charge. In most in- stances serious injury can be averted by prompt action. Determination of the Zero-point. Lower the beam and stirrups upon the knife-edges by slowly turning to the left the milled head at the front of the balance case. Then release the pan supports by gently pressing inwards the small button, also at the front of the case, and with the beam swinging smoothly, make a consecutive record of the number of scale divisions traversed by the pointer on either side of the center. Record 12 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the swings to the left as negative numbers and those to the right as positive numbers, making four observations on one side and three on the other. Take the average of each column, add these averages algebraically, and divide the sum by two. The quotient is the zero-point of the balance, i.e. the position on the scale at which the pointer would finally come to rest. Example : LETT RIGHT -6.8 -6.6 -6.4 -6.3 +4.7 +4-5 +4-3 Average: 6.5 Average : +4.5 Zero-point = i.o. Two methods of procedure are now open to the operator. He may either make his weighings with reference to this ob- served zero-point or he may adjust the balance so that the ob- served zero-point is the actual zero of the scale. The first method is preferable, unless the zero-point is more in error than one scale division. The zero-point is apt to change, and it must be determined each day, or even more often. Methods of Weighing. Weighings smaller than 0.005 g- ( r o.oi g.) are made with the rider. When the arms are divided into five divisions, a 5-milligram rider is used; in general, the rider should weigh as many milligrams as there are large divisions on the beam between the central knife-edge and the right-hand stirrup support. Each division on the beam then corresponds to a milligram. Ordinary Method. The object to be weighed is placed upon the left-hand pan of the balance and weights upon the right-hand pan, until, finally, the further addition of 5 mg. (or 10 mg.)more than counterbalances the object. This weight is then removed, the balance case closed, and the rider adjusted so that the pointer swings equal distances on either side of the zero-point. This INTRODUCTION 13 method of weighing is very common, and it is sufficiently accurate for ordinary analytical work. If necessary, the zero-point of the unloaded balance should be determined before each weighing. In special cases, as in the calibration of a set of weights, it is important to make more accurate weighings. It is here best to use the method of weighing by double vibrations, which from the following description may appear somewhat laborious; but the labor is more apparent than real. Method of Weighing by Double Vibrations, (a) Find the zero-position of the pointer in the case of the unloaded balance, according to the method already described. Let us suppose this to be at +0.1. (6) Find the deviation of the scale per milligram, that is, the sensitivity of the loaded balance. The object to be weighed is placed upon the left pan, the weights on the right pan. When the weights are so far adjusted that an additional 0.005 g- ( r o.oi g.) would be too much (e.g. weight on pan = 11.216 g.), close the door of the balance case, and adjust the rider until the pointer swings on both sides of the zero of the scale. Now find the position of rest, which we will suppose to be at +0.8. Move the rider one milligram division to the right, and again find the position of rest; this being at, say, 2.1. Hence, the zero-point is displaced -fo.8 ( 2.i) = 2.p divisions by increas- ing the weight i milligram; or 2.9 scale divisions correspond to i milligram for the given load. (c) Calculate the weight of the load on the pan. From the preceding results, it follows that the load weighs 11.216+0;. The zero-point of this load is displaced 0.80.1=0.7 scale division. Since 2.9 scale divisions correspond to i milligram, 0.7 scale division will correspond to =0.24 mg. Hence the 2.9 weight of the body is 11.216+0.00024 = 11.21624 g. These calculations may be summarized in the formula Correction = + r mg., a o 14 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS in which z represents the zero-point of the unloaded balance; a, the zero-point with not quite enough weight on the right pan ; and b, the zero-point with a milligram more on the right pan than corresponds to a. Analytical balances will rarely indicate with certainty less than o.oooi g. Hence, although the weight may be calculated as above to the fifth decimal, it should generally be rounded off by dropping the fifth decimal and raising the fourth decimal one unit when the dropped figure exceeds 5. In certain cases, as in the calibration of volumetric measuring apparatus, it is necessary for the weight found to be independent of any inequality in length in the beam arms. In such cases, and in the determination of absolute weights (reduction to weights in vacua), one of the following methods should be used. Method of Gauss. Weigh the object first in one pan, then in the other. Let W be the true weight, a the weights required to counterbalance the object when it is on the left pan, and b the weights required when the object is on the right pan. Accord- ing to the principle of moments : = ar That is, WHr = ablr^or W 2 = ab ; whence W Vab. Therefore the true weight is the square root of the product of the two observed weights. Borda's Method of Weighing by Tares. Here the object, placed on the right-hand pan, is balanced by a suitable tare (weights, wire, beaker containing shot, etc.) on the left-hand pan. The object is then removed, and weights are added in its place until equilibrium is restored. These weights are necessarily the same in value as the object for which they substitute, irre- spective of differences in the arms. The Calibration of a Set of Weights. Fairly accurate weights can be purchased for a reasonable sum, and for most analytical work the inaccuracies of the better class of weights are negligibly INTRODUCTION 15 small in comparison with the errors of experiment, and the im- perfections in the analytical processes. An analyst, however, should know that his weights are suffi- ciently accurate, and for this reason he should calibrate the weights. The errors due to imperfections in the weights can easily be reduced to o.oooi g. The weights should be tested at periodic intervals, say once or twice a year, depending upon the frequency with which they are used. In special cases, e.g. in the calibration of volumetric apparatus, absolute weights may be required, but for general analytical work absolute weights are not necessary. If the weights are consistent with one another, their absolute values have no in- fluence upon the accuracy of an analysis. Before beginning the calibration, distinguish all separate pieces of the same denomination by marking them with a small prick punch. One of the two lo-gram pieces may be marked ('), two of the i -gram pieces (') and ("), etc. The method of weighing to be followed in the calibration will depend upon the degree of equality in the lengths of the beam arms. If they are unequal, either the method of Gauss or that of Borda may be used for comparing the weights. If the method of Borda is used, a second set of weights will be found convenient for the substitutions. This method involves less work in calculating then does that of Gauss. If the beam arms are essentially equal, the simple method of double vibrations is used, without the necessity of a correction. In any case, the following comparisons are made, with the use of the rider to obtain equilibrium. I GRAM WEIGHTS i against i' 1 against i" 2 against i + i' 5 against 2-fi-fi'+i" 16 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS 10 against 5+2+1 + 1'+!" 10 against 10' 20 against 10+10' Etc. II FRACTIONAL WEIGHTS 500 against 200+ 100+100'+ 50+ 20+10+10' +5+rider at 5 200 against 100+ 100' 100 against ioo /l 100 against so+2o+io+io'+5+rider at 5 50 against 2o+io+io'+5+rider at 5 20 against 10+ 10' 10 against 10'. 10 against 5+ rider at 5 5 against rider at 5 Also unmarked i-gram piece against all of the fractional pieces + rider at 5 The calculation of the weight of each piece of i gram and upward is made upon the arbitrary assumption that the un- marked i -gram weight is correct. In calculating the weights of the fractional pieces, we first assume the unmarked lo-mg. piece as a standard and calculate provisional weights for each of the other fractional pieces upon this basis. We then add these provisionally corrected weights and determine by compar- ing the results with their collective weight as found in terms of the standard i-gram piece, how much each weight must be further corrected. If, for example, the sum of the provisional fractional weights were found to be 5 (rider) + 5 + 10+10+ 19.9 +49.7 + 100+100.1 + 200.1+499.2 = 999.0 mg., or 0.9990 g., while their collective weight in terms of the i-gram standard = i.ooi7g., then each of the provisional values should be multi- plied by =1.0027. I* 1 thi g way, we obtain the weight of 0.9990 each piece in the set in terms of the unmarked i-gram weight. If desired, we can then find the exact value of the unmarked i-gram piece in terms of an absolute standard weight. For example, if the unmarked i-gram weight is found by comparison INTRODUCTION 17 to weigh 0.9998 g., we have simply to multiply each weight in the table, based upon the unmarked i-gram weight as a standard, by 0.9998, in order to obtain the weight of each separate piece of the set in terms of the absolute. standard. Errors Due to Inequalities in Length in the Beam Arms. In the preceding discussion, it has mainly been assumed that the two arms of the beam are equal in length. This is not really the case. It is mechanically impossible to insure perfect equality. To find the relative lengths of the arms, place (corrected; weights of the same nominal value say, 50 grams upon each pan, and bring the balance into equilibrium by means of the rider. Interchange the weights on the two pans, and again bring the balance into equilibrium by means of the rider. Call the two weights W and w r and let / and r respectively denote the additional weights re- quired for equilibrium on the left and right sides. Then, on the first weighing, w+l = W ; and, on the second weighing, W = w+r. Let L and R respectively denote the length of the left and right arm. Then from the law of levers, L(w+t) = RW; zudLW = R(w+r) Solving each of these equations for W ', and equating the results, we find that whence, Suppose, for example, that the weighings were found to be : LEFT RIGHT 50 = 2o+io+io'+io"+o.i3 mg. Here, then, /= 0.00013, r = +0.00019, an d ^ = 50 g. Conse- quently if in the above expression we let R = i, we have r \w+r L = A I = i . 000003 2 \w+l i.e. L : R = i .0000032 : i i8 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS With this ratio, =1.0000032, a weight w on the left pan of K. the balance will be equivalent to a weight wX 1.000003 2 on the right pan. Hence, if a substance on the left pan balances the weight 50 g. on the right pan, the weight of the substance is 50.0000X1.0000032 = 50.0016 g. an error of only 0.003 P er cent. There is therefore no need to apply the correction. Each balance has its own constant L : R for a given load ; the numer- ical value of the ratio varies with the different loads. Most analytical balances do not require a correction on account of inequalities in the arms ; the arms are usually made sufficiently exact. Furthermore, if the weights are always placed, say, on the right pan, such a correction is unnecessary in ordinary analytical work, because, although there may be differences between the apparent and true weights of the substances weighed, these differences are proportional to the true weights and there- fore do not affect the ratios obtained. Errors Due to the Buoyancy of the Atmosphere. The assump- tion is made that, if two bodies are equal in weight at the same tune and place, they contain the same mass or quantity of matter. This is only true if the two bodies have the same volume, or if the weighing is carried out in a vacuum. A body weighed in air is buoyed up by a pressure equivalent to the weight of the air which it displaces. Suppose that exactly 100 grams of platinum (sp. gr. 21.55) are weighed in air with brass weights (sp. gr. 8.4). Then 4.5 cc. of air, at say 20 and 760 mm., i.e. about 0.0054 g., are displaced by the platinum; while the weight of the air displaced by 100 grams of brass is 0.0143 g. Hence, the weight of brass which exactly counterpoises 100 grams of platinum is 100+ (0.0143 0.0054) = 100.0089 g. The buoyancy of the air thus produces a sensible effect whenever the volume of the load differs materially from the volume of the weights. 1 1 To eliminate the buoyancy correction due to variations in temperature and pressure during the same experiment, it is customary, in weighing bulky glass apparatus (potash bulbs, etc.), to use a similar piece of apparatus as a counterpoise. INTRODUCTION The arithmetic of the above calculation is summarized in the formula : Corrected weight =w-\-co ( ) \5 SiJ in which w represents the apparent weight of the object ; s, the specific gravity of the object; si, the specific gravity of the weights ; and , the weight of a cubic centimeter of air under the conditions prevailing at the time of the experiment. To illustrate the effect of the buoyancy of air on the different substances usually weighed in clay analyses, the following table may be quoted : ERROR PER GRAM OF SUBSTANCE WEIGHED SUBSTANCE WEIGHED GRAVITY With Brass Weights With Platinum Weights Clay . 2.^ O.OOO3 C.OCO4. Silica 2.23 0.0004 o 0005 Aluminum oxide .... Ferric oxide 3.85 ^.12 O.OOO2 O.OOOI 0.0003 O.OOO2 Magnesium pyrophosphate . Calcium oxide 2.40 2 QO 0.0003 o 0003 O.OOO4 o 0004 Potassium chloride .... Sodium chloride I.Q9 2.13 0.0004 0.0004 0.0006 o 0005 Potassium chloroplatinate 3-34 O.OOO2 0.0003 In ordinary analytical operations we have to deal with differ- ences in weight, and with ratios, not with absolute weights. When the amount of a precipitate is determined from the dif- ference in the weight of an empty crucible and of the crucible plus the precipitate, the buoyancy correction is not needed for precipitates with a specific gravity near that of the substance undergoing analysis. If, however, the specific gravities are widely separated, it may be worth while to correct for buoyancy. For instance, since the specific gravities of pyrites and barium sulphate are nearly equal, it would be a waste of time to correct for buoyancy in determining sulphur in a sample of pyrites. 20 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS On the other hand, in standardizing a solution of silver nitrate by precipitating silver chloride from a specific weight of the solution, the buoyancy of air may affect the result by xtr of one per cent. Summary. The foregoing discussion serves to show that, in work involving delicate measurements, it is always advisable to make an estimate of the influence of the various sources of error on the final result. These errors can only be neglected when their effect is small in comparison with the error derived from other sources. The chief sources of error commonly intro- duced in the balance room are those arising from : (i) varia- tions in the zero-point of the balance ; (2) inconsistent weights ; (3) inequalities in length in the beam arms ; and (4) the buoy- ancy of the air. In weighings making any pretense to " accuracy to the TO milligram/' the following points should be noted : (1) The zero-point of the unloaded balance should be deter- mined and made use of in each weighing. (2) The weights should be calibrated, and periodically checked for consistency among themselves. (3) The errors due to inequality in length in the beam arms can be neglected in ordinary analytical work. (4) The correction of the weighings for the buoyancy of air is necessary when the determination involves the weighing of substances with appreciably different specific gravities. 1 II. PRECIPITATION Qualities Desirable in Precipitates Which Are to be Used in Gravimetric Determinations. Precipitation is made use of more often than any other means for the separation of inorganic substances. But, in carrying out such separations, precipitates should conform as nearly as possible to the following ideal specifications: (i) The precipitate should be insoluble in the x ln general analytical work this correction can almost always be neglected, since the resulting error is usually overshadowed by the errors associated with the preparation of the precipitates for the balance. ..*+ INTRODUCTION 21 mother liquid, and also in the wash liquid to be used ; (2) it should be compact, easy to filter and wash ; (3) it should be a pure chemical substance of known percentage composition; and (4) it should either be stable and non-volatile on heating, or it should yield upon ignition a pure, non-volatile substance of known composition. The last two conditions are of especial importance if the precipitate is the substance finally to be weighed. Moreover, other things being equal, it is conducive to accuracy if a precipitate can be obtained which contains a low percentage of the substance under investigation (cf . Part IV, Problem 46). But few processes satisfy all these requirements, and in the case of any analytical process it is important to know what conditions favor and what conditions hinder the separation and purification of a given precipitate. There are a few general principles of such wide applicability that they should be con- stantly borne in mind. Their discussion follows. Colloidal and Fine-grained Precipitates. Finely divided precipitates, such as newly precipitated silver chloride, barium sulphate, calcium oxalate, etc., are particularly liable to pass through the filter ; furthermore, they tend in large measure to stop up the pores of the filter, and thus to increase the time required for filtration and washing. Hence, the analyst employs various artifices in order to enlarge the size of the particles. (i) The grain size can frequently be increased by allowing the fine grains which originally separate to digest in the pre- cipitation liquid. This change is more rapid in the hot, than in the cold, mother liquid. In the case of crystalline substances, it often happens that the finer grains, which (owing to differences in surface tension) are somewhat more soluble than the coarser ones, redissolve; and since the solution is then supersaturated in respect to the coarser grains, these are augmented in size by the surface deposition of the material furnished by the finer grains. 22 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS The boiling of liquids containing colloidal substances fre- quently leads to the flocculation of a large number of fine par- ticles into a smaller number of coarser aggregates. (2) Precipitates produced in hot solutions are often coarser- grained than if produced in cold solutions. From what has just been said, the reasons for this fact will be apparent. (3) The flocculation of a precipitate which separates in a colloidal condition is frequently caused by the salts present in the mother liquid. When these salts have been almost removed, during the washing, the colloidal precipitate is apt to be defloc- culated, and it may then give a turbid filtrate, or become so slimy as to be almost impermeable to the wash liquid. In such cases it is necessary to wash the precipitate either with boiling water, or, better, with the solution of an electrolyte which will prevent the deflocculation of the precipitate, and which can be easily removed by drying or ignition. Sometimes dilute acids can be used, but usually, for obvious reasons, we have to de- pend upon volatile ammonium salts. The Contamination of Precipitates. Finally, it should be noted that the finer the grain of the precipitate, the greater will be the quantity of contaminating salts likely to be retained by the wet precipitate. The salts appear to be retained by a kind of surface attraction, called adsorption, 1 and, since fine-grained precipitates expose a larger surface of separation between the solid and the liquid phases, and also because of their compact- ness, the fine grained precipitates are more difficult to wash clean than those of coarser texture. Colloidal gelatinous pre- cipitates like ferric and aluminum hydroxides are in an ex- tremely fine state of subdivision, and, in consequence, they are most difficult to wash clean. In addition to their tendency to be contaminated by the adsorption of salts, precipitates are also frequently liable to contamination, owing to the formation during precipitation of 1 But see "The Contamination of Precipitates in Gravimetric Analysis," G. McP. Smith : Journal of the American Chemical Society, vol. jp, pp. 1152-73 (1917). INTRODUCTION 23 more or less stable insoluble complexes (and, in rare cases, possibly, to the carrying down of foreign substances by the precipitate in a state of solid solution) . These impurities, in what- ever form they may be present, cannot be completely removed by washing, and the wash water will frequently fail to show any in- dication of the impurities which are still present in the precipitate. It is therefore often advisable to redissolve the precipitate, and to repeat the precipitation. The objectionable impurity divides itself in a more or less definite concentration ratio be- tween the precipitate and the mother liquid. A relatively large amount may be retained by the precipitate in the first precipita- tion, but in a second precipitation, when only that amount of salt retained by the first precipitate is in solution, the division of the undesirable substance between the precipitate and the solution in the given concentration ratio means that a much smaller quantity of impurity will be retained by the second precipitate. Repeated precipitations will, in general, soon carry the amount of impurity outside the range of the balance ; but, in carrying out such operations, the solubility relations of the precipitate itself should never be left out of consideration. The Theory of Precipitation. Reversible Reactions. The reactions which are made use of in analytical chemistry belong for the most part to the reversible type. Instead of running to completion, the system may take up a state of equilibrium be- tween the initial stage and that of the completed reaction, and a certain quantity of the substance under investigation is apt to escape our notice. This is especially true in many reactions involving precipitation, neutralization, oxidation, etc. Furthermore, many reactions are more or less influenced by the presence of certain substances, and it is obvious that a thorough knowledge of the processes which take place in such cases will be of the greatest service to the analytical chemist. Therefore, it is of primary importance in analytical chemistry to study each process thoroughly in detail, with the object of finding out and understanding the conditions which will be most favorable 24 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS for the practical completion of each and every reaction involved in the process. ' Indispensable guides for such studies are found in the ionic theory and the law of mass action. It is taken for granted that, at this point, the student is already sufficiently familiar with the qualitative conception of ionization. Degree of Ionization. In a dilute aqueous salt solution, the greater, 1 and by far the most active, portion of the salt is almost invariably ionic. But with acids and bases there is a wider range, and of these a larger number are less highly ionized ; but even here the ions are nearly always much more active than the non-ionized molecules. The acids and bases that are commonly called " strong " are highly ionized, i.e. their solutions are especially active as acids or bases because they contain high hydrogen, or hydroxide-ion, concentrations. Composition of the Ions. It is usual to assume the simplest possible compositions for the ions formed upon the dissociation of any given electrolyte. A more careful study of the subject, however, shows that the ionization of even simple electrolytes may be a very complicated process. It is known, for example, that sulphuric acid contains ions of the formula HS0 4 ~, in ad- dition to SO 4 ions, and that phosphoric acid yields ions of the formulas H 2 P0 4 -, HPO 4 , and PO 4 . All of these are probably more or less highly hydrated; even hydrogen and hydroxide ions are supposed to be hydrated in aqueous solu- tion. Further, many metallic ions show a decided tendency to exist in combination with certain molecules and radicals, as OH2, OH, NH 3 , NH 2 , CN, C 2 O 4 , P0 4 , Cl, etc. ; but in very dilute solutions these complexes are apt to be more or less highly dissociated into their constituents. The Law of Mass Action as Applied to Ionic Equilibria. In aqueous solution, acetic acid is supposed to ionize as follows : HC 2 H 3 2 Z H++C 2 H 3 2 - 1 Noteworthy exceptions are mercuric chloride and cyanide, lead acetate, and a few others. INTRODUCTION The quantity of the molecular acid that is ionized per unit of time in a given volume of the solution is proportional to the concentration of the non-ionized molecules, Cnc 2 H 3 o 2 , while the quantity of the molecular acid that is simultaneously formed by the union of the ions depends upon the frequency of the en- counters of the two kinds of ions, which in turn is proportional to the product of their concentrations, C H +xC c ,H s02 -. The speeds of the respective actions will therefore be 5*1 = CHc 2 H 3 o 2 X FI and .5*2 = C H + X Cc 2 H 8 o 2 - in which FI represents the intrinsic tendency of HC 2 H 3 O 2 to ionize, and FZ that of H + and C^H-^Oz' to combine. When equal amounts of material are being transformed each way, i.e. at equilibrium, Si = S 2 , and therefore CnczHgOz XFi = C H + XCc 2 H,o,- or C H + xCc 2 H 8 o 2 - _ FI _ & -- ft- -^, being the ratio of two constants, is constant; and the value, rz k, of this ratio of the affinities driving the opposed actions is called the affinity constant of the reversible reaction. At any given temperature, provided the solution is dilute, the numerical value of k remains the same no matter what the total concentra- tion of the solution may be. 1 In the case of acetic acid, for example, the following figures have been obtained, at 18, from conductivity determinations. TOTAL MOLAL CONCENTRATION OF AGED PROPORTION IONIZED MOLAL CONCENTRATION OF H+ AND OF AC- (Cn+ AND Cc 2 H 3 o 2 -) MOLAL CONCENTRATION OFHAc (CHC 2 H 3 2 ) I.OOO O.IOOO 0.0100 0.0041 0.0130 0.0407 O.OO4I O.OOI3O O.OOO407 1.000-0.0041 O.IOOO-O.OOI3O 0.01000-0.000407 1 When data such as the following are applied to cases of soluble, highly ionized substances, the ^-values so obtained for any given compound are usually far from constant. The general conclusions arrived at through the application of such data are, however, as a rule, not invalidated by this fact. 26 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Substituting these figures in equation (i) above, we get : (0.0041)2 , (o.oois) 2 -V^ =0.0000160 ; =0.0000171 ; 0.996 0.0987 (0.000407) 2 and =0.0000172. 0.00959 It is seen that, although the third solution is a hundred times more dilute than the first, and although the degree of ionization has increased tenfold, the value of k is the same in both cases. The Common-Ion Effect. When, through the presence of two substances which furnish an ion in common, the concentra- tions of the positive and negative ions of an ionogen are unequal, the law of mass action still holds. Let us imagine, for example, that, by mixing equal volumes of the double-molal solutions, a solution is obtained which is uni-molal in respect to acetic acid and also to sodium acetate. Let us further suppose that the equilibria which exist in the mixture have been established in two separate stages, as follows : (i) that the concentrations of each undissociated compound and its ions have changed from those which exist in a double- molal, to those which exist in a uni-molal solution of that com- pound; and (2) that the concentrations of all the substances present have changed from those which exist in the separate uni-molal solutions of the compounds, to those which exist in the mixture which is uni-molal in respect to each compound. Let us now consider this latter stage in detail. In uni-molal solution, sodium acetate is 0.53 ionized, while acetic acid at that concentration is only 0.004 ionized. Each com- pound furnishes acetate ions, and the acetate ions present are all available, either for union with sodium ions, or for union with hydrogen ions. Initially, therefore, in the case of the sodium acetate, we have a53X ' 534 >k lt instead of '53Xo. 5 3 =fa 0.47 0.47 but the two expressions are so nearly identical that we see at a glance that the ionic equilibrium of the salt will not be INTRODUCTION 27 affected appreciably by the presence of the acid. In the case of the acetic acid, however, we have the initial relationship, 0.004X0x34 t j 0.004X0.004 , . - ^ ^ = 133 k. instead of - - = k. Since, at equi- 0.996 -0.996 librium, the fraction -^ - - 2 ^- remains constant, and since, owing to the low H + -ion concentration, CHC^O, cannot be in- creased appreciably, nor Cc 2 H 3 o a - be appreciably diminished, by the formation of the molecular acid, it follows that the value of C H + must be lowered to about y^ of its initial magnitude. That is to say, the sodium acetate in this solution diminishes the hydrogen-ion concentration from 0.004 to about 0.00003. The student should especially note that the concentration of a given ion can be lowered in this way to a value approximating zero only when that ion unites with an ion added to form a substance which is insoluble or which by nature has only a very slight tendency to dissociate. We might add sodium chloride in the hope of repressing the ionization of hydrochloric acid ; but, since both compounds ionize highly, we should obtain no appreciable effect. If, however, we add sodium acetate in excess to hydro- chloric acid, we can obtain a solution which is as weakly acid as the one discussed above. 1 The Solubility Product. One of the commonest and most interesting applications of the law of mass action is met with in connection with the precipitation and solution of relatively insoluble salts. Every substance possesses, when immersed in a liquid, a certain solution-tension, by which is meant an expansive force which tends to drive particles of the substance outward into the liquid. These particles move in every direction, and conse- quently some of them return to the solid and reattach them- 1 For example, i mol of HCl-f 2 mols of NaC 2 H 3 O 2 , in a volume of i liter, give a solution which is uni-molal in respect to acetic acid, to sodium acetate, and to sodium chloride. The hydrogen-ion concentration of this solution would also approximate 0.00003. 28 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS selves to it. This occurs the more and more frequently, as the concentration of the particles in the liquid increases, until, finally, a stage is reached at which the number of particles leaving the solid per unit of time is equal to the number deposited upon it. When the entire liquid is equally charged with dissolved particles, the liquid immediately surrounding the solid will lose none by diffusion, and a condition of equilibrium will be estab- lished. At any given temperature, the quantity of dissolved solute will remain thereafter unchanged, no matter how long the materials are left in contact. It is at this point that the solution is said to be saturated with the dissolved substance. In the case of silver bromate in water, we have the following scheme of equilibria : AgBrO 3 ^1 AgBr0 3 ^: Ag++Br0 3 - (solid) (dissolved) The solid AgBr0 3 molecules tend to enter the solution, while at the same time dissolved AgBr0 3 molecules tend to come out of solution, and the solution is saturated when these tendencies produce equal effects. The ions themselves (and any foreign materials present) are not supposed to take any direct part in the equilibrium which controls solubility. That is, in solutions saturated at a given temperature by a given solute, the concentra- tion of the non-ionized molecules will be constant no matter what other substances may be present, provided that the quantities of all the dissolved substances are not sufficient to alter the nature of the solvent. The total solubility of an ionogen, as we ordinarily use the term, is made up of a molecular and an ionic part. The quantity of the latter does not remain constant when a foreign substance giv- ing a common ion is added to the solution. In a solution of silver bromate, for example, we have the mathematical relationship : =k ' r (Ag+) x(Br0 3 -) =* x(AgBrOs). But, since, in the special case of a solution which is saturated with the salt at a given temperature, the concentration of the non- INTRODUCTION 29 ionized molecules, (AgBrOs), remains constant, it follows that the product, &x(AgBr0 3 ), also remains constant, or that in a saturated solution of a given slightly soluble ionogen the product of the concentrations of its ions is constant. This product is called the solubility product, because the two separate values jointly determine the magnitude of the total solubility of the ionogen. The concentration of the non-ionized molecules cannot be diminished, but the ionic part of the solute may become vanish- ingly small if the concentration of the common ion is made great as compared with that of the other ion of the solute. The relationships which exist in the case of silver bromate are illus- trated in the following table, where it will be seen that the ex- perimental values agree remarkably well with the calculated ones. SOLUBILITY OF AcBROs IN MOLS PER LITER MOLS PER LITER OF COMMON-ION SALT ADDED SOLUBILITY FOUND SOLUBILITY CALC. Addition of Silver Nitrate Addition of Potassium Bromate (Addition of either Salt) O 0.00850 0.0346 O.OoSlO O.OO5IO O.O02I6 0.008 1 o O.OO5I9 0.00227 0.00504 0.00206 The theory of the precipitation and solution of slightly soluble ionogens may be summed up as follows : 1 1 That is, of uni-univalent ionogens. In other cases, the solubility product would often contain ion-concentrations raised to the second, third, etc., powers; but in reality the question is very much complicated by interfering reactions. Thus, in the case of PbCU, if Nad is added to the saturated solution, some PbCU will be precipitated in accordance with the theory ; but the addition of Pb(NOs)2 actually increases the solubility of the PbCU- This is probably because of the formation of complexes, or of intermediate ions, such as PbCl~, or of both, whereby the addition of the salt giving the common bivalent ion may not only fail to increase the concentration of the bivalent ion, but may even lower that of the univalent ion. At any rate, enough is known to indicate that the theory may not be so much at fault as we ourselves, in our lack of methods for finding out just what ions and complexes are present in such solutions, and in what quantities. 30 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS If the product of the concentrations of any pair of ions in a solution is made to exceed the solubility product of the ionogen formed by their union, the latter will be precipitated until the ion- concentration product has been reduced to its solubility-product value. And conversely, if the ion-concentration product of any pair of ions in a solution is made less than the solubility product of the ionogen formed by their union, the latter will, if present in sufficient excess, continue to dissolve until the ion-concentration product has been increased to its solubility-product value. III. FILTRATION AND THE WASHING OF PRECIPITATES The purpose of nitration is to separate a solid from a liquid in which it is suspended. This is effected by causing the liquid to pass through a porous medium compact enough to retain the solid. The most important media in use are filter paper, as- bestos pulp, and platinum sponge. The Selection and Use of Paper Filters. Three qualities which are desirable in a filter are : (i) porosity, to insure rapid nitration; (2) sufficient compactness, to insure complete re- tention of the precipitate; and (3) low amount of ash. In quantitative work, only filters should be used which have been treated with hydrochloric and hydrofluoric acids, and which, on incineration, leave a small and definitely known weight of ash. Such filters are readily obtainable in the market. Rapid (porous) filters should be used for all precipitates which do not readily pass through the paper ; the slow, compact papers should be used only when necessary. A tremendous amount of time is consumed, often wasted, in the filtration and washing of precipitates. The size of the filter paper should be determined by the magni- tude of the precipitate, and not by the volume of the liquid to be filtered. A precipitate should not fill the paper more than half full, but if too large a paper is used, time is wasted in washing the filter. The filter, as well as the precipitate, has the property of retaining certain salts very tenaciously. INTRODUCTION 31 Funnels should be selected which have an angle of 60, with a narrow stem about eight niches long. The filter should be accurately folded to fit the funnel, and the top of the filter should be at least one half centimeter below the edge of the funnel. On no account should the paper project beyond the edge of the funnel. Place the filter in the funnel, wet it, and carefully bed it against the walls of the funnel. When the filter is filled with distilled water, the stem of the funnel should fill with a column of water, 1 and air should not pass between the funnel and the paper as the latter empties. When the filter is properly bedded, water should flow through it quickly, and filtration will usually pro- ceed quite rapidly ; at any rate, the paper will then do its best. The liquid at the apex of the filter is under a pressure approxi- mately equal to the weight of a column of water of the same diameter as the bore of the stem and of a height equal to the length of the stem plus the depth of liquid in the filter. When paper filters are employed, the use of a vacuum pump to promote filtration is of doubtful advantage in quantitative analysis. The increased tendency of precipitates to pass through the filter more than offsets the possible gain in time. Whenever suction is applied, a more compact paper should be used, or the point of the filter should be supported by a perforated platinum cone. The vessel used to receive the initial filtrate should invariably be replaced by a clean one, properly labeled, before the washing of a precipitate is begun. Precipitates which at first show no tendency to pass through the filter may enter into colloidal solu- tion as the washing proceeds. The advantage gained in such an instance by having removed the first filtrate is obvious. The precipitate is generally allowed to settle before filtration. The clear liquid should not be poured directly on to the filter, 1 If this fails to take place, either the stem of the funnel is too wide, or it is not free from grease. The latter can be removed by means of warm sulphuric acid- dichromate cleaning solution. 32 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS but down a glass rod, from which the stream should be directed towards the side, not the center of the filter paper. The re- ceiving vessel for the nitrate should be placed so that the liquid from the funnel will run down its side ; otherwise there is danger of loss by splashing. As far as possible, the precipitate itself should be kept in the precipitation vessel. Much time can generally be saved by washing the precipitate by decantation, i.e. by adding to it successive portions of wash liquid, allowing the precipitate each time to settle, and decant- ing the clear supernatant liquid through the filter, without un- duly disturbing the precipitate. This procedure is especially advantageous in the case of precipitates which tend to clog the pores of the filter. Finally the precipitate may be transferred to the filter, and the washing completed there. It will always be found that small portions of the precipitate adhere to the walls and bottom of the containing vessel. These can be rubbed loose by means of a so-called " policeman," a piece of soft rubber tubing tightly fitted on the end of a glass rod. Pieces of rubber tubing with closed ends are sold for the purpose. These rubber-tipped rods should be used only for the above-mentioned purpose; they should never be allowed to stand in analytical solutions, nor should they be used as ordinary stirring rods. Precipitates should never be allowed to dry before they have been completely washed. They are likely to shrink and crack, and, on further washing, the liquid will pass through these channels only ; this is especially true of gelatinous precipitates. Every original filtrate must be properly tested to insure com- plete precipitation, and the wash waters also must be examined. It is useless, however, to test the latter until several washings have been made. Only a few drops should be taken if the filtrate is to be used for a subsequent determination; but when the washing is nearly finished, at least 2 or 3 cc. should be used. The necessity of making these tests cannot be too strongly impressed upon the student; and no exception should ever be made. INTRODUCTION 33 Wash Bottles. Wash bottles, for distilled water, should con- sist of flasks of about 500 cc. capacity which are provided with rubber stoppers and with tubes gracefully bent and not too long. The jet should be connected in a flexible manner with the outlet tube by means of a short piece of soft rubber tubing, and should deliver a smooth stream about i mm. in diameter. For use with hot water, the neck of the bottle should be wrapped with heavy asbestos twine, or other suitable material. In order to avoid mistakes, wash bottles for other liquids than distilled water should always be plainly labelled. Gooch's Filtration Crucible. In 1878, F. A. Gooch proposed the separation of certain precipitates by nitration with suction through a mat of asbestos bedded on the perforated bottom of a crucible. The precipitate then could be washed, dried, and weighed in the crucible. Preparation of the Asbestos. There are several varieties of asbestos in the market, of which the long-fiber " silky " chrys- olite asbestos is the best. Rub the asbestos roughly over the surface of a jo-mesh brass sieve, placed in an inverted position on a sheet of paper, until a sufficient quantity has passed through. Shake this up with water, allow most of it to settle, and pour off the very fine particles. Now, digest the pulp on the steam bath for i hour with strong hydrochloric acid, in a covered porcelain dish. At the end of this operation, dilute the mixture with water, pour off the liquid through a funnel provided with a platinum filtration cone, and wash the asbestos with hot water, at first by decantation, and finally in the funnel, using gentle suction, until a 5 cc. portion of the filtrate fails to give an opales- cence with silver nitrate. Mix the washed asbestos with dis- tilled water and keep it in a bottle ready for use. Packing the Crucible with Asbestos Felt. Stretch a piece of rubber-band tubing over the upper edges of a cylindrical glass tube, about 3 cm. in diameter and 7 cm. long, which is closed at one end except for an attached stem of suitable size and length to pass through a rubber stopper. The tubing should project 34 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS 1.5-2 cm. above the top of the funnel tube. Fit the glass funnel tube into the stopper of a filter bottle, connect the latter with the vacuum pump, and then press into the short projecting end of the rubber tube the Gooch Crucible, so that it fits in an air- tight manner. Take some of the asbestos suspension referred to above, add water, and stir the mixture. Allow this to settle for some time, pour off the very fine particles, apply a gentle suction, and then pour some of the mixture cautiously into the crucible until an even felt of asbestos, not over 1.5 mm. in thickness, is formed. Place a small perforated disk (filter plate) upon the asbestos, and pour just enough more asbestos into the crucible to barely cover the disk. Run water through the crucible until no more asbestos fibers run through, and make sure that the washings are free from chlorides. If the water which has passed through the cruci- ble is held before a bright light, any suspended asbestos fibers can readily be seen. Usually, 250-500 cc. of water will suffice. The perforated filter plate is used to protect the asbestos felt, during the washing and subsequent filtration. Place the crucible in a small beaker, dry at 120-130 for an hour, cool in a desiccator, and weigh. Heat again, for ^ hour, cool, and again weigh, repeating until the weight is constant within 0.0002 g. The filter is then ready for use. How to Use the Gooch Crucible. The weighed crucible is re- placed in the funnel, and a gentle suction applied, after which the liquid to be filtered may be passed through the crucible, and the precipitate washed as if the crucible were a filter paper and funnel. When pouring liquid into the crucible, hold the stirring rod well down in the crucible, so as not to disturb the asbestos. Always examine the first portions of the filtrate with great care for asbestos fibers, and refilter the liquid if any are visible. When the precipitate has been washed and the crucible dried, the whole is weighed. The drying and weighing should be re- peated, as above, to constant weight. The increase in weight represents the weight of the precipitate. INTRODUCTION 35 The same crucible can be used for a number of determinations of the same substance. When the collection of precipitates in the crucible becomes too large, the upper part can be removed, and the crucible used as before. If. the felt has been properly prepared, filtration and washing are rapidly accomplished, and this, combined with the possibility of repeatedly using the same filter, is a strong argument in its favor, with any but gelatinous precipitates. If perforated platinum crucibles are available, which after removal from the funnel, can be fitted into platinum cups, the precipitates can be ignited as in ordinary crucibles. In this case, however, it is better to pack the Gooch crucible with a felt of platinum sponge. In this form, the apparatus is known as a Munroe Crucible. Sources of Error. It is important to remember that asbestos may absorb appreciable amounts of alkali, not removed by washing, so that, as a rule, solutions containing fixed alkalies should not be filtered through asbestos. Asbestos is also slightly attacked by water and feebly acid solutions ; but after the treat- ment indicated above, there is no real danger from this source. If the felt has been properly prepared, there is no danger of losing asbestos during the filtration and washing, but the liquid which runs through should nevertheless be carefully examined, and refiltered if necessary. The Theory of Washing Precipitates. The theory of wash- ing precipitates should include the consideration of several factors, among which may be mentioned the phenomena of adsorption (in which the filter also takes part), and the tend- ency of precipitates to enter the wash liquid in colloidal form. But these factors have been discussed under precipitation. Aside from these considerations, there is the important ques- tion concerning the most effective method of washing precipitates. Let us suppose that a precipitate is to be washed by decanta- tion, and for the sake of simplicity let us assume that neither it nor the filter exercises any physical or chemical action on the 36 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS salts dissolved in the mother liquid. The supernatant mother liquid has been decanted as far as possible into the filter, and the latter has been allowed to drain. Let v cc. be the total volume of solution which remains in contact with the precipitate and filter, and V cc. the volume of wash water added each tune ; and assume that the latter mixes uniformly with the liquid adhering to the precipitate and filter. Then, upon the addition of V cc. of water, the total volume of liquid is (V+v) cc. Further, let Co be the concentration in grams per cubic centimeter of the undesirable salts in the original solution; then the quantity contained in the v cc. left in contact with the precipitate and filter is vC grams. By the addition of V cc. of water, the concentration is reduced to Ci= -r-Co, an d if this liquid is removed until only v cc. are left, the quantity of undesirable salts present is reduced to flCo. A second addition of V cc. of water gives the ntrati salts in the v cc. left on draining is now concentration, C 2 = ~ Ci = f ~- J C , and the quantity of or, after n washings, the quantity of undesirable salts has been diminished to the value, This formula expresses mathematically the self-evident fact that, for a given number of washings, the quantity of undesir- able salts left behind will be the smaller, the more completely the precipitate and filter are drained, and the greater the volume of the wash water that is added each time. The formula enables us, however, to find the answer to a less simple question; viz., What is the most efficient method of washing a precipitate with INTRODUCTION 37 a given amount of wash liquid ? Suppose, for example, we wish to use only 150 cc. of wash liquid : is it better to wash six times with 25 cc. portions, or to wash 10 times with 15 cc. portions? Let us assume that C = o.i g. per 'cubic centimeter, and that v = 5 cc. ; then, in the two cases, the quantities of undesirable salts left behind will be ( A) 6 X 0.5 = 0.0000 107 g., and (A) 10 Xo.5 = 0.00000047 g., respectively. Disregarding adsorption, which greatly decreases the efficiency of washing, ten washings with 15 cc. portions are 23 times as efficient as six washings with 25 cc. portions. Both methods of procedure will require ap- proximately equal intervals of time, since, in either case, 153 cc. of liquid must run through the filter. // is much better to wash a precipitate many times with small portions of liquid, than a few times with larger portions. Each portion of wash liquid should be removed as far as possible by decantation and drainage, before the addition of a fresh portion. Another factor to be considered is the temperature of the solution to be filtered. Since the rate of flow of a liquid through a filter depends largely upon the viscosity of the liquid, and since the viscosity of water at 100 is only one sixth that of water at o, it is well to filter and wash at a high temperature, unless there is good reason to the contrary. Finally, in washing a precipitate on a paper filter, great care must be taken to wash the filter itself. Soluble salts are tena- ciously held back at the upper edges of the paper, and therefore this part of the filter should receive especial attention. It is best to fill the filter each time, and, before refilling, to allow it to drain completely. This is a strong argument in favor of filters which are not too large. IV. THE DRYING AND IGNITION OF PRECIPITATES Drying Ovens. There are on the market many types of drying ovens, heated by gas, by steam pipes, or by elec- tricity, in which the temperature may be more or less accu- 38 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS rately controlled. The oven consists essentially of a drying chamber, through which there is provided a slow circulation of hot air. A precipitate is dried on the filter by placing the funnel con- taining both in a drying oven, at 90-100, and leaving it there for a sufficient time. The funnel should be covered with a sheet of common filter paper, fastened in place by crimping its edges over those of the funnel. If the precipitate is suitable for weighing without ignition, e.g. silver chloride in a Gooch crucible, it should be dried to constant weight at a temperature considerably above the boiling point of water (in this case, at 120-130), in order to remove the last traces of moisture from the filter as well as from the precipitate. (Before it is used, the packed Gooch crucible should, of course, be dried to constant weight at the same temperature.) Many precipitates may, under proper precautions, be ignited without previous drying in an oven. But if such precipitates can be dried over night, or otherwise, without loss of time to the analyst, it is well to submit them to this process. The precipitate, with the filter folded over it, is placed at the bottom of the crucible, and the latter is supported, on a tri- angle, so far above the small flame of a burner as to preclude the violent escape of steam. When the filter and contents are dry, the open crucible is tilted on its side, and the heat slightly increased until the filter chars; the heating is then continued at this rate until the gases from the dry distillation of the paper have been completely expelled without taking fire. In this way, no material will be lost owing to strong draughts caused within the crucible by burning gases. During the dry distillation of the paper, the flame should be placed near the mouth of the crucible ; but afterwards it should be well at the base of the inclined crucible, to allow a ready access of air. After the filter has been freed from volatile matter, the crucible should be heated to redness until the ignition is com- plete. UNIVERSITY OF CALIFORNIA DEPARTMENT OF CIVIL ENGINEER1N Some precipitates are reduced or otherwise affected by con- tact with hot carbon or reducing gases from the filter paper; e.g. silver chloride, lead sulphate, etc. are reduced to metal. Since, however, these metals are volatile only at very high tem- peratures, there is no loss in their case, and the metal can by suitable treatment be transformed quantitatively into the original compound. In such cases it is advisable to separate the precipitate as far as possible from the filter, and then to ignite the latter. The small quantity of reduced metal is moistened with a few drops of nitric acid, and the resulting nitrate con- verted into silver chloride with hydrochloric acid, or into lead sulphate with sulphuric acid, and the excess of acid expelled by cautiously heating the crucible. The bulk of the precipitate is then added, and the whole ignited. Unless specially directed, precipitates should not be heated over the blast lamp. Desiccators. After an object has been dried and ignited, it must be permitted to cool before it can be weighed with accu- racy. In order to protect it from contamination with moisture, carbon dioxide, etc., it should invariably be allowed to cool in a desiccator. For general analytical work, desiccators should be charged with fragments of fused, anhydrous calcium chloride, some distance above which is placed a porcelain plate provided with holes of a size suitable for the reception of crucibles. In order to give the cover of the desiccator an air-tight fit, the ground- glass contact surfaces should be thinly coated with vaseline or some similar substance. 1 Desiccators should never be left uncovered. The dehydrat- ing agent is intended to keep the air inside the desiccator dry ; it rapidly loses its efficiency if exposed to the outside air. If the lumps of calcium chloride tend to stick together, the charge should be renewed. 1 A mixture made by melting together equal parts of vaseline and beeswax is very suitable. 40 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Pumice moistened with concentrated sulphuric acid is some- times used instead of calcium chloride. Crucibles. The most commonly used crucibles are of high grade porcelain. They withstand very high temperatures without appreciable change of weight, and they are compara- tively cheap in price. They cannot be used for fusions because most fluxes, particularly those of a basic nature, attack the glaze as well as the porcelain itself. Even in the ignition of non-basic precipitates, in spite of the most careful washing, traces of fusible materials always remain with the precipitate, and these in time destroy the glaze. After the lining is thus roughened it is difficult to clean, and the crucible is unsuitable for use. Crucibles more or less suitable for the ignition of precipitates are also made of alundum, and of fused silica. Crucibles of platinum are very desirable for ignitions; and for many fusions they are essential. Platinum melts at about 1770 and does not soften enough to preclude its use at tem- peratures slightly below its melting point. It is soluble in liquids containing free chlorine, such as nitrate-chloride mixtures of acid reaction, and to a lesser degree in acid ferric chloride solu- tions. Care must be taken to prevent injury to platinum vessels, or the introduction of platinum into solutions, by a disregard of these facts. Platinum easily alloys with most metals, and for that reason it should not ordinarily be heated in contact with metals, or with compounds of easily reducible metals, never, if carbon or reducing gases are also present. When heated for a long time in contact with carbon, platinum slowly takes up the latter and becomes brittle; therefore, the crucible should never be heated in a reducing flame. The flame should be carefully adjusted so that the tip of the inner cone is below the bottom of the crucible, and a flame showing yellow must never be used. Compounds of phosphorus or arsenic must not be heated under reducing conditions, since the free elements, as well as phosphides and INTRODUCTION 41 arsenides, render platinum brittle and lower its melting point. " Unknown " substances should never be heated in platinum vessels. Platinum ware should always be kept bright and clean. For this purpose it should be frequently polished with fine sea sand or with precipitated silica. These remove most impurities, and polish the platinum without serious loss. The fusion of potas- sium bisulphate in the vessel is a good method for cleaning the badly tarnished inside. The bisulphate should be poured out of the crucible while still liquid ; for if it has been strongly heated, the melt (pyrosulphate) is apt to expand so rapidly on cooling as to burst the crucible. Never heat the platinum crucible or dish in contact with iron or metals other than platinum, nor place hot platinum in con- tact with foreign metals. Use nothing but pipeclay, quartz, or platinum triangles, and platinum-shod tongs. 1 V. THE EVAPORATION OF LIQUIDS The greatest care must be taken to prevent loss of material during processes of solution and evaporation, either from the evolution of gas, from too violent ebullition, or from evaporation 1 Modern platinum ware is often inferior in quality to that on the market some years ago, and the cause has been the subject of special inquiry by a committee of the American Chemical Society. The main objections are: "(*) Undue loss of weight on ignition ; (2) undue loss on acid treatment, especially after strong igni- tion ; (3) unsightly appearance of the surface after strong ignition, especially after the initial stages of heating; (4) adhesion of crucibles and dishes to triangles, sometimes to such an extent as to leave indentations on the vessel at the points of contact with the triangle, even when complete cooling has been reached before the two are separated ; (5) alkalinity of the surface of the ware after strong ignition ; (6) blistering; and (7) development of cracks after continued heating." It is the general opinion that the trouble arises from the working of scrap platinum into chemical ware. The main difficulties here mentioned are not characteristic of platinum ware from some of the best manufacturers. The committee recommends that purchasers specify that platinum ware must show no marked uneven discoloration on heating, must give no test for iron after heating for two hours, and that the rate of loss per hour at 1100 over a period of four hours shall not exceed 0.2 mg. 42 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS nearly to dryness accompanied by spattering or by the " crawl- ing " of salts over the edge of the vessel. In order to prevent mechanical losses, solutions in which gases are being evolved, or which are to be boiled on the hot plate, should invariably be covered. And liquids which contain suspended matter (precipitates) should always be cautiously heated, since the presence of solid matter frequently occasions violent " bumping " which may lead to mechanical losses or to the destruction of the vessel. The evaporation of aqueous solutions rarely requires the use of temperatures above 100. Temperatures somewhat below this point, but sufficient for the evaporation of most aqueous solutions, are best attained by the use of the steam bath, which has the advantage of keeping the solutions at a temperature below that at which mechanical losses are to be feared. Evaporations should not be attempted in tall, narrow vessels, but should be carried out in low, wide-mouthed dishes or casse- roles : it is obvious that evaporation is promoted by the ex- posure of a large surface of liquid to the air. In evaporations on the steam bath, a watch glass should be supported above the casserole or dish by means of a glass triangle or other suitable device. If a large volume of liquid is to be evaporated, it is not neces- sary that the vessel should contain it all at once. Fresh portions of the liquid may be added, from time to tune, as the volume of that in the vessel is reduced by evaporation. Liquids should never be transferred from one vessel to another, nor to a filter, without the aid of a stirring rod held firmly against the lip or side of the containing vessel. In order to prevent the loss of liquid by running down on the outside of the vessel, a very thin coating of vaseline, applied with the finger to the out- side edge of the vessel, will suffice. If the vessel is provided with a lip, this is usually unnecessary. As few transfers of liquid as possible from one vessel to another should be made during an analysis. In such transfers, the solu- INTRODUCTION 43 tion must, of course, be quantitatively washed out. This can be accomplished better by the use of successive small portions of wash water, say of 5-10 cc. each, than by the addition of a few larger portions which unnecessarily increase the volume of the solution and lead to loss of time in subsequent nitrations or evaporations. VI. THE VOLUMETRIC MEASUREMENT OF LIQUIDS 1 Measurements with a good balance and weights can often be made with a precision even greater than is necessary for general analytical work. But, as has been intimated, the errors involved in the preparation of a troublesome precipitate for weighing may impair the value of an exact weighing. Although the measurement of volume, in volumetric analysis, is not apt to be so precise and reliable as the measurement of weight, 2 yet the results of volumetric processes, based on suitable reactions, are frequently more trustworthy than those of gravimetric processes, because the volumetric process for the determination of the substance is less liable to error. With proper precautions many volumetric processes yield excellent results ; and, es- pecially in technical work, where time is an essential factor, volumetric processes are very often used in preference to gravi- metric. In order, however, that dangerous errors may be- eliminated in volumetric work, it is of great importance that the analyst should have a clear idea of the precautions necessary for the attainment of a high degree of accuracy. 1 For more detailed information on this subject, see Bulletin of the Bureau of Standards, Vol. 4, pp. 553-601 (1908). 2 Even this difficulty can be obviated by the use of weight burettes; i.e. of burettes of such construction as to be readily weighable both before and after the removal of the quantity of solution required for the completion of the given reaction. The difference gives the weight of solution required, and, provided the solution has been standardized by the same method, the quantity of the substance under in- vestigation can be readily calculated. In this way, if based on suitable reactions, exceedingly exact determinations can be executed. (A similar weight burette should always be used as a counterpoise.) 44 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Volumetric Apparatus. The exact measurement of liquid volumes necessitates the use of certain special forms of apparatus, the most important of which will now be described. Burettes are graduated glass tubes of uniform, small diam- eter, designed to measure variable amounts of liquids delivered by them when supported in a vertical position. The outflow of the liquid is controlled either by a glass stopcock or by means of a rubber joint which connects the end of the tube with a glass nozzle, and which is provided with a pinchcock or other suitable device. The former require the use of some lubricant, such as vaseline, to permit easy control of the outflow, and the latter have the disadvantage that the rubber connection is acted upon to some extent by certain solutions, which in consequence are apt to experience a change in concentration; e.g. rubber stop- cocks should never be used with permanganate or iodine solu- tions. For accurate work, 50 cc. burettes should be graduated to o.i cc. and 25 cc. and 30 cc. burettes to 0.05 cc., and the graduation marks should be separated by at least i mm. Transfer Pipettes are tubes of much smaller bore than bu- rettes, designed to deliver specific volumes of liquid. They are provided with an enlargement at the center, which greatly reduces the length required, and with a single mark on the upper length, which indicates the point to which they must be filled in order to deliver the indicated volume of liquid. Pipettes are filled by suction and are allowed to deliver the liquid, while held in a vertical position, by the action of gravity. On account of the smallness of the bore at the point where the reading is made, the pipette may be made to measure liquids with a high degree of accuracy. However, certain errors of manipulation often render the measurements inexact. Measuring Flasks are employed for measuring relatively large volumes of liquid in one portion. The neck must be sufficiently small to permit a reading with only a slight per- centage error, but large enough to permit filling and emptying without trouble. The neck should also be of uniform bore and INTRODUCTION 45 of some margin above and below the mark. For the most accurate work the flask is always graduated for containing the amount indicated by the inscription upon it. Graduated Cylinders are cylindrical glass vessels provided with an enlarged base or foot and with a lip for pouring. They are marked to indicate the varying amounts of liquid which they may contain and are employed for rough measurements only. Necessary Precautions in the Use of Volumetric Apparatus. In making use of volumetric methods, the following are the most important sources of error; they must be fully reckoned with if the results are to be reliable. Errors Due to Water in the Apparatus. It is usually neces- sary to rinse the apparatus with water before using ; and the amount of water retained may appreciably change the concen- tration of the solution measured. The error can be avoided by drying the apparatus before use, or, more conveniently, by rinsing it out with several small portions of the solution to be measured, and discarding the washings. Errors Due to Drainage or Afterflow. When a liquid is per- mitted to flow somewhat rapidly from a burette, or pipette, small amounts adhere to its inner surface, and gradually flow down and unite with the liquid still remaining hi the vessel. In order to avoid errors from this source the rate of outflow must be limited by the size of the outlet, and a sufficient interval must elapse between the time at which the flow from the ap- paratus is stopped, and at which the reading is made. This inter- val, unless indicated on the instrument, may be taken as 30 seconds in the case of burettes and 15 seconds in the case of pipettes. In the case of transfer pipettes, the outlets must be of such size that the free outflow shall last not more than one minute and not less than 15, 20, and 30 seconds respectively, for 5, 10, and 50 cc. pipettes. The rate of outflow of burettes must not be more than three minutes, nor less than 90 and 50 seconds, respectively, for 50 and 30 cc. burettes. Burette and pipette tips should be made with a gradual taper of 2-3 cm. ; a sudden 46 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS contraction at the orifice is not permitted, and the tip should be well finished. In filling pipettes and burettes excess liquid adhering to the tip should be removed when completing the filling. In emptying pipettes and burettes, they should be held in a vertical position, and after the continuous unrestricted outflow has ceased the tip should be touched with the wet surface of the receiving vessel, to complete the emptying. Apparatus must be sufficiently clean to permit uniform wetting of the surface. (For the prepara- tion of cleaning solution, see the Appendix.) Errors Due to Parallax. In all apparatus in which the volume is. limited by a meniscus, the reading or setting is made, when possible, on the lowest point of the meniscus. Since this point lies at the center of the tube, the error from parallax must be avoided. The method of doing this is to support the tube so that its main axis is vertical, and to hold the eye at such a level that the line of sight makes an angle of 90 with the axis. In order that the lowest point of the meniscus may be ob- served, it is well to place a screen of some dark material im- mediately below the meniscus, which renders the profile of the meniscus dark and clearly visible against a light background. A convenient device for this purpose is a collar-shaped section of black rubber tubing, cut open at one side and of such a size as to clasp the tube firmly. Errors Due to Variations in Temperature. The volume occu- pied by a given weight of water, as well as the capacity of the measuring vessel, is dependent upon the temperature ; and the error involved in the measurement of the volume of a given mass of water, at any other temperature than the standard one, is due to the joint effect of the changed capacity of the vessel and the changed volume of the liquid. The coefficient of cubical expension of ordinary glass may be taken to be 0.000025 ; but the volume change of the water is much greater than that of the glass measuring vessel, and also much less uniform from degree to degree. The factors by INTRODUCTION 47 which a volume of water, measured at temperatures ranging from 10-29 m a vessel calibrated for 20 must be multiplied in order to obtain the true volume occupied by the liquid at 20, are given in the following table : TEMPERATURE or THE WATER """\^UNITS TENS^^-^ * 2 3 [4 1 i .001 24 I.OOII7 1 .00109 I.OOIOO 1.00089 2 I.OOOOO 0.99981 0.99961 0.99941 0.99919 TEMPERATURE OF THE WATER "~^\UNITS TENS^^\_ 5 6 7 8 9 1 1.00077 1.00064 I.OOO49 1.00034 I.OOOlS 2 0.99896 0.99873 0.99848 0.99822 0.99798 If the prevailing temperature does not differ by more than, say, 3 from the standard, this correction may ordinarily be omitted. In the case of solutions of 0.2 N concentration, or less, the corrections differ so little from those for pure water that the factors given in the table may be used without appreciable error. In order to illustrate the use of such factors, let us suppose that, in the standardization of a solution, a burette graduated correctly for 20 is used at an actual temperature of 27, and that the indicated volume of solution withdrawn is 28.75 cc - Then the true volume at 20 of this quantity of liquid is 28.75 X 0.99848 = 28.70 cc. And, if a determination is later made with this solution at, say, 17, and the indicated volume used is 28.68 cc., then the true volume at 20 is 28.68 X 1.00049 = 28.70 cc. That is to say, the same quantity of reagent is contained in an apparent volume of 28.75 cc - at 2 7> or in an apparent volume of 28.68 cc. at 17, as is contained in an actual volume of 28.70 cc. of the solution at 20 . 1 1 At 27 the actual volume of the liquid measured is a shade greater than 28.75 cc - and at 17 it is a shade less than 28.68 cc. The measuring vessel is larger at 27, and smaller at 17, than it is at 20. 48 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Errors due to Different Units of Volume Employed. Unfor- tunately, a number of different " liters " have been suggested for use in volumetric analysis. The normal liter, that is, the volume occupied by a kilogram of water, weighed in a vacuum and measured at 4, would manifestly be out of the question if it had to be determined in that way. The so-called " Mohr liter " is the volume occupied by a kilogram of water when weighed in the air with brass weights at a temperature of 17.5 ; but this volume varies with the atmospheric conditions. Other " liters " involving measurements at 15, 15.5, or 20 have been suggested by various chemists. It matters little, in analytical work, which liter is adopted, but it is of the greatest importance to have the pipettes, burettes, and measuring flasks rigorously consistent with one another. This matter is serious enough to require especial emphasis, since apparatus, if not specifically ordered, may be supplied by dealers, at different times, graduated according to different systems ; and mixed graduations may thus come into the hands of an individual analyst. As an example of the magnitude of the errors which might thus be introduced, it should be noted that the normal liter is related to the Mohr liter as 1000 : 1002.3. Much of the graduated apparatus on the market bears no mark by means of which the unit of volume represented can be recognized, and even when this is clearly designated the per- centage error represented may be large. It is not advisable, therefore, to use any piece of graduated apparatus, unless its actual value is well known. Owing to the great difficulty in measuring directly the re- lation between cubic capacity and the unit of length, the Inter- national Committee of Weights and Measures defines the liter as " the volume occupied by the mass of one kilogram of pure water at its maximum density under normal atmospheric pres- sure." This is almost exactly 1000 cc. 1 and for all practical pur- poses may be regarded as such. 1 About 1000.029 cc. INTRODUCTION 49 It is now customary to use this true liter as the standard, but of course it is out of the question to weigh a kilogram of water at 4 in a vacuum ; some convenient temperature preferably the average working temperature of the laboratory must be selected, and the necessary corrections made. If a liter flask is marked correctly at 20, this means that at 20 it will contain a mass of water (998.234 g.) which occupies a volume equal to that occupied by 1000 grams of pure water at 4. This quantity of water, if weighed with brass weights in air of mean humidity, at 20 and 760 mm., has an apparent weight of 997.18 grams. The Calibration of Volumetric Apparatus. The weight of brass (brass weights) which will be required to counterbalance one liter of pure water must be calculated from the temperature of the water and the density of the air. The following table indicates, for temperatures of the water (and room) ranging from 15-29, how many milligrams less than 1000 grams a quantity of water will weigh which is sufficient to fill to the mark a i -liter flask correctly calibrated for 20, the weighing being carried out in air of 50 per cent humidity at 760 mm. pressure (unreduced). 1 >N X N UNITS TENS\^ 1 2 3 4 5 6 7 8 9 1 ' 1950 2IOO 2260 2440 2630 2 2820 3030 3240 3470 3710 3960 42IO 4480 4760 2620 If such a flask is filled to the mark with water of 22.4, for example, the water will under the conditions of the table require a counterpoise of 1000-3.332 = 996.668 grams. The determination of the capacity of a measuring flask is carried out by weighing the water contained in it, while the volume of water delivered by a burette or pipette is determined by weighing this water after its delivery into another vessel. 1 These values depend upon the specific gravities of the water and the (brass) weights, the density of the air, and the coefficient of cubical expansion of the glass vessel. 50 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS The temperature of the water should be taken before and after the experiment ; it is important that it should be the same as the room temperature at the time of the weighing. In the calibration of a flask, the dry flask is placed upon the right-hand pan of the balance, together with the nominal weight of its capacity, i.e. with as many grams as it is supposed to con- tain cubic centimeters, and then tare material is placed upon the left-hand pan until the balance is brought into equilibrium. The weights are then removed from the right-hand pan, the flask is filled to the mark with water, and weights are added until the balance is again in equilibrium. The nominal capacity weight, minus the additional weight which is required upon the right-hand pan in order to reestablish equilibrium, is equal to the weight of the water in the flask. In the case of burettes and pipettes, a covered beaker is placed upon the right-hand pan, together with the nominal weight in grams of the volume to be delivered, after which the balance is brought into equilibrium by the addition of tare material to the left-hand pan. The water is then allowed to run into the beaker, which is replaced upon the right-hand pan. The subse- quent procedure is the same as that described above. The difference between the additional weight required to reestablish equilibrium and that calculated from the above table indicates directly the error of the vessel. If, for example, it be found necessary, in order to reestablish equilibrium in the case of a 5oo-cc. flask filled to the mark with water of 22.4, to add 1.832 g., instead of the calculated 1.666 g., then it follows that the vessel is 0.166 cc. too small. On the other hand, in test- ing the 25 cc. segment of a 30-cc. burette, at a temperature of 17, the additional weight required on the right-hand pan should be (25 X 2 260) -f- 1000 = 57 mg. ; if, instead of this, it be found that an additional weight of 15 mg. is required on the left-hand pan, then the 25 cc. segment is 57 mg. ( 15 mg.) =0.072 cc. too large. 1 1 This error is too large to be tolerated. INTRODUCTION 51 In the calibration and use of burettes, the liquid should in general be allowed to flow from the zero mark to some second level in the burette. D. THE PREPARATION OF SAMPLES FOR ANALYSIS It is not easy to give general rules for the preparation of sub- stances for analysis, because it is necessary to proceed differ- ently in different cases. In all cases, however, the samples should promptly be transferred to tightly stoppered bottles or weighing tubes. In technical analyses, for the purpose of determining the commercial value of an article, or of controlling processes of manufacture, materials must be analyzed as they are. But, in every case, especial care should be taken to make up a sample which will represent as nearly as possible the average composition of the whole lot. If, on the other hand, it is desired to determine the atomic composition of a compound, it is necessary to select or prepare pure material for analysis. This may seem simpler than it really is. Many compounds absorb or give up moisture upon exposure to the air, and their treatment should vary with their nature, as illustrated in the following cases. Salts such has Na 2 SO 4 . 10 H 2 O and Na 2 CO 3 . 10 H 2 O, which effloresce in ordinary air, may be dried, after recrystallization, by strongly pressing the powdered material between several layers of filter paper, the paper being renewed as long as moisture continues to be taken up ; MgSO 4 . 7 H 2 O and NaKC 4 H 4 O 6 . 4 H 2 0, which do not lose water of constitution in ordinary air, may be spread out upon filter paper, covered with another sheet, and allowed to dry at the ordinary temperature. Compounds such as HFe(SO 4 ) 2 . 4 H 2 O and CaC 4 H 4 O 6 . H 2 O, which do not effloresce in artificially dried air, but which undergo chemical change at 100, may be conveniently dried in a desiccator, over calcium chloride. Substances, as KHC 4 H 4 O 6 , sugar, etc., which readily give up hygroscopic moisture at 100, without other 52 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS alteration, are best dried in an oven at that temperature ; while K 2 PtCl 6 , which retains moisture, or dries only slowly at 100, but which decomposes below a red heat, should be dried in an oven at, say, 130. Finally, substances such as NaCl, Na 2 S0 4 , etc. may be given a preliminary drying, in a covered vessel, at 130, or higher, to prevent decrepitation, and then be ignited, more or less strongly, depending upon their nature. In every case, the sample should be dried, without decomposition, to constant weight. Substances used in testing the accuracy of analytical pro- cesses, or in standardizing volumetric solutions, must also be extremely pure. In fact, compounds are generally favored which are non-hygroscopic, and which may readily be prepared in a pure condition; if possible, it is well to select compounds which normally do not contain water of crystallization. Many salts can be obtained sufficiently pure in the market ; but their purity should never be accepted on faith. If tests indicate the presence of impurity, and, often, if the salt contains water of crystallization, the material should be recrystallized. For this purpose, a convenient weight of the salt is dissolved in the least possible quantity of hot water, using a quantity of water not quite sufficient to dissolve the whole lot; the hot solution is poured into a fluted filter, held in a stemless funnel, and the filtrate is received with continuous stirring in a beaker, which itself is immersed in cold water, in a larger vessel. The rapid cooling and constant stirring cause the formation of a fine crystalline powder, which is almost free from inclosed mother-liquor. The crystalline powder is filtered off in a funnel containing a perforated platinum cone, the adhering mother- liquor being removed by suction or in a centrifuge. Two such recrystallizations will nearly always suffice. According to the nature of the substance, it is dried in the air at a specific tem- perature, or in a desiccator, to constant weight. Concerning the preparation of samples for analysis by beginners in quantitative analysis, the reader should consult the appendix. PART II GRAVIMETRIC ANALYSIS EXERCISES WITH THE BALANCE Before beginning work at the balance, read carefully the rules given on pp. 9-11 of Part I, and observe them always. Determination of the Zero-Point. Determine the zero-point of the unloaded balance, according to the method on p. n. If the zero-point found is not more than one division from the center of the scale, the balance may be used by the student; otherwise it will be adjusted by an instructor, upon request. The beginner should not attempt this adjustment. Determination of the Weight of an Object. Clean two porce- lain crucibles, rinse them with distilled water, and allow them to drain. Place each crucible upon a pipes tern triangle, sup- ported upon a tripod, and heat with the colorless flame of a Bunsen or Tyrill burner, gently at first, and then to a red heat. Allow the crucibles to cool off somewhat, but while still warm, place them in a desiccator, using the crucible tongs. (After any piece of apparatus has been cleaned and ignited for weighing, it must never be handled with the fingers before the weight is taken.} Allow the crucibles to cool in the desiccator for at least 20 minutes. Now with the crucible tongs place a crucible on the left-hand pan of the balance, and, by means of the forceps in the weight box, place weights on the right-hand pan until they balance the crucible to within 0.005 g- Begin with a weight which you think will approximately balance the object, lower the balance beam, and gently release the pan supports. It will then be seen 53 54 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS which side is the heavier. Finally adjust the rider, so that, when the beam is swinging freely, the pointer traverses the same number of divisions on either side of the observed zero-point. Always try the weights in the order in which they occur in the box, beginning with the heavier weights, and using the rider for weights smaller than 5 or 10 milligrams, according to the num- ber of large divisions on the beam. As soon as the object appears to be balanced, raise and lower the beam, and make another observation. Read the weight of the crucible by noting in order the vacant spaces in the box, begin- ning with the largest missing weight; and check this reading as the weights are returned to the box . Be sure also to note the weight recorded by the rider, and then lift it from the beam. Always record the weight, in pencil as it is first read, and in ink after it has been checked, and always in the record book, never on a loose sheet of paper. In this manner, weigh the two crucibles separately, and then weigh them together, entering all three results in the notebook. (In connection with the keeping of records, see the remarks on p. 4.) The sum of the separate weights should agree closely with the result obtained upon weighing both crucibles together, within, say, 0.0002 g. THE DETERMINATION OF CHLORINE IN A SOLUBLE CHLORIDE The sample may be pure sodium chloride, or it may be an artificially prepared mixture of sodium chloride and sodium carbonate. Method. The aqueous solution of the chloride is acidified with nitric acid and treated with silver nitrate in excess. The chlorine is quantitatively precipitated as silver chloride, which is filtered off, washed, dried, and weighed. Other acids which yield silver salts insoluble in nitric acid must of course be absent. A. A Paper Filter is Used: Procedure. Carefully clean the weighing tube containing the sample, without handling it di- GRAVIMETRIC ANALYSIS 55 rectly with the fingers, and weigh it accurately to a tenth of a milligram. Record the weight at once in the notebook. Hold the tube over a clean 300-0:. beaker (plainly labelled " I "), remove the stopper, allowing no particles to fall from it or from the tube elsewhere than into the beaker, and carefully pour into the beaker from 0.2 to 0.3 g. of the sample. Replace the stopper in the tube, weigh accurately, and record the weight in the notebook. The difference of these two weights is the weight of the portion taken for analysis. Weigh a second portion of 0.2-0.3 g. into another beaker (labelled " II "), entering the weights and their difference in the notebook, as before. Dissolve each portion in about 150 cc. of distilled water, and acidify the solutions with nitric acid, adding the acid slowly and with stirring, until a strip of blue litmus paper shows an acid reaction when moistened by means of the wet stirring rod with the least possible quantity of the liquid. Assuming the sample to be pure sodium chloride, calculate the volume of silver nitrate solution required in each case to effect complete precipitation (for the strength of the reagents, see the Appendix), and add slowly and with stirring about 5 cc. more than that amount. Cover the beaker with a watch glass, and heat the solution gradually to boiling, with occasional stirring. Con- tinue the heating and stirring until the precipitate coagulates and the supernatant liquid is clear. The beaker should be kept away from direct sunlight, and the heating and stirring should be so conducted as to avoid any possibility of loss. Finally, add to the clear liquid a drop or two of silver nitrate solution, to test for complete precipitation; if a precipitate forms, add 5 cc. more, and test again. Prepare two ashless filters (9 cm. in diameter), according to the directions given on p. 30, and decant the hot liquid through the filter in each case, leaving the precipitate as far as possible in the beaker. Unless the filtrates are perfectly clear and free from particles of silver chloride, they must be refiltered through the same filters. If they are perfectly clear, and the tests show 56 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS complete precipitation, pour them into one of the laboratory receptacles for " Silver Residues," wash the beakers with tap water and then with distilled water, and replace them under the funnels. Now wash the precipitates twice by decantation with 10 cc. portions of hot water, acidified with a drop or two of nitric acid, pouring the washings through the filters, and finally transfer each precipitate to the corresponding filter by means of a stream of hot water from the wash bottle, loosening the adher- ing particles with the aid of a " policeman" (see p. 32). Wash the filters and precipitates with hot water until 3 cc. of the wash- ings show no cloudiness or opalescence with one drop of dilute hy- drochloric acid. After allowing the filters to drain, cover each funnel with an ordinary filter paper, crimping the edges of the paper over the sides of the funnel. The funnels, properly numbered, and labeled with the student's name and desk num- ber, should then be placed in a drying oven, at a temperature of 9o-ioo, and left there until completely dry. Now, in the case of each precipitate, open the filter over a piece of smooth glazed paper, about six inches in diameter, and, by finally rubbing the sides of the filter gently together, transfer the precipitate as completely as possible to the center of the glazed paper. Be careful not to rub off any appreciable quantity of the paper, nor to lose any of the silver chloride in the form of dust. Cover the precipitate on the paper with an inverted fun- nel or watch glass, to protect it from dust and air currents. Carefully refold the paper, flat, bend the top over, and roll the paper into a small bundle ; then place it in a weighed porcelain crucible. Place the crucible upon a triangle, incline it about 45, and ignite gently until the volatile products are expelled from the paper. Then, with the flame well at the base of the inclined crucible, ignite strongly until all the carbon is consumed. (See Part I, p. 38.) Allow the crucible to cool, add two drops of 6-normal nitric acid and one of hydrochloric acid, and heat with the greatest caution, to avoid spattering, until the acids are ex- pelled. Transfer the bulk of the precipitate quantitatively from GRAVIMETRIC ANALYSIS 57 the glazed paper to the cooled crucible, placing the latter on a second piece of glazed paper and brushing the precipitate into it, with a small camel's hair brush. Moisten the precipitate with two drops of nitric, and one drop of hydrochloric acid, carefully expel the acids, and then gradually raise the temperature until the salt just begins to fuse. Allow the crucible to cool in a desiccator, and weigh it. Repeat the heating, without the addition of acids, and weigh the cooled crucible. The heating and weighing must be repeated until the weight is constant within 0.2 mg. after two consecutive heatings. From the weight of silver chloride obtained in each case, calculate the percentage of chlorine in the sample. Finally, place the ignited precipitates in one of the laboratory receptacles for silver residues. The chloride which adheres to the crucible may be loosened by covering it with dilute sulphuric acid and adding a small quantity of granulated zinc. NOTES. i. The solution is acidified with nitric acid, before precipita- tion with silver nitrate, to prevent the precipitation of substances such as silver oxide, carbonate, phosphate, etc., which are insoluble in water but soluble in nitric acid. The acid also helps to coagulate the precipitate. A large excess of the acid is to be avoided, since it would slightly increase the solubility of the precipitate. 2. It is safer not to boil the acidified solution until after silver nitrate has been added in excess, since otherwise a slight amount of chlorine might be set free by the nitric acid and lost by volatilization. The presence of an excess of silver nitrate can easily be recognized at the time of its addition, by the increased readiness with which the precipitate coagulates and settles. 3. The precipitate should not be exposed to strong sunlight, since by its action a slight amount of chlorine is set free. The superficial alteration which the chloride undergoes in diffused daylight may readily be counter- acted by the treatment with nitric and hydrochloric acids ; but the loss in weight due to this cause is really too insignificant to affect the accuracy of the determination. 4. The precipitate and filter are washed with water to remove the non- volatile nitrates of silver and sodium, as well as any other soluble impurities. It may be assumed that these are all removed as soon as 3 cc. of the wash- ings give no cloudiness with a drop of hydrochloric acid. Only a single drop should be added, because silver chloride is somewhat soluble in this 58 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS acid. The wash water should be hot in order to prevent the precipitate from going into colloidal solution ; it is still safer to acidify the water slightly with nitric acid. 5. The bulk of the precipitate must be separated from the filter, be- cause the burning organic matter would reduce a considerable quantity of the precipitate to metallic silver, and its complete reconversion to the chloride within the crucible, by means of acids, would be uncertain. The small quantity which adheres to the filter, and which is more or less com- pletely reduced during ignition, is easily reconverted to chloride by the treatment with nitric and hydrochloric acids. 6. Silver chloride should not be heated to complete fusion, since a slight loss by volatilization might take place. The temperature of fusion is sufficient to completely remove adsorbed water and acids, but it is not always sufficient to destroy filter shreds. Although these would probably be completely oxidized by the nitro-hydrochloric acid and subsequent ignition, they should not be allowed to contaminate the precipitate. 7. The ignited precipitate of silver chloride, as well as the filtrates which contain an excess of silver nitrate, should be placed in the receptacles for silver residues ; the silver can easily be recovered. Assuming that, on the average, duplicate determinations require 55 cc. of o.2-normal silver nitrate solution, the residues returned in the case of each student should contain about 1.2 g. of metallic silver. Taking into account the number of analyses which have to be repeated, a class of one hundred students will usually return in residues at least 150 g. (about 5 oz.) of metallic silver. 8. Silver chloride is almost insoluble in water. The solubility varies somewhat with its physical condition, and is about 1.12 mg. per liter at 20 C. Owing to the common-ion effect, the solubility is still less in a very dilute solution of silver nitrate or of hydrochloric acid. In hot water the salt is more soluble, 21.8 mg. per liter at 100 C. ; but, fortunately, the speed of solution is so slow that the precipitate may be thoroughly washed with hot water, without undue error from this cause. Silver ion has a great tendency to enter into the formation of complex ions, as [Ag(NH 3 ) 2 ] + , [Ag(CN) 2 ]-, [AgS 2 O 3 ]-, etc., and the chloride is therefore readily soluble in aqueous ammonia, in alkali cyanide solutions, and in sodium thiosulphate ("hypo") solution. Silver chloride is also soluble in strong hydrochloric acid and in other chloride solutions, probably with the formation of a complex anion, such as [AgQ 2 ]~; it is also soluble in con- centrated silver nitrate solution, and in strong nitric acid. When boiled with concentrated sulphuric acid, it is converted into silver sulphate, and by zinc and dilute sulphuric acid it is reduced to metallic silver. GRAVIMETRIC ANALYSIS 59 B. A Gooch Crucible is Used: Procedure. Weigh out two samples of the substance, of about 0.25 g. each, and convert the chloride into silver chloride as described in procedure A . Mean- while, prepare two Gooch crucibles, following the directions on pp. 33-35 ; and finish the analysis according to the details there given. NOTE. Bromides, iodides, cyanides, sulphocyanates, etc., as well as silver itself, may be determined in a similar manner, with the use of Gooch crucibles. Chlorates, etc., may be determined by first reducing them to chlorides, and then precipitating with silver nitrate. For the determination of these substances when two or more of them are present in the same sample, the student is referred to the larger works on quantitative analysis. (But see also Part IV, Problems 37, 38, 90, 91, and 93.) THE DETERMINATION OF IRON AND OF SULPHUR IN A SOLUBLE SULPHATE OF IRON The sample may be pure ferrous ammonium sulphate, pure ferric alum, or an artificially prepared mixture of anhydrous ferric sulphate, sodium carbonate, and potassium sulphate. This mixture is readily soluble in dilute hydrochloric acid. Method. The sample is dissolved in water, with the addition of hydrochloric acid, after which the iron is oxidized to the ferric condition, unless it is already wholly present in that state. The iron is then separated, by double precipitation with ammonium hydroxide, as ferric hydroxide. The precipitate is ignited, and weighed as ferric oxide. From the combined filtrates and washings, which must have a large volume, and which must be free from nitrates, etc., the sulphate is precipitated by means of a dilute solution of barium chloride. The precipitate is ignited, and weighed as barium sulphate. A. Procedure for the Determination of Iron. Weigh out into dry 200 cc. beakers two portions of about one gram each, and add to each portion 50 cc. of water and 10 cc. of 6-normal hydrochloric acid, keeping the beakers covered with watch- 60 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS glasses to prevent loss by effervescence. Treat each solution as follows : Heat to boiling, and add, drop by drop, nitric acid (sp. gr., I.42), 1 until the brown coloration at first imparted to the liquid gives place to a yellow or red. (Note the volume of nitric acid which is added ; i cc. will always be found sufficient.) Boil for three minutes, and pour the solution, with stirring, into an excess of ammonia which has been diluted with water to a volume of 200 cc. (For this purpose, calculate the volume of 6-normal ammonium hydroxide required to neutralize the acids added, and use 5 cc. in addition.) Heat the mixture to boiling, and allow the precipitate to settle. Decant the boiling-hot, clear liquid through an ashless filter (9 cm. in diameter), leaving the precipitate as far as possible in the beaker, and wash twice by decantation with 50 cc. portions of very hot water, still leaving the precipitate in the beaker. (At once neutralize the filtrate and washings with hydrochloric acid, and reserve for the sul- phate determination. Their evaporation, as directed under the deter- mination of sulphur, should be begun at this point. See p. 4.) Dissolve the precipitate by pouring through the filter a boiling mixture of 5 cc. of water and 10 cc. of 6-normal hydrochloric acid, adding the acid in small portions, and collecting the filtrate (and washings) in the beaker containing the bulk of the precipitate, which also should completely dissolve. After thoroughly washing the filter, tear it into small bits and add these to the ferric chlo- ride solution. Pour this mixture into an excess of ammonia, as before, and heat to boiling. Filter boiling-hot through a fresh filter, wash the precipitate twice by decantation with hot water, and finally transfer it to the filter ; wash continuously with hot water until 3 cc. of the washings show no turbidity when treated with a drop of nitric acid and one of silver nitrate solution. (The combined filtrate and washings should at once be neu- tralized with hydrochloric acid and added to those from the first precipitation ; the evaporation should be allowed to continue.) Ignite the precipitate, together with the filter, in an inclined 1 Before adding nitric acid, see Note 2. GRAVIMETRIC ANALYSIS ' 6l platinum or porcelain crucible. After the volatile matter of the filter has been expelled, raise the temperature to the full heat of the burner, and, with the flame well at the base of the inclined crucible, continue the heating for about 15 minutes. Cool in the desiccator, and weigh. Repeat the heating until the weight is constant within 0.2 mg. Report the percentage of iron in the sample. NOTES. i. In ferrous salt solutions the iron is slowly oxidized by oxygen from the air (4 Fe ++ +0 2 +2 H 2 = 4 Fe +++ +4 OH~), and, unless the solution contains free acid to remove the OH~ ions, the iron will partially precipitate in the form of a basic ferric sulphate. Moreover, owing to hydrolysis, upon boiling an aqueous solution of ferric sulphate there results a partial precipitation of the iron, again in the form of a basic ferric sulphate (e.g. Fe 2 (SO 4 ) 3 +2H 2 0=2 Fe(OH)S0 4 +H 2 SO4). This action, however, is prevented by the presence of sufficient hydrochloric acid. 2. The complete oxidation of the iron is necessary, since ferrous iron is not quantitatively precipitated by ammonia. In the absence of air, in- deed, ammonium salts are capable of preventing entirely the precipitation of iron from ferrous salt solutions. Ferric iron, in the absence of organic matter, is completely precipitated by ammonia, even in the presence of ammonium salts. The nitric acid oxidizes the iron according to the equation : 6 FeS0 4 + 2 HN0 3 +6 HC1= 2 Fe 2 (S0 4 ) 3 +2 FeCl 3 +2 NO+4 H 2 0, and the dark color imparted to the solution is due to the union of the nitric oxide with ferrous salt which has not yet been oxidized, to form an unstable nitroso-compound similar to that formed in the "ring- test" for nitrates. The nitric oxide is expelled by heat, and the solution finally acquires the yellow color which is characteristic of ferric chloride in the presence of hydrochloric acid. To insure the presence of iron wholly in the ferric condition, a very small quantity of the oxidized solution should be tested on a porcelain plate with a drop of very dilute, freshly prepared potassium ferricyanide solution (a piece of the salt the size of a pinhead, in 20 cc. of water) . If the solution has a volume of 50 cc., and an ordinary drop a volume of 0.05 cc., then the loss of the latter would occasion an experimental error of 0.1% in the iron (and in the sulphur) determination. In this case, therefore, we might add a couple of drops of the solution to i cc. of water in a clean watch glass, and use one drop of this mixture for the test ; the remainder can then be washed back into the beaker containing the bulk of the solution. The error is thus reduced to 0.01%, which is negligible. 62 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Much time can often be saved by testing in this way the solution made from a very small portion of the (un weighed) original sample, for ferrous iron ; in its absence the addition of the nitric acid, and the subsequent evapo- ration to dryness, should be omitted. 3. If ammonia is added to ferric sulphate solution, the ferric hydroxide precipitate is apt to be contaminated with sulphuric acid, in the form of a basic ferric sulphate. A gradual neutralization with ammonia is almost sure to lead to the separation of an insoluble basic sulphate, owing to a deficiency of hydroxide ions. If, however, the iron solution is added with stirring to an excess of aqueous ammonia, the precipitate obtained will be comparatively free from basic sulphate. But, since even here the basic salt is likely to be present in small quantity, the precipitate is redissolved and the solution again added to an excess of dilute ammonia. 4. To avoid errors due to the solvent action of ammonium hydroxide upon glass, the precipitate should be filtered off without unnecessary delay. It is for this reason, also, that the filtrates and washings should be promptly neutralized with hydrochloric acid. The ferric hydroxide precipitate should under no circumstances be allowed to dry before the washing has been completed; it would be sure to crack, and in a subsequent washing the wash water would simply run through the crevices. 5. During the combustion of the filter, a portion of the precipitate may be reduced to FesO^ and it is essential that any of this substance should be oxidized back to ferric oxide. Therefore, during the ignition, it is im- portant that there should be a ready access of air to the precipitate. For this reason it is directed to macerate the filter with the solution of ferric chloride before the second precipitation ; this insures a very porous mass which is readily reoxidized. 6. The foregoing method may be used for the gravimetric determination of aluminum or chromium, with the additional precaution that the solution, before it is filtered, must be heated until but a very slight excess of ammonia remains, the hydroxides of these metals being more soluble in aqueous ammonia than ferric hydroxide. If it is desired by this method to determine the chromium in an alkali chromate, the latter is boiled with hydrochloric acid and alcohol, in order to reduce the chromium to the trivalent condition. (K 2 Cr 2 7 +3 C 2 H 6 0+8 HC1=2 KC1+2 CrCl 3 +3 C 2 H 4 0+ 7 H 2 0). The hydroxides of all three metals are precipitated by sodium or potas- sium hydroxide, but the precipitates are always contaminated with alkali. Furthermore, aluminum and chromium hydroxides dissolve readily in an excess of caustic alkali and form anions, to which the formulas A10 2 ~ and CrO 2 - are usually ascribed. When freshly precipitated, all three hydroxides GRAVIMETRIC ANALYSIS 63 dissolve in hydrochloric acid ; but aluminum hydroxide, after standing for some time, dissolves with considerable difficulty. While their precipitation is favored by the presence of ammonium salts (coagulation), it is entirely prevented by the presence of sufficient tartaric acid (formation of soluble com- plexes) ; citric acid, glycerol, sugars, etc. resemble tartaric acid in this respect. Upon ignition, each hydroxide yields an oxide suitable for weighing Fe 2 O 3 , A1 2 O 3 , Cr 2 3 . Chromic oxide, however, upon ignition, is partially oxidized to Cr 2 (CrO 4 )3; it should be ignited in a current of hydrogen (G. Rothaug, Zeitschrift jur anorganische Chemie, Vol. 84, pp. 165-189 (1913)). B. Procedure for the Determination of Sulphur. Evaporate to dryness, on the steam bath, the combined nitrates and wash- ings from the iron determination ; add to the residue 10 cc. of 6-normal hydrochloric acid; and again evaporate to dryness. 1 Dissolve the residue in 100 cc. of water, and filter the solution if it is not perfectly clear. Transfer the solution to a 700 cc. beaker, dilute to 400 cc., and then add 1.5 cc. of 6-normal hydrochloric acid. Heat the solution to boiling, and, with stirring, quickly pour in a boiling mixture of 15 cc. of i-normal barium chloride solution and 100 cc. of water. Continue the boiling and stirring for two or three minutes ; allow the precipitate to settle, and, at the end of half an hour, after testing for complete precipitation, decant the liquid through a filter. Substitute a clean beaker for the beaker containing the clear filtrate, wash the precipitate by decantation with hot water, and subsequently upon the filter with hot water until 3 cc. of the washings give no cloudiness or opalescence with a drop of silver nitrate solution and one of nitric acid. The precipitate is then dried and ignited to constant weight, with the flame well at the base of the inclined crucible. Report the percentage of S0 4 in the sample. NOTES. i. Barium sulphate, to a greater degree than most precipi- tates, tends to carry down other salts which are present in the solution from which it separates, and these substances cannot be removed by simply washing the precipitate. This is especially true of nitrates and chlorates, 1 In case the original sample contained no ferrous iron, and the addition of nitric acid was omitted, it is sufficient only to evaporate the neutralized nitrates and washings to about 300 cc. This liquid is then transferred to the beaker, diluted to about 400 cc., and treated further as described in the procedure. 64 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS and of the salts of trivalent metals, such as iron, chromium, etc. There- fore, if nitric acid has been used to oxidize the iron, it must be completely removed by evaporation with a large excess of hydrochloric acid. Iron is always present in the sulphate precipitated from hot solutions in the presence of ferric salts, and the precipitate then loses sulphuric acid upon ignition, and gives low results in spite of its iron content. Pure barium sulphate itself is not decomposed at a red heat, but suffers loss, probably of sulphur trioxide, at a temperature above 900. Barium sulphate requires about 400,000 parts of water for its solution, but it is more soluble in hydrochloric acid, even very dilute. In many salt solutions it is still more soluble than in water acidified with hydrochloric acid. 2. In the precipitation of sulphuric acid with barium chloride, the solu- tion should contain only salts of the alkali metals and ammonium, and it should be free from nitrates and chlorates. Even alkali salts and barium chloride are carried down to some extent by barium sulphate, more or less in proportion to their concentration, and consequently the solution should be dilute. Since, further, the solubility of barium sulphate, as well as the amount of barium chloride carried down, increases with the concentration of the hydrochloric acid present, the quantity of the latter should be re- duced to a minimum; some, however, must be present, since otherwise the precipitate would be very fine grained, and therefore difficult to filter. Barium sulphate carries down quantities of chlorine varying from traces to as much as i%, depending upon the conditions, and these should there- fore be very carefully regulated. For a quantity of sulphate corresponding to 1-2 g. of BaSC>4, the latter should be precipitated from a solution which has been diluted to about 400 cc., and which should contain 1.5 cc. of free 6-normal hydrochloric acid. This solution should be boiling hot, and, for each gram of barium sulphate, 10 cc. of i-normal barium chloride solution diluted to 100 cc., and boiling hot, should be poured in, all at once, with con- stant stirring. In this way exact results can be obtained, but only owing to a compensation of errors : although a very small quantity of barium sulphate remains dissolved in the acidified salt solution, an approximately equal weight of barium chloride is contained in the ignited precipitate. 3. Owing to the tendency of the precipitate to pass through the pores of the filter, the filtrate and washings should always be carefully examined for minute quantities of the sulphate. This is best accomplished by im- parting to the liquid a gentle rotary motion, so that, if present, the sulphate will collect at the center of the beaker. 4. A partial reduction of barium sulphate to sulphide may be caused by the action of the burning carbon of the filter ; in order to prevent the reduction of any considerable quantity, the crucible should not be heated GRAVIMETRIC ANALYSIS 65 above dull redness until after the carbon has been consumed. Subsequent ignition, with ready access of air, will then suffice to reconvert the sulphide to sulphate. If considerable sulphate is reduced, it may be necessary to moisten the precipitate with sulphuric acid, and then to heat cautiously until the excess of acid is expelled. THE DETERMINATION OF SULPHUR IN ORES Method. The ore is heated with strong nitric acid and potas- sium chlorate, which oxidize the sulphur to sulphuric acid. After removing the nitric acid and chlorate, as well as the iron, lead, etc., the sulphuric acid is precipitated with barium chloride, and weighed as barium sulphate. Procedure. Treat 0.25-0.50 g. samples of the finely pul- verized ore (depending upon the sulphur content) in 250 cc. Erlenmeyer flasks with 10 cc. of nitric acid (sp. gr. 142), and heat very gently until the red fumes have somewhat abated. Then increase the heat, and add to the quietly boiling liquid potassium chlorate, from time to tune, in o.i g. portions, until any free sulphur which has separated is entirely oxidized and dissolved; finally add 0.5 g. of solid sodium chloride and evapo- rate the solution to dryness. After cooling, cautiously add 10 cc. of hydrochloric acid (sp. gr. 1.19), heat gently until solution is as complete as possible, and evaporate to dryness. Take up in 5 cc. of strong hydrochloric acid, heat to boiling, and dilute with 100 cc. of cold water. To the cold solution add three drops of methyl orange, and ammonia to alkaline reaction ; then add 5 cc. more of ammonia and 10 cc. of ammonium car- bonate solution. Heat to boiling, allow the precipitate to settle in the hot liquid, and filter while still hot, washing thoroughly with hot water, and receiving the filtrate and washings in a yoo-cc. beaker. Neutralize the filtrate with hydrochloric acid, and add i . 5 cc. of the 6-normal acid in excess. Dilute the solution to 400 cc., heat to boiling, and add with stirring a boiling-hot mixture of 10 cc. of i -normal barium chloride solution and 100 cc. of water. Allow the mixture to stand for half an hour, test for complete pre- 66 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS cipitation, and finish the determination as described in the pre- vious Procedure. Report the percentage of sulphur in the sample. NOTES. i. Barium sulphate, if present in the ore, remains practically unaffected by the above acid treatment. If it is desired to determine the total sulphur hi ores containing barium, the hydrochloric acid solution, after the removal of nitrates and chlorates and dilution with 100 cc. of water, may be treated with 5 g. of solid ammonium chloride (to hold any lead in solution), heated to boiling, and filtered from the insoluble residue. The filter containing the latter is destroyed by ignition in a platinum crucible, and the residue fused with an.excess of sodium carbonate. The fusion is extracted with hot water and the residue washed with sodium carbonate solution; the filtrate and washings, which contain sodium sulphate, are added to the hydrochloric acid filtrate containing the bulk of the sulphur. The united filtrates are then treated with ammonia and ammonium car- bonate, as described above. 2. The small amount of sodium chloride is added before the first evapora- tion in order to prevent the possible loss of any free sulphuric acid which might be present, as, for example, in the analysis of pyrites. The potas- sium chlorate added would, however, probably be sufficient in most cases to accomplish this result. In the analysis of pyrites, which contains a very high percentage of sulphur, samples should be used of only 0.25 g. Otherwise the procedure is the same. 3. Upon adding ammonia in excess to the solution and heating to boiling, there is practically no danger of losing sulphur in the form of basic ferric sulphate. (In this connection see Note 3 of the foregoing Procedure.) The ammonium carbonate is added in order to remove any lead which may be present, as the carbonate, and thus prevent the loss of sulphur, as PbSC>4, before the precipitation with barium chloride. 4. In neutralizing a solution with the use of methyl orange, the solution should be cold, since otherwise the methyl orange is not a satisfactory indicator. 5. The student should be sure to read the notes on the determination of sulphur in iron sulphate, and also Problems vi, 12 and 13, of Part IV. THE DETERMINATION OF PHOSPHORIC ANHYDRIDE IN PHOSPHATE ROCK Method. The finely ground mineral is heated with nitric acid, the mixture evaporated to dryness, and the residue ex- tracted with hot nitric acid; the solution is then filtered from GRAVIMETRIC ANALYSIS 67 the insoluble silicious material. The filtrate is made almost neutral with ammonia, and is treated with a solution of am- monium molybdate, in excess, to separate the phosphoric acid from calcium, iron, aluminum, etc. The precipitated ammonium phospho-molybdate is washed with acidified ammonium nitrate solution, dissolved in ammonium hydroxide, and the phosphoric acid precipitated with magnesia mixture. The magnesium am- monium phosphate is finally ignited to magnesium pyrophos- phate, which is weighed. Procedure. Weigh out two portions of the finely ground mineral, not to exceed 0.2 g. each, into 2oo-cc. beakers, and treat each as follows. Pour over the sample 20 cc. of 6-normal nitric acid and warm gently until solvent action has ceased; then evaporate the mixture to dryness on the steam bath. Allow the dry residue to remain for half an hour on the steam bath, and then heat it again for a few moments with 20 cc. of the nitric acid. Filter off any siliceous residue and wash several times with small portions of hot water, receiving the filtrate and washings in a 4oo-cc. beaker. Finally test the washings with ammonia for calcium phosphate, but add to the original filtrate all such test solutions in which a precipitate appears. Cautiously, and with stirring, add ammonia to the filtrate and washings until the precipitate just fails to redissolve, then nitric acid, drop by drop, until the cloudiness disappears. The volume of the liquid at this point should not exceed 100 cc. Heat the solution until it cannot be held comfortably in the hand, remove the burner, and add 75 cc. of a freshly filtered solution of ammonium molybdate which has been gently warmed. Digest for an hour at 60-65, an< ^ then decant the supernatant liquid through a filter ; wash the yellow precipitate by decanta- tion with acid ammonium nitrate solution 1 (still keeping the bulk of the precipitate in the beaker), until 3 cc. of the washings give no test for calcium with ammonia and ammonium oxalate. 1 Made by mixing 50 cc. of 6-normal ammonium hydroxide with 100 cc. of 6-normal nitric acid, and diluting the mixture with 350 cc. of water. 68 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS The filtrate should not be thrown away, but should be tested for complete precipitation by renewed digestion with 5 cc. of moybdate reagent ; it should then be placed in a receptacle for " Molybdate Residues." Dissolve the ammonium phospho-molybdate by pouring over the filter four separate 10 cc. portions of a warm 2.5 per cent solution of ammonia l (afterwards washing the filter five tunes with 10 cc. portions of hot water), and receiving the filtrate and washings in the beaker containing the bulk of the precipitate. To the clear solution add hydrochloric acid, drop by drop, with stirring, until the yellow cloudiness produced disappears only slowly upon stirring. To this solution add 20 cc. of magnesia mixture from a pipette, at the rate of about i drop per second, with vigorous stirring (see Note 7 ; the last 10 cc. may be added somewhat faster). Let stand for 15 minutes, add 15 cc. of am- monia (sp. gr., 0.90), and then, after a period of 2 or 3 hours, de- cant the clear liquid through a filter and transfer the precipitate to the filter by means of 2.5 per cent ammonia water. Continue the washing with this liquid until 3 cc. of the washings, after acidification with nitric acid, give no opalescence with silver nitrate solution. Finally test the filtrate for complete precipi- tation. Dry the filter in the covered funnel, and then ignite, being careful to raise the temperature slowly and to insure the presence of plenty of air. Do not raise the temperature above moderate redness until the precipitate is white. Finally ignite to constant weight at the blast lamp, over a large Meker burner, or, prefer- ably, in an electric furnace. Report the percentage of P 2 5 in the sample. NOTES. i. The dehydration and removal of any dissolved silicic acid is necessary, since otherwise it would tend to partially separate with the phospho-molybdate precipitate, and render the latter more or less insoluble in ammonia. 1 Made by mixing 10 cc. of 6-normal ammonium hydroxide with 30 cc. of hot water. GRAVIMETRIC ANALYSIS 69 When washing the siliceous residue the filtrate may be tested for calcium by simply adding ammonia, which neutralizes the acid holding calcium phos- phate in solution and causes precipitation. 2. Nitric acid is used as the solvent because the phospho-molybdate is somewhat soluble in hydrochloric acid. Nitric acid exerts a slight solvent action, but this is counteracted by the presence of ammonium nitrate; hence the partial neutralization of the nitric acid with ammonia, and the washing with nitric acid containing an equivalent of ammonium nitrate. It should be noted that the molybdate reagent contains ammonium nitrate and free nitric acid. (See the Preparation of Reagents, in the Appendix.) 3. The precipitation of the phosphoric acid as magnesium ammonium phosphate from the original solution of the rock is not possible, owing to the presence of metals such as iron, aluminum, calcium, etc., which form phosphates insoluble in ammonia. For that reason the phosphoric acid is first separated from the metals, in the presence of nitric acid, by means of ammonium molybdate. 4. While the composition of the yellow precipitate varies somewhat with the conditions, it nevertheless seems to correspond pretty closely to the formula (NH 4 ) 3 P0 4 . 12 Mo0 3 . 2 HNO 3 . H 2 ; at any rate, the ratio P 2 O 5 : 24 Mo0 3 holds good. The yellow precipitate dissolves in ammonia to give ammonium phosphate and ammonium molybdate, and molybdic acid is not precipitated by magnesia mixture. 5. The precipitation of the phospho-molybdate is more prompt in warm than in cold solutions, but the temperature should not exceed 65 ; at higher temperatures molybdic acid, which is white, tends to separate. Vigorous stirring also promotes the separation of the yellow precipitate. A large excess of the molybdate reagent is required to effect a complete precipitation of the phosphoric acid. Theoretically 1.95 g. of Mo0 3 are required to combine with the phosphorus in 0.2 g. of rock containing 40 per cent of P2O 5 ; while the quantity of the reagent actually used (75 cc.) contains about 5 g. of MoOg. The presence of ammonium nitrate in the solution is also conducive to complete precipitation. These substances, by mass action, prevent the partial dissociation of the complex into its more soluble constituents. 6. If magnesium ammonium phosphate is washed with pure water, it is hydrolyzed according to the equation, MgNH 4 P0 4 +H 2 ^ MgHPO 4 +NH 4 OH. It thereby loses its crystalline form, and is almost sure to give a cloudy filtrate. In the presence of ammonium hydroxide, however, this decom- position is prevented by mass action. 70 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS 7. The precipitate of magnesium ammonium phosphate should be per- fectly crystalline ; the slow addition of the reagent, with constant stirring, is essential to this end, but the stirring rod should not be allowed to scratch the beaker. A large excess of magnesia mixture tends to cause the precipitate to carry down molybdic acid, as well as magnesia (shown by a persistently flocculent precipitate). In such cases the precipitate should be redissolved by adding a small quantity of hydrochloric acid, the solution treated with 2 cc. of magnesia mixture, and the hot liquid slowly neutralized with 2.5 per cent ammonia. Strong ammonia is then added, and the analysis continued as above. "Magnesia Mixture" is prepared by putting together in solution mag- nesium chloride, ammonium chloride, and ammonium hydroxide. The function of the ammonium chloride is to prevent the precipitation of mag- nesium hydroxide, so that the composition of the precipitate may corre- spond to the formula, MgNH 4 P0 4 . 6 H 2 0. 8. Upon ignition, the magnesium ammonium phosphate gives off am- monia and water, and is converted into magnesium pyrophosphate: 2NH 4 MgP0 4 .6H 2 = Mg 2 P 2 7 +2NH 3 +i3H 2 O. The precautions de- tailed hi Part I should be observed with great care during the ignition of this precipitate. There is danger here of a partial reduction of the phos- phate by the ammonia or by the carbon of the filter, and also, if too soon heated very strongly, the precipitate becomes glazed over and it is then practically impossible to remove the carbon by further heating. The precipitate is much more readily ignited to whiteness in platinum than hi porcelain ; but in case platinum is used, especial care should be taken to provide a plentiful supply of air. Reduction of the phosphorus would play havoc with the crucible. The most satisfactory procedure is to filter off the magnesium ammonium phosphate in a Munroe crucible of platinum (with a platinum sponge filter) , in which the precipitate can be ignited without danger of loss. If a good muffle furnace (preferably electric) is available, a Gooch crucible of porce- lain may be used with advantage. THE DETERMINATION OF CALCIUM AND MAGNESIUM OXIDES IN LIMESTONE Method. The hydrochloric acid extract of the limestone x is freed from dissolved silica, treated with bromine water and ammonia, to remove iron, aluminum, manganese, etc., and from 1 See Note i. GRAVIMETRIC ANALYSIS 71 the filtrate the calcium is precipitated with ammonium oxalate, the precipitate being ignited to the oxide. The filtrate from the calcium oxalate is treated with sodium phosphate and am- monia, the precipitate of magnesium ammonium phosphate being ignited to magnesium pyro-phosphate. A . Procedure for the Determination of Calcium. Weigh out into two casseroles 0.5-0.6 g. samples of the finely ground rock, and treat each as follows : Cautiously moisten the powder with 5 cc. of water, cover the casserole, add 10 cc. of 6-normal hydrochloric acid in small portions, and evaporate to dryness on the steam bath. Pour over the residue 5 cc. of water and 10 cc. of the hydrochloric acid, evaporate to dryness, and heat the dry residue for half an hour on the steam bath. Pour over this residue 5 cc. of wate'r and 10 cc. of the 6-normal acid, and heat gently for 10 minutes; filter and wash twice with 5 cc. portions of dilute hydrochloric acid, and finally with water until free from chlorides. (See Note i in regard to the insoluble residue.) Add to the filtrate and washings enough bromine water to impart a distinctly yellow tinge, boil, and then add ammonia until its odor persists in the solution. Heat until the excess of ammonia is largely expelled, and filter promptly. Wash the filter twice with hot water, allowing the washings to run into the beaker containing the filtrate. Now pour through the filter 25 cc. of hot hydrochloric acid (one volume of the 6^normal acid to 5 of water), and if there is a brown insoluble residue allow it to remain on the filter ; the acid solution should be re- ceived in the beaker in which the ammonium hydroxide pre- cipitate was obtained. Wash the filter five times with hot water, and then reprecipitate the iron, etc., from the filtrate and washings with bromine water and ammonia as already described. Collect the precipitate on the filter already used, and wash it free from chlorides with hot water ; add the filtrate and washings to those at first obtained. (Concerning the precipitate, see Note i.) Evaporate the combined filtrates and washings to a volume of about 200 cc. Heat the solution to boiling; if necessary, add 72 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS ammonia until its odor is plainly perceptible; and then add ammonium oxalate solution slowly and with stirring, in moderate excess. Boil for two minutes, allow the precipitate to settle for half an hour, and decant through a filter into a beaker, wash- ing the precipitate twice with hot water containing a few cubic centimeters of ammonium oxalate solution and a very little ammonia. Test the filtrate with ammonium oxalate for com- plete precipitation, and if no precipitate forms in 15 minutes, acidify the solution with hydrochloric acid and reserve it for the magnesium determination. Redissolve the calcium oxalate with warm hydrochloric acid (one volume of the 6-normal acid to one of water), pouring the acid through the filter and receiving it in the beaker containing the bulk of the] precipitate. Wash the filter three times with water, and twice with very dilute ammonia. Dilute the solu- tion to 250 cc., heat to boiling, add i cc. of ammonium oxalate solution, and ammonia in slight excess; boil for two minutes, and set aside for half an hour. Filter off the precipitate upon the filter previously used, and wash it free from chlorides with hot water containing a few drops of ammonium oxalate solution and a very little ammonia. (The filtrate and washings should at once be acidified with hydrochloric acid and combined with those from the first precipitation. If not already started, the evaporation of these filtrates should be begun at this point.) Gently ignite the dried precipitate and filter until the latter is consumed, and then heat with the full flame of the burner for 45 minutes ; finally heat for three minutes over the blast lamp. Repeat the heating at the blast lamp, until the weight becomes constant. (A muffle furnace is preferable for the ignition, after the filter is consumed.) Report the percentage of CaO found. NOTES. i. The chief component of limestone, calcium carbonate, is readily attacked by hydrochloric acid, as also are some of the other com- ponents ; but few limestones are so pure as to dissolve completely in hydro- chloric acid. The residue may contain quartz, silicates, pyrites, or other refractory materials, and carbonaceous matter may also be present. GRAVIMETRIC ANALYSIS 73 Furthermore, the insoluble silicates of the residue are apt to contain some calcium and magnesium. The thorough analysis of a limestone necessi- tates the use of an elaborate system of procedures and the determination of numerous substances, but for many - technical purposes the analysis may be confined to the determination of the insoluble matter and silica, of the oxides of iron and aluminum (including small quantities of the oxides of titanium, manganese, phosphorus, etc.), of calcium oxide, and of mag- nesium oxide. In this exercise, the insoluble residue, if ignited and weighed, would give a more or less accurate approximation of the insoluble matter and silica, and the ignited ammonium hydroxide precipitate would roughly approxi- mate the summation of the oxides of iron, aluminum, titanium, etc., in the soluble portion. It should be remembered, however, that substances such as hydrous silicates, pyrites, and carbonaceous matter in the insoluble residue, and ferrous and manganous oxides (originally present as carbonates) in the soluble portion, would not be correctly indicated by this method. It should be noted that the amount of insoluble residue and also its character will often depend upon the concentration of the acid used for the solution of the limestone, and that the determination of this residue is essentially empirical. For a description of the complete analysis of limestones, the student should refer to Bulletin 422 of the United States Geological Survey, by W. F. Hillebrand. 2. Some of the silicates present are apt to be at least partly decomposed by the acid, and the soluble silicic acid must be dehydrated and rendered insoluble by evaporation and heating. The residue is washed first with dilute acid to prevent the separation on the filter of basic salts of iron, aluminum, etc., owing to the hydrolytic action of water. 3. The addition of bromine water serves to oxidize any ferrous iron, and also manganese, which precipitates as MnO(OH) 2 . The ammonium hydroxide precipitate should be filtered off promptly, since the alkaline solution absorbs carbon dioxide from the air, with the consequent pre- cipitation of a little calcium as the carbonate. This is always possible, and for that reason, as well as because the precipitated hydroxides also tend to carry down the hydroxides of calcium and magnesium, the precipi- tate is redissolved and again precipitated, to free it from these metals. 4. The accurate separation of calcium and magnesium by means of ammonium oxalate requires considerable care. The calcium oxalate tends to carry down some magnesium oxalate, probably in the form of a double salt, but this can be removed by dissolving the precipitate and reprecipi- tating the calcium in the presence of only this small amount of magnesium. 74 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS If the proportion of magnesium is not very large, the calcium can be sepa- rated by precipitation from a rather dilute solution, with the addition of more than enough ammonium oxalate to convert both the magnesium and calcium into oxalates. (In this connection, see T. W. Richards, C. T. McCaffrey, and H. Bisbee; Zeitschrift fiir anorganische Chemie, 28, p. 71 (1901).) 5. The small quantity of ammonium oxalate solution is added before the second precipitation of the calcium, because an excess of the reagent reduces the solubility of the calcium oxalate, and also tends to hold the magnesium in solution in the form of a double magnesium ammonium oxalate. For the first reason, the precipitate is washed, not with pure water, but with water containing ammonium oxalate and ammonia. These sub- stances are volatilized in the ignition. 6. Calcium oxalate is practically insoluble in water (5.6 mg. of the anhydrous salt per liter of saturated solution), and only very slightly soluble in acetic acid, but it is readily dissolved by the strong mineral acid?. This behavior with acids is explained by the fact that oxalic acid lies about halfway in strength between acetic acid and the strong mineral acids. In acetic acid solution, the hydrogen-ion concentration is too low to appre- ciably diminish the concentration of C 2 4 ions, and practically no solvent action takes place. In the solution of a strong mineral acid, however, the high hydrogen-ion concentration causes the calcium oxalate to dissolve according to the following scheme : CaC 2 O 4 ^ CaC 2 4 ^ Ca ++ + C 2 O 4 ~ 1 f HC 2 4 - or (solid) (dissolved) | ^T HC1 $ Cl- + H+ J I H 2 C 2 4 . The oxalate is immediately reprecipitated from such a solution upon the addition of a base; the hydroxide ions unite with the hydrogen ions of both the mineral acid and the oxalic acid to form water, and the Ca ++ and C 2 4 ions left in the solution recombine to form CaC 2 4 . (Compare the precipitation of Ca 3 (P0 4 ) 2 from the acid solution of apatite by am- monium hydroxide, in the preceding exercise.) 7. Upon ignition, calcium oxalate becomes anhydrous slightly above 1 80, and at low redness it is decomposed into calcium carbonate and carbon monoxide. Strong ignition converts the carbonate into the oxide ; in a platinum crucible, this conversion may be carried to completion over a large Meker burner. With porcelain crucibles, however, an electric fur- nace is to be preferred. Since calcium oxide absorbs moisture and carbon dioxide from the air, it should be weighed as rapidly as possible. GRAVIMETRIC ANALYSIS 75 8. By burning off the filter and then evaporating with 2-3 cc. of 6-normal sulphuric acid, the calcium oxalate may be converted into sulphate, and this may be heated to constant weight. While this procedure is preferred by some analysts, it is nevertheless disadvantageous, since it involves danger of loss by spattering. Moreover, calcium sulphate is more readily decom- posed upon ignition than barium sulphate, and there is some danger of loss on this account. B. Procedure for the Determination of Magnesium. Evapo- rate the acidified filtrates and washings from the calcium oxalate on the steam bath until the salts begin to crystallize. Dilute the solution cautiously with small portions of water, and with stirring, until the salts are brought back into solution, adding a little hydrochloric acid if the solution has been evaporated to a very small volume. (If the acid solution contains solid matter at this point, it should be filtered.) Carefully add ammonia to the clear solution, just to alkaline reaction (methyl orange) ; add sodium ammonium phosphate solution, drop by drop with stirring, as long as a precipitate continues to form, and then 10 cc. in excess. Finally add to the solution one third of its volume of 6-normal ammonia, stir vigorously for 10 minutes, and allow the mixture to stand for at least 6 hours, preferably overnight. Decant the solution through a filter, and wash the bulk of the precipitate on to the filter with 2.5 per cent ammonia (one volume of 6-normal ammonia to three of water) ; do not bother to clean the beaker completely. Dissolve the precipitate from the filter with the least possible quantity of hydrochloric acid (6-normal acid diluted with twice its volume of water), receiving the acid solution in the precipitation beaker. Wash the filter with small portions of hot water until the washings are free from chlorides. Add to the combined filtrate and washings 2 cc. of sodium ammonium phosphate solution and then aqueous am- monia, drop by drop with constant stirring, until the liquid smells distinctly of ammonia. Stir for two minutes, add to the solution one third its volume of 6-normal ammonia, and allow the mixture to stand for 2 hours. Decant the clear liquid through a filter and transfer the precipitate to the filter by means of 2.5 per 76 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS cent ammonia. Continue the washing with this liquid until 3 cc. of the washings give no opalescence with nitric acid and silver nitrate. Dry the filter completely in the covered funnel, and then ignite, taking great pains to raise the temperature very slowly and to insure the presence of plenty of air. Do not raise the temperature above moderate redness until the precipitate is white. Finally ignite to constant weight over the blast lamp or in a muffle furnace. Report the percentage of MgO found. NOTES. i. The filtrates from the calcium oxalate should be slightly acidified immediately after filtration, in order to avoid the solvent action of the alkaline solution upon glass. 2. The precipitation of the magnesium should be made in a small volume of liquid, and the ratio of ammonia to the total volume of solution should be carefully regulated, on account of the relative solubility of the magnesium ammonium phosphate. (Compare Note 6, under the determination of Phos- phoric Anhydride.) 3. In the presence of ammonium salts in large quantity, the first precipi- tate is rarely wholly crystalline ; it is apt to contain the mixed phosphate, Mg[(NH 4 ) 2 PO4]2, which upon ignition leaves magnesium metaphosphate Mg(P0 3 ) 2 ; and, if produced in the presence of too much ammonia, it may also contain some tri-magnesium phosphate, which upon ignition remains unchanged as Mg 3 (PO 4 ) 2 . Such precipitates can be purified by dissolving in a very little hydrochloric acid, adding a small quantity of sodium am- monium phosphate, and reprecipitating the magnesium, in the practical absence of ammonium salts, by means of ammonia. 4. In order to avoid a partial reduction of the phosphorus, the precipi- tate should be heated gently at first, until all the ammonia is expelled and the filter consumed. Also, if heated too soon to a bright red heat, the precipitate becomes glazed over, and it is then impossible to remove the carbon by further heating. Concerning the use of Munroe or of Gooch crucibles, see Note 8 under the determination of phosphoric anhydride. THE DETERMINATION OF CARBON DIOXIDE IN LIMESTONE Method. The weighed carbonate is placed in an apparatus which contains acid in a separate compartment; the whole apparatus is then weighed. After this the acid is run in upon GRAVIMETRIC ANALYSIS 77 the carbonate, and the carbon dioxide set free is removed from the apparatus through a tube filled with calcium chloride, which prevents the escape of moisture from the apparatus. Finally, the apparatus is weighed again, and the loss in weight indicates the quantity of carbon dioxide in the sample. Many different forms of apparatus have been devised for this purpose. The one shown in the accompanying figure is an im- proved form of the so-called alkalimeter of Mohr. It consists of a small, wide-mouthed, flat-bottomed flask F, which has a ground-glass connection with the tubes A and B, which are for acid and calcium chloride. The ground-glass joints are lubri- cated with a mixture of vaseline and beeswax, or other suitable substance. Procedure. Thoroughly clean the apparatus, allow it to drain, and finally dry it by gently heating the flask while drawing a current of dry air through it. As aspirator an inverted wash- bottle (shown in the figure on a very much reduced scale) may be used, from which the water is caused to run out slowly through the shorter tube. During aspiration the calcium chloride tubes C and D should be connected with c and d, as shown in the figure, so that no moisture may enter the apparatus. After drying the apparatus, place a loose wad of cotton at the bottom of B ; introduce into the neck of the tube a cylinder of glazed paper about 3 cm. wide, and through this cylinder pour in small pieces of calcium chloride until the tube is about three fourths full; remove the glazed paper, taking care to keep the upper walls of the tube free from calcium chloride. Place another cotton wad in the tube, insert the stopper, and close the tube temporarily at d by means of a short piece of glass rod within rubber tubing. 1 1 The tube must be kept closed when not in use, to prevent the gradual absorp- tion of moisture from the air. Each of the ordinary calcium chloride tubes pre- viously mentioned is filled in the same way about two thirds full, but in this case a softened cork stopper, pierced by a short piece of glass tubing with rounded ends, is introduced and shoved far into the tube with the help of a stirring rod, leaving the outer 2 or 3 mm. empty. This space in the tube is filled with molten sealing- 78 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS GRAVIMETRIC ANALYSIS 79 When all is ready, weigh out into the flask about 1.5 g. of the finely powdered substance, which has been dried at 100 C. and allowed to cool in a desiccator, and add 3-4 cc. of water. Close the stopcock T and fill the tube A about three fourths full with hydrochloric acid (i volume of 6-normal acid to 1.5 volumes of water) by means of a small funnel. The whole apparatus, with the tubes open at c and J, is now accurately weighed ; the two calcium chloride tubes are connected at c and d\ and the stopcock T is slightly opened so that the acid from A slowly drops into the flask. As soon as the evolution of carbon dioxide begins to take place quietly, the apparatus is allowed to stand without watching for about half an hour. All of the acid will then have entered the flask, and the decomposition will be practi- cally complete. It now remains to remove the carbon dioxide absorbed by the liquid and contained in the apparatus. To this end, connect the calcium chloride tube D with the wash bottle W, as shown in the figure (the wash bottle is of course much larger than the figure would indicate), and regulate the flow of water through e so that not more than 3 or 4 bubbles of air per second pass through the flask F. Then heat the flask F gently, by means of a small flame, until the acid just begins to boil ; at once remove the flame, and continue to aspirate air through the apparatus until it is. cold. Stopper the tubes at c and dj wipe the apparatus with a clean dry towel, and allow it to stand for one half hour near the balance. Finally, remove the stoppers from c and d, and weigh the apparatus. Report the percentage of C02 found. NOTES. i. This method yields excellent results in the estimation of large amounts of carbonic acid such as are present in limestones and baking powders. But it is unreliable for the determination of small quantities, e.g. in cements. 2. Since baking powders are decomposed by water, they should be kept dry until after the apparatus has been weighed ; and since their efficiency wax, so that an air-tight connection is made. These tubes also are closed, when not in use, by glass rods within rubber tubing. 8o INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS as leavening agents depends upon the volume of gas liberated under the conditions of actual usage, water should be placed in the tube A, instead of acid. Otherwise the procedure is the same. The loss in weight is then a measure of the available carbon dioxide of the sample. 3. Carbon dioxide is readily displaced from the apparatus by the method described, but in order to insure its complete removal at least a liter of air should be drawn through the apparatus. The small tube from A should project well below the surface of the liquid, for in order to remove carbon dioxide efficiently from the solution, the air must be made to bubble through the liquid. 4. Since commercial calcium chloride is apt to contain free lime, it should for the best results be treated with carbon dioxide before the deter- mination is made. For this purpose a current of the dry gas is passed through the apparatus for a minute or two, the tubes at c and d are closed, and the apparatus allowed to stand overnight. The carbon dioxide is then removed by aspirating dry air through the apparatus for about 20 minutes, after which the sample may be placed in the flask. 5. The most serious objection to this method is the fact that, owing to the size and weight of the apparatus, there is likely to be an appreciable error in the difference between the two weights. This danger, however, can be largely overcome if a similar piece of apparatus is available as a tare. THE DETERMINATION OF SILICA IN A REFRACTORY SILICATE Method. The finely ground sample is fused with an excess of sodium carbonate, whereby it yields sodium silicate and other compounds, depending upon the nature of the mineral. The melt is then decomposed with hydrochloric acid, which should dissolve everything except a portion of the silicic acid. Upon evaporating the liquid, the silicic acid in solution loses water and becomes much less soluble; upon extracting the residue with hydrochloric acid, filtering from silica, and evaporating the filtrate, however, appreciable amounts of silica are recovered in a second filtration, leaving negligibly small amounts in the filtrate. Upon strongly igniting the precipitates, the silica (contaminated with iron oxide and alumina) is left in the anhydrous condition. After weighing, this is evaporated with hydrofluoric acid and a few drops of sulphuric acid, and the GRAVIMETRIC ANALYSIS 81 residue is subjected to strong ignition. The weight of the im- pure silica less that of the ignited residue gives the weight of the silica originally in the sample. Procedure. Grind about 3 g. of the material in an agate mortar until it will entirely pass through a sieve of fine silk bolt- ing cloth. 1 Weigh out into two platinum or palau (also called rhotanium) crucibles portions of the silicate of about 0.75 g. each. Also weigh out, on a rough balance, two portions of anhydrous sodium carbonate of about 4 g. each. In each case, add about three fourths of the sodium carbonate to the silicate sample in the crucible, place the latter on a piece of glazed paper, and thoroughly mix its contents with a dry glass rod. Place the remaining fourth of the flux on top of the mixture, after first stirring it with the rod to remove from the latter any adhering particles of the mixture. Cover the crucible and heat it gradually to the highest heat of the Bunsen or Tirrill burner, and then, if necessary to secure complete fusion, heat the mixture over a Meker burner or a blast lamp. As soon as the mass is in quiet fusion, evolving no gas bubbles, take up the crucible in tongs applied to the upper edge, and, by means of a slow rotary motion, cause the liquid melt to spread around the walls of the crucible, where it will solidify. As soon as this takes place, and while the mass is still red-hot, plunge the lower portion of the crucible for a few seconds into cold water, but with care not to allow any water to enter the crucible. Then set the crucible aside to cool. The solid material may then be loosened from the crucible by gentle tapping. (Do not deform the crucible?) Place the solid melt in a rather tall beaker, add 100 cc. of water and, with stirring, gradually add 50 cc. of 6-normal hydro- chloric acid. Also clean the crucible and lid with a little of the acid, and add this to the main portion in the beaker. (In case 1 Place the ground material in a small beaker and stretch over the top a piece of the bolting cloth, fastening the cloth in place by means of a rubber band below the rim of the beaker. By gently tapping the inverted beaker over a piece of clean paper, the fine particles are caused to pass through the sieve. The coarser par- ticles which fail to pass through must be returned to the mortar and reground. G 82 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the melt adheres obstinately to the crucible, place both in the beaker and treat with water and acid as described.) Heat the beaker gently, and, if necessary, aid the disintegration of the melt by gentle pressure with the broadened end of a glass rod. After complete disintegration, transfer the mixture to a porce- lain casserole, evaporate to dryness on the steam bath, stirring frequently towards the end until the residue is a dry powder. Heat the dry residue on the steam bath for at least an hour, cover it with 5 cc. of i2-normal hydrochloric acid, warm gently, see that the residue is wholly moistened with acid, add 100 cc. of water, and heat to boiling. Filter promptly, and wash five times with hot dilute acid (i volume of 6-normal hydrochloric acid to 3 of water), collecting the nitrate and washings in a por- celain casserole ; evaporate to dryness, heat the dry residue on the steam bath for one hour, and proceed as before, using a fresh filter to remove the silica. Wash the filter with hot dilute hydrochloric acid, as before, and then wash both filters with hot water until the washings are free from chlorides. Transfer both filters to a weighed platinum (or palau) crucible, and ignite cautiously until the paper is consumed, then at the full heat of the burner for half an hour., Moisten the cold residue with 2 or 3 drops of strong sulphuric acid, heat cautiously to expel the free acid, ignite to low redness, and finally for half an hour over the blast lamp. Repeat the blasting for periods of 5 minutes, to constant weight. Now add to the silica in the crucible i cc. of 6-normal sul- phuric acid and 3 cc. of pure hydrofluoric acid. (This acid should not be allowed to come in contact with the skin, as it produces painful wounds.) Evaporate as far as possible on the steam bath in a well-drawing hood, adding more hydrofluoric acid if any solid residue remains. Cautiously fume off the sul- phuric acid, heat to low redness, and finally ignite over the blast lamp, for 5-minute periods, to constant weight. Deduct the weight of this residue from that of the impure silica, and from the difference calculate the percentage of SiO 2 in the sample. GRAVIMETRIC ANALYSIS^ 83 NOTES. i. The whole of the sample must be ground very fine, or the coarser particles will resist the action of the flux ; unless all of the material is passed through the bolting cloth, the sifted portion may not represent an average sample. 2. Upon fusion with sodium carbonate, silicates are decomposed with the evolution of carbon dioxide. The other products of the decomposition are sodium silicate and aluminate, ferrous carbonate or ferric oxide, cal- cium and magnesium carbonates, etc. Owing to the evolution of gas during fusion, the heating should be gradual and the crucible should be kept covered. 3. Upon disintegrating the mass with a considerable volume of dilute acid the silicic acid at first largely enters the solution, but upon evaporation it is rendered almost insoluble. Treatment of the fused mass with strong acid would be likely to cause the separation of gelatinous silicic acid which would inclose metallic salts and withhold them from the solution. 4. A gritty residue remaining after the disintegration of the melt with acid indicates that the original silicate has been but imperfectly decom- posed. In such a case the fusion should be repeated with another sample, which should be sufficiently well ground and thoroughly mixed with the flux. 5. Silicic acid cannot be rendered wholly insoluble by a single evapora- tion and heating; nor are repeated evaporations and moistenings before filtration as effective in separating the silica as are alternate evaporations and filtrations. The underlying causes are as yet obscure. 6. To free the silica as far as possible from mineral salts, the residue after evaporation should be thoroughly extracted with warm hydrochloric acid ; and the solution should be diluted to a large volume to prevent the inclosure of impurities by the silica. The silica is first washed with dilute acid, to prevent the partial separation of basic salts of iron, aluminum, etc., by hydrolysis ; the washing is then completed with hot water. 7. The finely divided silica holds moisture so tenaciously that prolonged ignition over the blast lamp is necessary. Even then the ignited powder tends to absorb moisture, and it should therefore be weighed as rapidly as possible. 8. Notwithstanding all the precautions, the ignited silica is rarely pure. Upon evaporation with hydrofluoric and sulphuric acids, however, the silica is volatilized as silicon tetrafluoride and water, and a sulphate residue is left. If the contaminating substance is an alkali salt, as sodium chloride, the residue will remain as sulphate, even at high temperatures ; but certain other sulphates, as those of iron, aluminum, and titanium, evolve sulphur trioxide on ignition and leave the corresponding oxides. In the estimation of silica, the weight of impurities in the silica is always determined by weighing the residue from the hydrofluoric and sulphuric acid treatment; 84 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS in order then that the impurities weighed with the silica may be as nearly as possible identical with the residue from this treatment, it is best to treat the silica before ignition with a few drops of sulphuric acid. The final residue from the hydrofluoric and sulphuric acids should also be subjected to the same temperature employed in the ignition of the silica. 9. The procedure for the determination in the united filtrates from the silica of the mixed oxides of iron, aluminum, etc. and of calcium and mag- nesium, does not differ materially from that given under the determination of calcium and magnesium in limestones. 10. For a thorough study of the analysis of silicate and carbonate rocks, the student is referred to Bulletin No. 422 of the United States Geological Survey, by W. F. Hillebrand. THE DETERMINATION OF POTASH IN SOLUBLE SALTS The sample may be a pure salt, a soluble industrial product, or an artificial mixture of potassium chloride and sodium car- bonate. Principle. The determination of potassium by this method depends upon the insolubility of potassium per chlorate, and the solubility of sodium and certain other perchlorates in 96% alcohol. This is not a precipitation method, but one of ex- traction. If heavy metals are present, they are first removed. Procedure. Weigh out samples sufficient to contain about 0.25 g. of K 2 0, into 50 cc. beakers, and treat each as follows : Warm the sample with 25 cc. of water, and, if sulphates are absent, filter into a loo-cc. porcelain dish ; if, however, sulphates are present, acidify with 6-normal hydrochloric acid, stir, treat the hot acid liquid with barium chloride solution in slight excess, filter into a loo-cc. porcelain dish, evaporate the filtrate to dryness on the steam bath, and warm the residue with 15-20 cc. of water, with stirring. Add to the solution sufficient perchloric acid to contain 1.7 times the sample's weight of HC104 (See Note 2), and evaporate to a sirupy consistency. Add 15 cc. of hot water and 2 cc. of perchloric acid, and again evaporate. Once more add 15 cc. of hot water, and evaporate until heavy fumes of perchloric acid appear. GRAVIMETRIC ANALYSIS 85 Allow the mixture to cool thoroughly, add 20 cc. alcohol con- taining 0.2% by weight of HC1O4, 1 and stir for some time, keeping the salt as coarsely granular as possible. Let settle, decant the liquid through a weighed Gooch crucible 2 (containing a mat moistened with the wash liquid), and to the residue add a second 20-cc. portion of the wash liquid. Stir, let settle, again decant, and then drive off the remaining alcohol on the steam bath. Dissolve the residue in 15 cc. of hot water, add a few drops of perchloric acid, and evaporate to heavy fumes. Cool, add i cc. of the wash liquid, decant, and test a few drops of the washings for complete extraction. (The extraction with the alcoholic liquid must be continued until a few drops of the filtrate leave no residue when evaporated to dryness on platinum foil.) Finally, cool, add i cc. of the wash liquid, and sweep the salt into the Gooch crucible with a policeman, washing at last with a very little pure 96% alcohol. Dry the salt for half an hour at 130, and weigh. Report the percentage of K 2 in the sample. NOTES. i. This method was proposed in 1831 by Serullas, but, owing to some mistaken ideas concerning the properties of perchloric acid, the proposition did not receive the attention it deserved. Perchloric acid solutions of satisfactory grade can now be obtained in the market, they can be kept indefinitely in glass-stoppered bottles, and the method rivals in results the chloroplatinic acid process ; and this at a greatly reduced cost. 2. The specific gravities of perchloric acid solutions are as follows: 70% HC10 4 , 1.67 ; 60% HC1O 4 , 1.54 ; 50% HCLO 4 , 1.41 ; 30% HC10 4 , 1.20 ; 20% HC10 4 , 1. 1 2. The strength of a solution of the pure acid may easily be determined by the dilution of a known amount and titration with sodium hydroxide, with phenolphthalein as indicator. 3. In order to obtain the potassium as pure KC1O 4 by this method, it is essential that no strong acids be present, other than perchloric acid, which yield salts insoluble in alcohol. Sodium chloride and sulphate are such salts, and it is therefore necessary to remove chlorides and sulphates before 1 Made by mixing 1.7 cc. of the 60% acid, or 4.4 cc. of the 30% acid, with one liter of 96% alcohol. 2 It is better to use a Munroe crucible, with a filter of platinum sponge. The crucible itself may be of gold, to save expense. 86 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the treatment with alcohol. HC1 may be removed by repeatedly evaporat- ing the aqueous solution with the less volatile HC1O 4 ; but H 2 S0 4 is less volatile than HC1O4, and cannot be expelled in this way. Before the first evaporation, therefore, the latter should be precipitated from the hot acid solution by means of BaCl2 in slight excess. Phosphates, though often insoluble in alcohol, need not be removed ; but, in their presence, a larger excess of HC1O 4 should be used, to insure their complete removal by the wash liquid, as H 3 P0 4 . (See Note 5 of this procedure, and also Note 6 under the determination of calcium.) 4. Since NH 4 C10 4 is only sparingly soluble in alcohol, ammonium salts should be carefully expelled by gentle ignition, before the treatment with HC1O 4 . Moderate amounts of barium, calcium, and magnesium do not interfere with the procedure ; their perchlorates are soluble in alcohol. 5. The perchlorate mixture is extracted with alcohol containing a small amount of HC10 4 because, owing to the common ion effect, the solubility of KC10 4 is less in it than in pure alcohol ; the two solubilities are about 4 mg. and 16 mg. per 100 cc., respectively. The solubility of KC10 4 is still less in the presence of sodium, and other soluble perchlorates ; i.e. in the first portions of the alcoholic extract. Alkali and alkali-earth phosphates are decomposed by perchloric acid and the H 3 P0 4 dissolves in the acid-alcoholic liquid ; in the presence of phosphates, therefore, perchloric acid should be present in considerable excess. 6. If a known weight of NaCl KC1 mixture, obtained for example in a silicate analysis, is converted into a mixture of the perchlorates, and the KC10 4 isolated and weighed, the method yields both the K 2 and Na 2 O contents of the original sample. (Cf. Part IV, Problem 29.) 7. In order to prevent the loss of KC10 4 , in the separation of sodium and potassium, it has been suggested to extract the perchlorate mixture with an alcoholic liquid which has previously been saturated with KC10 4 . This procedure, however, is apt to lead to high results, owing to the precipi- tation of small amounts of potassium from the wash liquid by the NaC10 4 entering into solution ; it is therefore not to be recommended. In order to obtain exact results, it suffices to avoid the use of unnecessary quantities of the wash liquid. THE ELECTROLYTIC DETERMINATION OF COPPER The sample to be analyzed may be pure copper sulphate, an artificial mixture of the carbonates of copper and sodium, a copper ore, or a nickel coin. In case stationary electrodes are employed, the solution should contain not over 0.2 g. of copper GRAVIMETRIC ANALYSIS 87 and 5 cc. of nitric acid (sp. gr., 1.42), and should have a volume of 100 cc. ; in the case of a rotating anode, however, the solution may contain as much as 0.5 g. of copper, and it should contain, in a volume of 100 cc., 3-5 cc. of nitric acid (sp. gr., 1.42), or i cc. of sulphuric acid (sp. gr., 1.84) and 3 g. of ammonium sulphate. Method. The copper salt is decomposed by the electric current, and the copper deposited upon the cathode (negative electrode). The cathode is weighed before and after the opera- tion, and the increase in weight indicates the quantity of copper in the sample. The polarity of the terminals may be determined by bringing the wires, about 0.5 cm. apart, into contact with a piece of filter paper moistened with potassium iodide solution. At the positive terminal iodine will separate and color the paper. Cleaning the Platinum Electrodes. 1 The electrodes are freed from grease by heating with dilute sodium hydroxide solution, after which they are washed with water; the cathode is then dried, allowed to cool in a desiccator, and weighed. To clean the platinum cathode after the determination, cover the deposit completely with 6-normal nitric acid, heat for at least 15 minutes, and wash. A. Procedure with Stationary Electrodes. 2 Dissolve a 0.5- o.6-g. sample of copper sulphate, CuS04 . 5 H^O, in 50 cc. of water, in a tall 150-0:. beaker; stir to complete solution, add 4 cc. of nitric acid (sp. gr., 1.42), and dilute to 100 cc. Im- merse the electrodes in the solution and connect them in such a way that the electrode with the larger surface is made the cathode. The electrolysis, which should be carried out at a potential of 1.9-2.0 volts, may be completed in the cold over night, or in two or three hours if the temperature is kept at 70-80 by means of a heated sheet of wire gauze placed a short distance below the beaker. Finally test for complete deposition by adding a little water to raise the level of the solution on the 1 In case a silver cathode is used, see Note 12. 2 If the sample is an ore, see Note n. 88 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS cathode; if after 30 minutes no copper is to be seen upon the fresh platinum surface, the deposition is probably complete. (Test a few cubic centimeters of the solution with sodium acetate and a drop of potassium ferrocyanide solution.) Without dis- connecting the electrodes, siphon off the electrolyte while in- troducing distilled water, until the current ceases to pass ; this is to prevent the re-solution of any of the copper by the acid liquid. Remove the cathode, wash it with water, then with alcohol, and dry it for a short time in an air bath at 85-90. Allow the cathode to cool in a desiccator, and weigh. B. Procedure with a Rotating Anode. 1 Heat a five-cent coin with sodium hydroxide solution to free it from grease, then wash it with water, and dry at 100. After cooling, weigh the coin, and dissolve it in 50 cc. of 6-normal nitric acid, in a covered casserole. Evaporate the solution to dryness on the steam bath, dissolve the residue in about 100 cc. of cold water, add 10 cc. of sulphuric acid (sp. gr., 1.84), allow to cool, and transfer the whole to a 5oo-cc. measuring flask, diluting to the mark with water. Measure off one tenth of the well-mixed solution into a 5o-cc. graduated flask and transfer this quantitatively to the electrolytic vessel; before transferring the solution to the elec- trolytic vessel, however, pour 100 cc. of water into the latter and adjust the electrodes so that, when four fifths covered with water, they do not come into contact with one another, nor cause a loss of liquid, when the anode is rotated. The water can then be siphoned, or drawn off, and the solution transferred to the vessel and diluted to 100 cc. without disturbing the vessel or the electrodes. To perform the electrolysis, attach the anode to the shaft of the rotator (which is connected by means of a mercury cup, or otherwise, with the positive terminal), and the cathode to the negative terminal. Adjust the levels so that, with 100 cc. of solution, the cylindrical cathode is about four fifths im- mersed in the liquid. Then, by means of the sliding contact, 1 If the sample is an ore, see Note u. GRAVIMETRIC ANALYSIS 89 throw in the maximum resistance of the rheostat, see that all connections are well made, start the motor, and close the switch ; immediately decrease the resistance of the rheostat until the ammeter registers about i ampere, and allow the electrolysis to proceed. In the presence of nickel, the voltage should not ex- ceed 2.7. After about 55 minutes, test for complete deposition by adding a little water to raise the level of the solution on the cathode ; if after 10 minutes no copper is visible on the freshly exposed platinum, the deposition is complete. When this is the case, without disconnecting the terminals, stop the rotator and draw off the solution into a large beaker, carefully pouring in water as fast as the solution flows out. 1 As soon as the ammeter indicates that no current is passing, throw off the switch, remove the cathode and wash off the water with a little alcohol ; dry below 100, allow to cool in a desiccator, and weigh. NOTES. i. If two platinum plates, immersed in an aqueous solution of copper sulphate, are connected by wire with the poles of a storage battery, metallic copper will be deposited upon one of the plates; under certain conditions, all of the copper will separate in the form of a compact, firmly adherent metallic film. The process of decomposition is called electrolysis; the solution under- going decomposition is called an electrolyte; the two poles by which the current enters and leaves the electrolyte are called electrodes. When salt solutions are electrolyzed, the positive ions (cations) move towards the negative electrode (cathode), and the negative ions (anions) towards the positive electrode (anode). The quantity of electricity which passes through the solution in unit tune, or the speed of the current, is measured by an ammeter. The unit, 1 If it is desired to determine the nickel electrolytically, evaporate this dilute solution to a volume of 25-30 cc., make slightly alkaline with ammonia, filtering off any ferric hydroxide which may be precipitated, and to the solution (40 cc. in volume) in the electrolytic vessel add 60 cc. of ammonia of sp. gr. 0.90. Elec- trolyze at 3.0-3.5 volts with a rotating anode. After about an hour, test for com- plete deposition by adding to a few drops of the solution, neutralized with acetic acid, a drop or two of dimethyl-glyoxime solution (a red color indicates nickel). When the deposition is complete, proceed as directed in the copper determination. Finally remove the nickel from the cathode by heating for at least 15 minutes with 6-normal nitric acid. QO INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS called an ampere, is represented by the unvarying current which, when passed through a solution of silver nitrate, deposits metallic silver at the rate of 0.001118 g. per second. 1 The electromotive force, i.e. the electrical pressure which drives the current along the circuit, is measured by the voltmeter. The unit, called a wit, is represented by the electrical pressure that produces a current of one ampere when steadily applied to a conductor whose resistance is one ohm. The unit of resistance, called an ohm, is represented by the resistance offered to an unvarying electric current by a column qf mercury 14.4521 g. in mass, of a constant sectional area and a length of 106.3 cm -> at the tem- perature of melting ice. These magnitudes are always related to one another as follows (Ohm's law): Quantity of electricity (amperes) = Electromotive force (volts) Qr . = Resistance (ohms) R The most satisfactory current producer for electro-analysis, in which a steady non-fluctuating current is desired, is the secondary or storage ele- ment. The E. M. F. of the lead cell is about 2 volts, and the necessary voltage for the work may be obtained by connecting several cells in series. In practice, the potential difference between the electrodes is regulated by means of incandescent lamps, coils of wire, or other devices, which offer resistance to the flow of electricity along the circuit and convert electrical energy into heat. Any rheostat will do for this work, provided the range in resistance is properly related to the other factors which determine the current strength. The sliding-contact coil resistances which are on the market are very satisfactory. Voltmeters and ammeters should have the scales graduated with a range as limited as is consistent with the current conditions to be employed, so that each subdivision may represent a small fraction of a unit. The manner of connecting the instruments is illus- trated in the accompanying figure. 2. The passage of an electric current of suitable voltage through the solution of an ionogen is associated with physical and chemical changes which often may be utilized in exact gravimetric analysis. 1 The quantity of a given metal deposited by a current of electricity is directly proportional to the quantity of electricity which passes through the solution; and the quantities of different metals deposited by a specific quantity of electricity are directly proportional to the chemical equivalents of the metals in the solution. These two statements are known as Faraday's laws, though these apply to non- metallic ions as well. GRAVIMETRIC ANALYSIS 9 1 The chemical effect at the cathode is always some form of reduction. Simple metallic ions, as those of copper, tin, nickel, cobalt, cadmium, etc., travel towards the cathode, where they give up their charges and separate Electrodes in the metallic condition; while the hydrogen ion here loses its positive charge and either acts directly as a reducing agent (e.g. nitric acid to am- monia) or is evolved as gaseous hydrogen. At the anode, on the other hand, the chemical effect is always some form of oxidation. The anions of the halogen group are liberated as free chlorine, bromine, or iodine and may act as oxidizing agents, while from solutions containing hydroxide, sulphate, or nitrate ions, oxygen separates at the anode and either acts directly as an oxidizing agent or is evolved in gaseous form. (It should be borne in mind in this connection that aqueous solutions always contain the ions of water.) Although positive ions always move towards the cathode, certain metals (e.g. lead, cobalt, nickel, and a few others) may, under specific conditions, be oxidized (possibly to complex oxy-anions) and deposited more or less completely at the anode in the form of insoluble peroxides. In fact, lead can be determined accurately in this way, as an oxide. 3. Metals, like all other substances, possess when immersed in water a characteristic solution tension, by which is understood an expansive force which seeks to drive particles of the metal into the solution ; when a metal is immersed in the solution of one of its salts it will either send more of its atoms into the solution as ions, or some of its ions will be discharged from the solution on its surface as atoms. In the first case the metal will be- come negatively charged, and in the second case positively charged with respect to the solution ; in either case equilibrium will be reached when the solution tension of the metal is exactly counterbalanced by the electrostatic charges and the osmotic pressure of the metallic ions in the solution. Upon comparing the different elements from the standpoint of the potential difference between them and their salt solutions, at identical normal ion-concentrations, a characteristic series of values is obtained. 92 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS In the case of solutions of normal ion concentration, for example, some of the values are as follows : Zn= +0.493 Ag= 1.05 Cd= +0.143 I = -0.80 Fe= +0.067 Br= 1.27 Co= 0.045 O= 1.50 (At N. H + -ion concentration.) Ni= 0.049 Cl= 1.63 H= -0.277 S0 4 =-2.i8 Cu=- 0.606 4. In the electrolysis of a salt solution, both electrodes soon become coated with the products deposited (i.e. each becomes essentially an elec- trode of the deposited material, no matter what its original composition) and they are said to be polarized. Hence a system similar to that just dis- cussed may be considered to exist at each electrode, and it is evident that electrolysis must act in opposition to the solution tension of the elements and in conjunction with the osmotic pressure of their ions. In order then to decompose a salt solution continuously, a voltage at least slightly in excess of the polarization voltage must be applied; i.e. a voltage greater than the numerical difference between the single potential differences normally established at the cathode and anode. Assuming normal ion- concentrations, the decomposition voltage of copper chloride, for example, is 1.63 minus 0.606=1.02 volts; but the decomposition voltage of a solution as calculated in this manner, especially in the case of a salt of an oxyacid, frequently fails to agree with that found by experiment. The separation of gases at the electrodes is often accompanied by "overvoltages," which vary more or less markedly with the material and physical nature of the electrodes ; moreover, the ion-concentration is generally unknown, and it always changes as the electrolysis proceeds. The important matter here lies not so much in the calculation of decomposition voltages as in the recognition of the existence of a minimum decomposition voltage for every ionogen, under definite conditions. In the case of an ionogen of known decomposition voltage, we should simply use a somewhat higher voltage, but if other metallic ions were present it might be impossible to completely deposit one metal without using a voltage that would start the deposition of the second metal also. While copper can readily be separated from cobalt or from nickel by elec- trolysis, it is not possible to separate nickel from cobalt in this way ; and in general only metals whose deposition voltages differ by several tenths of a unit can be separated from each other by maintaining an intermediate voltage during the electrolysis. GRAVIMETRIC ANALYSIS 93 The addition of certain reagents, as ammonia, potassium cyanide, am- monium oxalate, etc., to solutions containing two metals sometimes reduces the concentration of one metallic ion very much more than that of the other, owing to the formation of more or less stable complexes, and makes it possible to perform a separation by the "constant voltage" method that otherwise might not be possible. 5. The current strength will of course depend upon the voltage used, since, according to Ohm's law, i = . In performing an electrolysis, the R voltage actually available is diminished by the decomposition voltage of the electrolyte (polarization voltage) ; hence the current which passes is equal to the available voltage minus the decomposition voltage of the electrolyte, divided by the resistance of the circuit. 6. The quantity of metal deposited in a given time is dependent upon the strength of the current in amperes. A current of i ampere is capable of depositing 1.118 mg. of silver, and, according to Faraday's law, equiva- lent amounts of other elements, per second. This law might be used to calculate the time necessary for the complete deposition of the metal if, under the analytical conditions, it were the only cation taking part hi the electrolysis. In the neighborhood of the cathode, however, the concentra- tion of the solution with respect to this cation gradually decreases to an infinitesimal value, and the resistance and the decomposition voltage of the solution therefore rise ; finally a point is reached at which other ions begin to be discharged. Since circulation of the solution tends to maintain a uniform distribution of the ions, mechanical stirring favors the rapid deposi- tion of those ions which have the lowest discharge voltages. 7. Unless the solution is mechanically stirred, the rate of deposition of a given metal decreases rapidly, owing to the decreasing concentration of its ions around the cathode and to the continually increasing proportion of the current which is carried by the hydrogen (or other) ions. Since a rapid circulation of the solution tends greatly to prevent the local decrease in metallic ion concentration around the cathode, and since with improved circulation currents of much higher density may be used than would other- wise give satisfactory deposits, it is possible to greatly reduce the time neces- sary for a determination by performing the electrolysis with the use of a rotating electrode. 8. Owing to the reduced viscosity at higher temperatures, the resistance offered by an aqueous solution to the passage of electricity decreases with a rise in temperature, and in this way the voltage required to produce a given current may be reduced to a minimum; this may sometimes be of importance in electrolytic separations. Moreover, in case stationary 94 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS electrodes are used, heating the solution during electrolysis gives rise to more or less rapid convection currents, and also increases the speed of diffusion, and these effects are equivalent to a gentle mechanical stirring. The solution should never be heated to the boiling point, however, since the deposit might in that case be loosened from the cathode. 9. The deposited metal tends to redissolve in the electrolyte (cf. "polar- ization"), and consequently the rate at which the metal is deposited must exceed that at which it redissolves. The metal is only deposited from the solution in immediate contact with the cathode, so that the greater the area of the cathode, the more metal there is available for deposition ; but also the greater the rate of re-solution. Hence it follows that the current strength necessary for the satisfactory deposition of the metal is propor- tional to the area of the cathode. The current strength per unit area is called the current density; a square decimeter is generally taken as the unit area. Hence a "normal current density of 2 amperes " means a current of 2 amperes per 100 sq. cm. of cathode area, or of i ampere for 50 sq. cm. of cathode area, etc. While the tendency of a metal to redissolve fixes a lower limit for the current density to be used, a higher limit is set by the tendency of the metal to form spongy, non- adherent films when deposited too rapidly. It is highly important for accurate work to deposit the metal in the form of a compact film which can easily be washed and weighed without loss. The condition of the deposit depends not only upon the current density used, but also upon the concentration of the metallic ions in the solution, the amount of free acid and other substances present, the tem- perature, etc. The best conditions for specific cases have been determined by repeated experiments. 10. Concerning the effect upon the nature of the deposit of the products that accumulate in the solution during an electrolysis, that are purposely added to the solution, or which were originally present in the sample, it may be stated that a very marked influence is often exerted by certain acids, bases, and other substances. A solution of copper sulphate, if electrolyzed without the addition of another substance, is almost sure to give a reddish brown, non-adherent deposit of spongy copper ; the addi- tion of a little sulphuric acid gives rise to a much more compact deposit, while the addition of nitric acid leads to a still better deposit of bright red firmly adherent metal. A small quantity of urea, in addition to either acid, appears to favor still more the formation of a satisfactory deposit. On the other hand, high current densities, which cause a rapid discharge of hydro- gen, are apt to yield loosely adherent deposits of spongy metal. A current density which gives a bright red, coherent deposit of pure copper when no GRAVIMETRIC ANALYSIS 95 interfering substance is present, will often give a very dark, loosely adherent deposit when arsenic is present, even in small amount. Such impurities must be removed before the electrolysis is begun. 11. If the sample to be analyzed by this method is a copper ore, and is not known to be free from arsenic and other interfering substances, it should be subjected to special treatment, in order to obtain a solution suitable for electrolysis. In most cases, a satisfactory solution may be prepared accord- ing to the procedure detailed under the volumetric estimation of copper (which see) ; the ore is evaporated with aqua regia, the residue extracted with dilute hydrochloric acid and water, and the copper, arsenic, etc., pre- cipitated with sodium thiosulphate ; the arsenic is then driven off by igni- tion, the residue evaporated to dryness with nitric acid, and finally taken up in 4 cc. of nitric acid (sp. gr., 1.42) and 50 cc. of water. This solution is diluted to 100 cc. and electrolyzed. 12. The electrode material should preferably be insoluble in the electro- lyte, with or without current action, and for that reason platinum is most often used for electrodes ; but the continued advance in the price of this metal has led to a search for less expensive materials. Other metals, as silver and copper, are in some cases suitable for use as cathodes in the deposition of metals (as is also the more expensive palladium-gold alloy which is in the mar- ket), but the anode must still be made of platinum or of something equally resistant. In the determination of copper, for example, a silver cathode is about as satisfactory as one of platinum ; the deposit can be removed by means of dilute hydrochloric acid, with the addition of a little hydrogen peroxide or nitric acid, and, after washing with ammonia, the cathode is again ready for use. Since a practical limit is placed upon the current density, the time neces- sary for a deposition is inversely proportional to the area of the cathode ; for this reason the electrode to receive the deposit should present the maxi- mum of surface to the solution. The platinum dish electrode designed by Classen is quite thin and presents a relatively large surface ; a Classen dish weighing 40 g. has a capacity of about 250 cc. and presents an inner surface of about 150 sq. cm. to the solu- tion. A platinum disk or a flat spiral of platinum wire may be used as the anode. The electrodes commonly used, however, are more economical. They are open cylinders of thin foil or of fine mesh gauze, and elongated spirals of heavy platinum wire ; the cylinders, which are used to receive the deposit, weigh 10-12 g. and the wire spirals about 8 g. Of all cathodes, those of gauze are the most efficient ; they present a relatively larger sur- face, all parts of the surface are equally effective, they permit a much better circulation of the solution and consequently the use of higher current den- g6 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS sities. These cylindrical cathodes and spiral anodes are the ones assumed in the foregoing procedures. 13. In case the electrolysis is to be performed with stationary electrodes, it is best to use a tall beaker of small diameter, which can be heated. If heating is not desired, however, or if a rotating electrode is used, the most suitable vessel is a 150-0:. glass cylinder, with a rounded bottom ending in an outlet tube provided with a stopcock ; the electrodes should reach nearly to the bottom of this cylinder, to insure efficient mixing. After the deposition is complete, without interrupting the current, the electrolyte can easily be drawn off with the simultaneous introduction of distilled water above. PART III VOLUMETRIC ANALYSIS GENERAL DISCUSSION Fundamental Principles. It has already been pointed out in Part I that in volumetric analysis the amount of an element or compound present in a sample is calculated from the volume of some reagent of known concentration which is required, after suitable treatment of the sample, to complete a definite reaction. The analytical balance is equally requisite as a starting point for both gravimetric and volumetric systems; in addition to the balance, volumetric processes demand graduated measur- ing instruments and standard solutions (i.e. solutions of ac- curately known value). The concentration or value of a solu- tion for a specific reaction is determined by a procedure called standardization, in which the solution is brought into reaction with a definite weight of a substance of known purity ; from the volume of solution required to complete the reaction, the strength or value of the solution can be calculated, and it is then a stand- ard solution. The value of standard solutions may be expressed in terms of the weight of reagent actually present in each cubic centimeter, or, better, in terms of the weight of a given substance with which one cubic centimeter of the solution will react; but since the weight of reagent present in a unit- volume is always chemically equivalent to the weight of substance with which the unit- volume reacts, it is in general more convenient to express the value of the standard solution in terms of chemical equivalents per unit- volume. Such solutions are often made to bear some simple H 97 98 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS relation to a normal solution of the specific reagent; they are, for example, half-normal, tenth-normal, or fiftieth-normal solu- tions. A normal solution contains in one liter one gram-equiva- lent of the active reagent; i.e. that quantity of the active re- agent which contains, unites with, replaces, or in any way, di- rectly or indirectly, brings into reaction 1.008 g. of hydrogen. Thus a liter of normal acid solution will contain 1.008 g. of available hydrogen ion (e.g. one mol of HC1, or one half mol of H 2 S0 4 , etc.) ; and a liter of normal alkali solution will contain sufficient available hydroxide ion to combine with 1.008 g. of hydrogen ion, or 17.008 g. (e.g. one mol of NaOH, one half mol of Ba(OH) 2 , etc.). A normal solution of an oxidizing agent will have the same oxidizing value per liter as one gram-equiva- lent, 8.00 g., of oxygen (e.g. one sixth mol of K^C^Oj, etc.) ; and a liter of normal reducing agent will have the same reducing value as i .008 g. of hydrogen (e. g. one half mol of SnCy . It will be seen that a liter of normal acid solution will exactly neutralize a liter of normal alkali solution, and a liter of normal oxidizing solution will exactly oxidize a liter of normal reducing solution, and so on. It should be especially noted, however, that the equivalent or normal weight of a substance may vary according to the reaction in which it is used. Thus the normal weight of oxalic acid is one half its molecular weight, whether it be used as a neutraliz- ing, a reducing, or a precipitating agent; whereas the normal weight of nitrous acid would be the molecular weight, if used either as a neutralizing agent or to oxidize hydriodic acid, but only one half the molecular weight if used to reduce potassium permanganate. In the case of potassium permanganate, two molecules yield three atoms of available oxygen (equivalent to six hydrogen atoms) in neutral solution, and five atoms of avail- able oxygen (equivalent to ten atoms of hydrogen) when used in acid solution. The normal weight of this compound as an oxidizing agent is therefore one third or one fifth of the molec- ular weight, according to the conditions of its use. VOLUMETRIC ANALYSIS 99 The preparation of exactly normal, half-normal, or tenth- normal solutions generally requires considerable time and care, and is usually carried out only when a large number of analyses are to be made, or when the analyst has some other specific pur- pose in view. It is much easier to prepare standard solutions which differ but slightly from half-normal or tenth-normal, and these still have the advantage of approximate equality; two approximately half-normal solutions are much more con- venient to work with than two which are widely different in strength. When these approximate solutions are used, the volumes can readily be reduced to the corresponding values in terms of solutions which are exactly normal, half-normal, or tenth-normal. For example, 25.75 cc - f a 0.0987 N solution are equivalent to 25.75X0.0987 = 2.542 cc. of the normal, or to 2.542 X 10 = 25.42 cc. of the tenth-normal solution. Reactions Suitable for Volumetric Processes. Volumetric processes are usually based upon definite chemical reactions, and in general only such reactions are suitable as can be made to take place completely and very rapidly when equivalent amounts of the reacting substances are brought together. Vol- umetric determinations, however, do not always consist in the direct titration of the substances under investigation. In many cases an excess of the standard solution is used, and this excess is then titrated with a second standard solution. The volumetric relation between the two standard solutions being known, the proper correction for the excess of the first solution is easily made. This is known as the method of back titration. In other cases the substance under investigation is capable under suitable conditions of setting free or of carrying down as a precipitate a definite proportion of some other substance which can subse- quently be titrated with a standard solution ; from the volume of the latter required it is an easy matter to calculate directly the weight of the original substance. The processes of volumetric analysis are readily classified, according to their character, into : 100 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS A. Neutralization Methods; such as those of acidimetry and alkalimetry. B. Methods of Oxidation and Reduction; as exemplified in the determination of ferrous iron by oxidation with potassium permanganate. C. Precipitation Methods; as, for example, the titration of silver with potassium thiocyanate solution. Determination of the End-point. In order to utilize a reaction for volumetric purposes, it is necessary to have some means of ascertaining the point at which an equivalent volume of the standard solution has been added. In the neighborhood of this point certain physical and chemical properties of the solu- tion change very rapidly; in many cases there is a marked change in color, electrical conductivity, or oxidation potential of the solution, and in some a precipitate just ceases to form. In numerous cases the presence of another reagent, called an indicator, gives rise to a decided color change in the solution, or in a few instances it causes a precipitate to form and thus renders the solution turbid. The point at which the standard solution has been added in sufficient quantity to make these changes apparent is called the end-point of the titration. If the process is to be sufficiently accurate, the difference between this point and the true end-point of the reaction (i.e. the point at which an equivalent quantity of the solution has been added) must be exceedingly small; and this is actually the case in nearly all of the established volumetric processes. In a few cases, however, a principle which can be used to great advantage is that of compensating errors ; here the errors which are involved in the actual determination are counteracted by equal errors in the standardization of the solution used. Assum- ing that there is a noticeable discrepancy between the true and the observed end-point, it will generally hold good that this discrepancy will remain constant so long as the conditions are the same. If the solution can be standardized under conditions identical with those which obtain in the actual determination, VOLUMETRIC ANALYSIS'' ioi' all errors can be practically eliminated; the standard solution becomes merely an instrument for comparing two solutions of the same substance, one representing a known amount of the standard solution and the other an approximately equal but unknown amount. Considered from this point of view, when- ever it is possible to use the same reagent for the determination of different substances, strict accuracy would demand that the solution be standardized by comparison with a known weight of the substance for which it is to be used in a given case. General Theory of Indicators. Whenever a substance is titrated in the presence of an indicator, the physical change which enables us to recognize the end-point is the visible result of a chemical reaction in which the indicator itself is an active reagent. In case the indicator is acted upon by the standard solution employed in making the titration, the reactions involved may be represented by the following equations, in which X represents the substance being titrated, R the reagent in the standard solution, and I the indicator employed : (1) X+R^RX, (2) I+R3ZIR, (3) IR+X^RX+L The appearance of the end-point is here dependent upon the concentration of IR, 1 which should remain equal to zero as long as an appreciable amount of X is present, but should increase in direct proportion to the amount of R that is added after the concentration of X has been reduced to an infinitesimal value. In other words, reaction (i) must be completed before reaction (2) begins to take place, but reaction (2) must take place promptly from left to right, even at an exceedingly low concentration of 1 IR is not necessarily a compound of the indicator with R, but even so a definite concentration of the free reagent R must finally be present in the solution in order to produce the visible change which is characteristic of the indicator. In such cases reaction (2) may be written /i^/ 2 , and the appearance of the end-point is dependent upon the concentration of 7 2 , which in turn depends upon that of R in the solution. COURSE IN QUANTITATIVE ANALYSIS Ry also, the least possible concentration of IR should cause a marked change in the appearance of the solution. It is further necessary that the small amounts of IR which are formed locally during the titration, in consequence of imperfect mixing, should react with X according to equation (3), thus preventing the appearance of false end-points. This will of course insure the completion of the reaction of X with R before the indicator begins to be permanently influenced by R. It will be seen, therefore, that the closeness of agreement between the observed and the true end-point in any specific case will depend upon the relative magnitudes of the three equilibrium constants con- cerned, as well as upon the experimental conditions, e.g. the concentration of the indicator, the temperature, etc. If, on the other hand, the indicator added reacts with the substance undergoing titration, the reactions concerned must take place essentially according to the following equations : (a) (6) (c) IX+R^RX+I. The appearance of the end-point is here dependent upon the final completion of reaction (c) from left to right. The accuracy of the process in any given case will depend here also upon the relative magnitudes of the three equilibrium constants, and upon the experimental conditions. In a few cases, in which IX will not react with R, it is necessary to use a special pro- cedure ; as the end-point is approached, the solution under- going titration is frequently tested, one drop at a time on a white test-plate, with a drop of the indicator solution. A factor which is usually of great importance in volumetric analysis is the concentration of the indicator in the solution. In those cases in which the end-points obtained depend upon a change from one specific color to another, the two colors may tend to mask one another and give rise to a series of indeter- minate transition tints. It is then desirable that the entire VOLUMETRIC ANALYSIS 103 amount of indicator present should be promptly transformed by the slightest possible excess of the titrating solution ; in such cases only a very slight amount of the indicator should be used. 1 If, however, the solution containing the indicator is colorless, and the addition of an excess of the standard solution produces a specific color, a relatively larger amount of indicator is less likely to be harmful ; in some cases, in which IR is a substance subject to dissociation, it is necessary, in order by mass action to insure its prompt formation, to add the indicator in consider- able amount. (Cf . the use of starch in iodometric methods, and also the use of ferric alum in the titration of silver with thio- cyanate solutions.) The Advantages of the Volumetric System. Volumetric determinations can usually be carried out' much more rapidly and conveniently than the corresponding gravimetric processes. The actual titration requires a few minutes only, but the neces- sity of removing interfering substances and of transforming the substance itself into a form suitable for titration often increases the time of the analysis to several hours. In many cases vol- umetric processes are more accurate than the corresponding gravimetric processes, but in other cases the reverse is true. Volumetric processes often avoid the errors which are involved in making gravimetric determinations; that is, the errors re- sulting from solubility, from the contamination of precipitates, and from actual mechanical losses. On the other hand they necessarily involve certain errors in the preparation and measure- ment of the standard solutions (see, however, what is said in Part I concerning the use of weight burettes) and in the deter- mination of the end-points of the reactions which are utilized. General Directions. For successful volumetric work it is essential that uniformity of practice prevail throughout with 1 In the case of methyl orange, for example, the indicator is prepared by dis- solving 0.02-0.03 g. of the solid compound in 100 cc. of water, and in any one titration only about three drops of this solution should be used. Counting 20 drops to i cc., and using 3 drops in a titration, the amount of methyl orange actually present is less than 0.05 mg. 104 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS respect to all matters which can influence the accuracy of meas- urement of liquids. For example, whatever time is allowed for drainage in the calibration of a measuring vessel, should also be allowed whenever the vessel is used ; and parallel conditions should be insured during both standardization and analysis, with respect to the quantity of the indicator and the final volume of the reaction mixture after titration. In some cases the stand- ard of the solution will vary appreciably with variation of the experimental conditions. It is of course necessary that standard solutions should be protected from concentration or dilution; they should be kept stoppered and away from direct sunlight or heat. The bottles should be shaken before use to collect any water that may have evaporated from the solution and condensed on the sides, and, after use, before replacing the stoppers, the necks and stoppers of the bottles should be wiped dry with a clean, lintless towel. Needless to say, the measuring vessels must be clean and free from grease, and great care must be taken to thoroughly rinse out all burettes and pipettes with the standard solutions they are to contain, in order to remove all traces of water or other liquid, which would act as a diluent. It is best to rinse them three times with small portions of the solution, allowing each portion to run out through the tip, before assuming them to be in a condition to be filled and used. Much time may be saved by estimating, if possible, before beginning the operation, the approximate volume of standard solution which will be required for the titration. This makes it possible to run in rather rapidly almost the required amount, after which, of course, the end-point must be very carefully determined. In case such a calculation cannot be made, it is often worth while to ascertain this approximate volume by means of a very rapid preliminary titration : this of course will necessitate the use of an extra sample. (See, for example, the determination of manganese by titration with potassium per- manganate.) VOLUMETRIC ANALYSIS 105 A. NEUTRALIZATION METHODS ALKALIMETRY AND ACIDIMETRY Standard solutions of acid and alkali are required for these processes, together with suitable indicators. Standard Acid Solutions. These are generally prepared from hydrochloric or sulphuric acid. Hydrochloric acid has the advantage of forming soluble compounds with the alkali earths, but its solutions cannot be boiled without danger of loss. Both acids may be used with all indicators. Standard Alkali Solutions. These may be prepared from sodium or potassium hydroxide, sodium carbonate, or barium hydroxide. Sodium and potassium hydroxides, if free from carbonate, may be used with all indicators, but they absorb carbon dioxide readily and attack the glass of bottles; sodium carbonate may be weighed directly for the preparation of stand- ard solutions, provided its purity is assured, but with many indicators the liberation of carbonic acid is a disadvantage. Barium hydroxide solutions are free from carbon dioxide; if any of this gas is absorbed, it at once causes the formation of a precipitate. Barium hydroxide may be used with all indicators, but it is not freely soluble in water. In many cases in which it is desirable to have a more concentrated carbonate-free solution of alkali, a sodium hydroxide solution from which the carbonate has been removed by means of a slight excess of barium chloride is very serviceable. Carbonate-free solutions should be pro- tected by means of a soda-lime absorption tube. Half-normal or tenth-normal solutions are employed in most analyses, the latter strength being convenient when small amounts of acid or alkali are to be determined. Indicators for Use in Alkalimetry and Acidimetry. The in- dicators used in these processes are organic substances, often of very complicated structure, each being capable of existence in two forms of different color, which are mutually transform- able into one another at specific H + - and OH~-ion concentrations. io6 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS An aqueous solution which is neutral contains H + and OH~ ions in equal concentration, and, so long as the solution is very dilute, the number of mols of water ionized in one liter is o.ooooooi, or io~ 7 , at 25. Since in the ionization equilibrium, H 2 OH + +OH-, we must have (H+) (OH-) = fc, it follows that in dilute solutions (H+)(OR-) = io~ 14 . If an acid is added to water, or to a dilute neutral solution, in sufficient amount to increase the H + -ion concentration from io~ 7 to io~ 6 , then the OH~-ion concentration will fall to io~ 8 . Only 0.0000009 mol of H + ion would be required to produce such a change in a liter of neutral solution, and this amount is contained in about o.oi cc. of tenth-normal hydrochloric acid, while the addition of only o.i cc. to one liter would increase the H + -ion concentration from io~ 7 to io~ 5 . The color changes which are characteristic of these indicators do not as a rule occur in exactly neutral solutions. The accom- panying table gives for some of the more common indicators, the H + -ion concentrations at which the color changes occur. It might seem, at first thought, that only an indicator which changes exactly at the neutral point would be suitable, but this is by no means true. In titrating a strong acid against a strong base a few hundredths of a cubic centimeter of tenth-normal acid or alkali in excess will carry the concentration of hydrogen or hydroxide ion so far to one side of the neutral point that any of the indicators for which the characteristic point lies between io~ 5 and io~ 9 will give a sharp and accurate end-point. In titrating a weak acid with a strong base, e.g. acetic acid with sodium hydroxide, the acetate ions from the highly ionized sodium acetate drive the reaction, HC2H302<^H + +C2H302~, far to the left, long before an equivalent quantity of the base has been added; the concentration of hydrogen ion becomes exceedingly low (io~ 7 ) before the major portion of the acetic acid has had its hydrogen replaced. In such a case the change in color for methyl orange (see the Table) will appear gradually, during the addition of a cubic centimeter or more of VOLUMETRIC ANALYSIS 107 3 * l-a 25 o> h .a ?l io8 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the alkali, long before an equivalent quantity of alkali has been run in, and no sharp end-point will be indicated. If phenol- phthalein is used, however, the change in color will not occur until the true neutral point, (H+) = io~ 7 , has been passed ; a very slight excess of alkali will then reduce the concentration of the hydrogen ion far below io~ 8 , a sharp end-point will be indicated, and the total alkali added will correspond very exactly to that required for the actual neutralization of the acid. When a weak base is titrated with a strong acid, e.g. ammonium hydroxide with hydrochloric acid, the conditions are reversed and such indicators as methyl orange, methyl red, or cochineal, which change color in a faintly acid solution (i.e. H + >io~ 7 ) are most suitable. The following rules should be observed in the use of neutraliza- tion methods for the determination of acids and bases : 1. In the titration of a strong acid with a strong base, or vice versa, use any indicator in the list, from methyl orange to phenol- phthalein. 2. In the titration of any acid other than a strong mineral acid with a strong base, use phenolphthalein, trinitrobenzene, or a similar indicator. 3. In the titration of a weak base with a strong acid, use methyl orange, Congo red, or a similar indicator. 4. Do not attempt to titrate a weak base against a weak acid. The sensitivity of a given indicator may vary under widely differing conditions of temperature and dilution, and for that reason it is important to titrate approximately equal volumes of solution in standardization and in analysis; and when it is necessary, as is often the case, to titrate the solution at the boiling temperature, the standardization should take place under the same conditions. It is also obvious that since some acid or alkali is required to produce the change in the indi- cator itself, the amount of indicator used should be uniform and not excessive ; usually three or four drops of the solution are ample. VOLUMETRIC ANALYSIS 109 Methyl orange solution is most readily prepared by dissolv- ing 0.02-0.05 g- f tne s lid compound (also known as Orange III) in a very little alcohol and diluting with water to 100 cc. It may be successfully used for the titration of strong acids and bases, and is particularly useful in the determination of weak bases, such as ammonium hydroxide and certain weak organic bases. It can also be used in titrating with a strong acid the soluble salts of very weak acids, such as carbonates, sulphides, arsenites, borates, and silicates, because in such cases the acids which are liberated are too weak to affect the indicator, and the reddening of the solution does not take place until a very slight excess of the strong acid has been added. It should be used in cold, not too dilute solutions. Its sensitivity is less in the presence of large quantities of alkali salts, or of smaller quan- tities of certain other metallic salts. Phenolphthalein solution is prepared by dissolving one gram of the solid compound in 100 cc. of alcohol. This indicator is particularly valuable in the determination of weak acids, es- pecially of organic acids. It should not be used with ammonia or weaker bases. It is decolorized by carbonic acid, which must therefore be removed by heating when other substances are being determined ; unlike methyl orange, it is sensitive in boiling- hot solutions The volume of the solutions titrated should be approximately uniform in standardization and in analysis, and for the best results this volume should not in general exceed 125-150 cc. THE PREPARATION AND STANDARDIZATION OF APPROXI- MATELY HALF-NORMAL SOLUTIONS OF HYDROCHLORIC ACID AND SODIUM HYDROXIDE Procedure. Pour into a zoo-cc. measuring cylinder a volume of hydrochloric acid (sp. gr. 1.19, with 37.2% of HC1) sufficient to contain 36.5 g. of hydrogen chloride, transfer it quantitatively to a one-liter measuring flask, and dilute to the mark with dis- tilled water. Pour this solution without loss into a clean, well- no INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS drained 2 liter bottle, refill the measuring flask with distilled water, and pour this also into the bottle. Finally shake the solution thoroughly for a full minute in the stoppered bottle, to insure uniformity of concentration. Weigh out, upon a rough balance, about 42 g. of stick sodium hydroxide. Dissolve the alkali in water, in a large beaker, dilute this solution also to two liters, and shake for a full minute in the stoppered bottle. Fill two clean, grease-free burettes with the respective solu- tions, first thoroughly rinsing each burette three times with 10 cc. portions of the corresponding solution and allowing the wash liquid each time to run out through the tip. See that all air bubbles are expelled from the tips, note the exact position of the liquid in each burette, and record the readings in the note- book. Run out from one burette about 20 cc. of the acid into a beaker, and add three drops of methyl orange solution ; dilute the acid to about 80 cc., and run in alkali solution from the other burette, with stirring, until the color of the solution changes from pink to yellow. Wash down the sides of the beaker with a little distilled water, replace the beaker under the acid burette, and add acid to restore the pink; continue these alternations until the point is accurately fixed at which a single drop of either solution will produce a distinct change of color. Select as the end-point either the appearance of the faintest tinge of pink which can be recognized, or its disappearance, and always titrate to the same point. If the titration has occupied more than three minutes, the time required for draining, the readings of the burettes may be taken at once and entered in the notebook. Refill the burettes and repeat the titration. Correct the burette readings as indicated by the burette calibrations, and if necessary for temperature (see Part I), and obtain the ratio of the solutions as shown in the following example : cc. acid 21.53 r -j r n T = = 1.022 cc. of acid per i.ooo cc. of alkali, cc. alkali 21.07 VOLUMETRIC ANALYSIS in When this ratio has been satisfactorily established, the hydro- chloric acid solution is standardized as follows : (a) By Titration against Pure Sodium Carbonate. If a suit- able oven is available, dry the salt on a watch glass for an hour at 130-150 ; otherwise heat about 5 g. of pure sodium carbonate in a small porcelain dish, on a wire gauze over a small Bunsen flame, for one half hour, and then allow the salt to cool in a desiccator. Transfer the cold salt to a dry, well-stoppered weighing tube, and weigh out into 4oo-cc. beakers two portions of 0.5-0.6 g. each, 1 noting the exact weights ,in the notebook. Pour over the salt about 80 cc. of water, stir until dissolved, and add three drops of methyl orange solution. Fill the burettes with the acid and alkali solutions, note the initial readings of the burettes, and run in the acid, with stirring, until the solution assumes the faintest tinge of pink. Wash down the sides of the beaker with a little water, and if the solution loses its pink tint add acid, one drop at a time, with stirring, until the faint pinkish tinge just returns. (If too much acid is added, the excess can be determined by means of the alkali in the other burette, since the ratio of the alkali to the acid is known.) After three minutes, note the burette readings and enter them in the notebook. From the data recorded, calculate the normality factor of the acid and also that of the alkali. The standardization of the acid must be repeated until the duplicate values agree within two parts in one thousand, and the same of course applies to the determination of the ratio of the two solutions. (b) Gravimetrically with Silver Nitrate. Measure out accu- rately from a pipette 10.00 cc. of the acid into each of two 3OO-CC. beakers, and dilute in each case with 150 cc. of water. Precipitate the chlorine from these solutions with silver nitrate according to the procedure given in Part II, and filter the silver chloride off through Gooch crucibles, prepared and weighed as there indicated. Wash the precipitates with hot water until 1 The weights of samples in this book are based upon the use of 30-00. burettes. If 50-cc. burettes are used, it is better to take samples f as large. 112 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS free from soluble silver salts, dry at 120-130 to constant weight, and from the weight of silver chloride found in each case calcu- late the normality factor of the acid. The duplicate values should agree very closely with one another, and also with those previously found by titration against sodium carbonate. NOTES. i. Although silver chloride is insoluble, the normality factor of the hydrochloric acid may nevertheless be calculated directly from the weight of the precipitate obtained. For example, if 10.00 cc. of the acid were found to yield 0.7317 g. of AgCl, then (since an equal volume of normal HC1 would yield 1.4334 g. of AgCl) the normality factor of the acid is 0.7317/1.4334, or 0.5105. In the same way, the normality factor of the acid may be calculated in the case of Method (a). For example, if 0.5682 g. of Na 2 C0 3 require 21.00 cc. of acid, the normality factor is equal to 0.5682/1.1130=0.5105 (1.1130 g. is the amount of Na 2 C0 3 contained in 21.00 cc. of the normal solution). If it has previously been found, for example, that i.ooo cc. of alkali solution is equivalent to 1.022 cc. of the acid, then it follows that the nor- mality factor of the alkali is 0.5105X1.022 = 0.5217. 2. If it is desired to prepare solutions of exactly one half normal con- centration, slightly stronger solutions are first prepared, and, after stand- ardization, they are diluted with the calculated volume of water. For example, the 0.5105 N acid should be diluted according to the proportion, 0.5105 : 0.5000= x : icoo, and the 0.5217 N alkali according to the proportion, 0.5217: 0.5000=37 : 1000; i.e. one liter of each solution should have added to it 21.0 cc. and 43.2 cc. of water, respectively. The water is added from a burette. After dilution the solutions should be thoroughly shaken, and then restandardized. 3. Solutions should be thoroughly mixed to insure uniformity of con- centration before standardization. They should be allowed to attain the temperature of the room, and they should be shaken to take up any water which may have evaporated and later condensed on the inner walls of the bottles. Before replacing the stopper in a bottle, always dry it, as well as the neck of the bottle, with a clean, lintless towel. 4. The liquid is diluted to 100 cc. during standardization in order that the volume may be the same as that which will prevail during analysis. 5. The exact point at which the color changes should be chosen as the end-point; any deeper tint is unsatisfactory, since it is not possible to duplicate shades of color from day to day by memory. VOLUMETRIC ANALYSIS 113 6. The selection of the best compound to be used as a standard for acid solutions has been the subject of much controversy. In many works cal- cium carbonate has been recommended (see Part IV, Problem 54), and also sodium carbonate, which can now be purchased sufficiently pure for this purpose. The most reliable standard would seem to be sodium carbonate prepared from recrystallized sodium bicarbonate by heating the latter between 280 and 300. The bicarbonate is easily purified by crystallization, and between the temperatures named it is decomposed quantitatively according to the equation, 2 NaHCO 3 =Na 2 C03+H 2 0+C0 2 . 7. Instead of standardizing the acid solution, it is equally practicable to standardize the alkali solution, with the use of phenolphthalein, against any of the following pure acids : oxalic acid, H 2 C2O 4 . 2 H 2 O ; acid potas- sium oxalate, KHC 2 4 . H 2 ; potassium tetroxalate, KH 3 (C 2 4 ) 2 . 2 H 2 O ; potassium bitartrate, KHC 4 H 4 6 ; succinic acid, H 2 C 4 H 4 4 . The last two are probably the most suitable, since they are free from water of crystalliza- tion. It should be noted that the acid oxalate and the bitartrate contain one replaceable hydrogen atom each, while the tetroxalate contains three such atoms, and the oxalic and succinic acids two. 8. While it is permissible to standardize hydrochloric acid solutions (provided they are free from other chlorides) gravimetrically with silver nitrate, a solution of sulphuric acid should not be standardized in this way by precipitation as barium sulphate ; the results would be less reliable on account of the difficulty in obtaining large precipitates of barium sulphate which are free from contamination and which are not partially reduced on ignition. THE DETERMINATION OF THE TOTAL ALKALINE VALUE OF SODA ASH The sample may be one of commercial soda ash, or it may be an artificial mixture of sodium carbonate and sodium chloride. Procedure. Weigh out roughly on a watch glass 5 g. of the soda ash, and dry it for one hour at 110 ; allow it to cool in a desiccator. Now accurately weigh the sample on the watch glass, transfer it quantitatively to a 5oo-cc. beaker, washing off the glass with about 50 cc. of water ; dry and weigh the watch glass, and take the difference as the weight of the sample. Gently warm the sample with the 50 cc. of water and filter off any in- 114 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS soluble residue. Wash the filter at least five times with 20 cc. portions of warm water, and receive the filtrate and washings in a 250-00. measuring flask which has been freed from grease by means of cold cleaning solution. Cool the liquid to the room temperature, add distilled water until the lowest point of the meniscus is level with the mark on the neck of the flask, and thoroughly mix the solution by pouring it from the flask into a dry beaker and back into the flask two or three times. Measure off 25 cc. of the solution with a pipette (which should be previously rinsed out with small quantities of the solution) into a 300-cc. Erlenmeyer flask, allowing the pipette to drain for a sec- ond or two with the tip in contact with the inside wet surface of the flask (unless it was standardized otherwise). Dilute the solu- tion to about 80 cc., add three drops of methyl orange solution, and titrate with the standard acid, using the standard alkali to complete the titration as already described. From the volumes of acid and alkali used, corrected for temperature dif- ference and burette errors, calculate the percentage of alkali present, assuming the alkali to be wholly sodium carbonate. Measure out other portions of 25 cc. from the main solution, and repeat the titration until satisfactory checks are obtained. NOTES. i. Let us assume, for example, that 5.890 g. of soda ash were used in the preparation of 250.0 cc. of solution, and that 25.75 cc. of 0.5105 N acid and 4.13 cc. of 0.5217 N alkali were used in the titration of 25.00 cc. of this solution. Then it follows that 5.890 g. of the soda ash are equivalent to ioX(25. 75X0.5105 4.13X0.5217) = 109.90 cc. of N acid; and, since this volume of normal acid would neutralize an equal volume of normal alkali, therefore the 5.890 g. of soda ash contained 0.053 X 109.90= 5.8246 g., or 98.9% of Na 2 C0 3 . 2. Soda ash is crude sodium carbonate. When made by the Solvay process it is apt to contain also sodium chloride, sulphate, and either the bicarbonate or hydroxide ; if made by the Le Blanc process, which however has gone out of use, sodium sulphide, silicate, aluminate, and other impurities are likely to be present. Many of these contribute to the total alkaline value? but it is customary to calculate this value in terms of sodium carbonate alone. 3. In order to obtain uniform results, it is customary to dry the soda ash at 110 before analysis. Complete expulsion of the moisture would VOLUMETRIC ANALYSIS 115 require a very much higher temperature. At least 5 g. are taken, in order to secure a representative sample ; but since this is too much for convenient titration, an aliquot portion of the solution is measured off. 4. For other methods of analyzing soda ash, the student should refer to Part IV, Problems 60, 61, and 95. THE DETERMINATION OF THE NEUTRALIZATION VALUE OF AN ACID Procedure. Weigh out accurately into 3oo-cc. beakers two 0.6-0.7 g. portions of the unknown acid (oxalic acid, acid potas- sium oxalate, potassium tetroxalate, potassium bitartrate, succinic acid, or some similar compound), dissolve each sample in about 80 cc. of warm water, add two or three drops of phenol- phthalein solution, and run in half-normal alkali from a burette until the solution is pink. Add half-normal acid from the other burette until the pink color just disappears, and then exactly 0.30 cc. in excess. Heat the solution, and boil for three minutes. If the pink color reappears upon boiling, discharge it with acid, again add 0.30 cc. in excess, and repeat the boiling. Discharge the pink color if it again reappears, again adding 0.30 cc. in excess. Repeat this treatment until the pink color fails to return upon boiling for three minutes. Finally add alkali until the color just reappears, then a drop or two of acid in excess and boil for one minute. If no color appears during this time, complete the titration with alkali in the hot solution. From the corrected volume of alkali required to react with the acid solution, calculate the normality factor of a solution containing 10.00 g. of the un- known acid in one liter. The results should check within two parts in one thousand. NOTES. i. Although it is desirable to employ the same indicator throughout standardization and analysis, the difference resulting from the change of indicator is in this instance insignificant, and the student may neglect it. It should be remembered, however, that in order to obtain the greatest accuracy possible, a restandardization throughout would be essential. 2. Since commercial sodium hydroxide always contains some carbonate, and since phenolphthalein is sensitive to carbonic acid, the solution must i n6 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS be boiled to free it from this acid. Phenolphthalein does not show an alkaline reaction with cold dilute sodium bicarbonate solution ; hence cold, dilute solutions of sodium carbonate become colorless with this indicator as soon as the carbonate has been transformed into bicarbonate by the acid. Upon boiling, the bicarbonate is partially hydrolyzed according to the equation, Na + HCO 3 -+H+OH-5:Na + +C>H-+H2C03, the solution loses carbon dioxide, and the pink color returns. This must again be dis- charged, the solution boiled, and so on. 3. Hydrochloric acid is volatilized from aqueous solutions upon boiling, unless they are very dilute. If the excess of acid added is not more than 0.30 cc., however, no loss need be feared. 4. When a large number of acidimetric determinations are to be made with phenolphthalein as the indicator, it is well worth while to prepare and standardize a carbonate-free alkali solution. The acid solutions are in such cases boiled, to free them from carbonic acid, or they are made up with freshly boiled water, and titrated hot with carbonate-free alkali (cf. the general discussion under alkalimetry and acidimetry). THE DETERMINATION OF PROTEIN NITROGEN BY THE KJELDAHL METHOD Principle. When an organic substance is heated with con- centrated sulphuric acid, especially in the presence of an oxygen carrier, the organic substance is completely decomposed, and any protein (or other similarly combined) nitrogen is converted into ammonia. This at once combines with acid to form am- monium acid sulphate, NH 4 HS0 4 , which remains in solution in the sulphuric acid. Upon diluting the mixture with water and adding sodium hydroxide in excess, the ammonia is liberated, and can be distilled over and collected in a known volume of standard acid, which it partially neutralizes. By titrating the excess of acid with a standard alkali, the volume of the standard acid neutralized by the ammonia can be found, and from the data obtained the percentage of nitrogen in the sample may be calculated. Procedure. Accurately weigh out from a weighing tube, upon separate sheets of quantitative filter paper, two samples of about i g. each of the substance to be analyzed. Wrap each VOLUMETRIC ANALYSIS 117 sample carefully in the paper, and introduce the bundle into a clean 500 cc. Kjeldahl flask. To each flask add about 0.5 g. of powdered copper sulphate, and 25 cc. of concentrated sulphuric acid. See that the samples are thoroughly wet by the acid, and then place the flasks on the digestion rack in the Nitrogen Lab- oratory, with the necks. resting in the circular openings of the lead ventilating pipe; place the flasks in unoccupied positions as near as possible to one of the exhaust flues. Heat gently until frothing ceases, add 10 g. (weighed roughly) of potassium sulphate, or an equivalent weight of sodium sulphate, and heat to gentle ebullition for two or three hours until the liquid is of a clear green color, without any trace of brown (do not allow the flame to reach above the surface of the liquid). Con- tinue the heating for half an hour longer, and allow to cool. While the flasks are cooling, accurately measure from a burette two 30.00-cc. portions of 0.5 N. hydrochloric acid, into 4oo-cc. Erlenmeyer flasks, and add to each about 25 cc. of distilled water. Place these flasks under the distilling apparatus, so that the delivery tubes just dip into the acid solutions. After cooling, carefully dilute the contents of the digestion flasks with 150 cc. of distilled water, and cool again. Carefully pour down the inclined neck of each flask, so that it shall not mix with the acid solution, 75 cc. of sodium hydroxide solution (300 g. of NaOH per liter). Place the flasks on the distilling rack, add one or two pieces of granulated zinc, and quickly connect the flasks with the distilling heads, using well-fitting rubber stoppers. Finally, mix the contents of each flask by gently rotating it, and then begin to heat the mixture. Distill off about two thirds of the contents of each flask, with great care that they do not boil over. The distillation will re- quire about 45 minutes. Disconnect the distilling flasks and rinse out the delivery tubes into the receiving flasks with a little distilled water. Add 3 drops of methyl orange to each of the receiving flasks, and titrate the contents with 0.5 N. sodium, hydroxide. n8 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS From the data obtained, calculate the percentage of nitrogen in the sample. NOTES. i. The sulphuric acid hydrolyzes the NH 2 -group, to give ammonia, and also acts as an oxidizing agent, converting the organic matter into carbon dioxide, water, or other volatile products. For example, CO(NH 2 ) 2 +H 2 0+2 H 2 S0 4 =C0 2 +2 NH 4 HS0 4 ; C 6 H 10 5 +w H 2 S0 4 =6 C +5 H 2 O+w H 2 S0 4 ; and C+2 H 2 S0 4 =CO 2 +2 H 2 0+2 S0 2 . 2. The CuS0 4 gives up oxygen more readily to the organic matter than the H 2 S0 4 does ; but the H 2 S0 4 then reoxidizes the copper so that at the end of the operation the copper is still present as copper sulphate. That is to say, the copper salt acts catalytically as an oxygen carrier. 3. Mercuric sulphate is often used instead of copper sulphate as an oxygen carrier, a few small globules of metallic mercury being added to the acid digestion mixture. Although the mercury salt is somewhat more efficient, it tenaciously retains ammonia, as H2N Hg S0 2 Hg NH 2 , even in the presence of an excess of hot alkali, and it is therefore necessary to add also a large excess of sodium sulphide. This converts all the mercury into HgS, which combines with Na 2 S to form soluble Hg(SNa) 2 , and the ammonia is liberated. 4. The K 2 SO 4 forms with the acid KHSO 4 , and this serves to raise the boiling point of the sulphuric acid ; the higher temperature hastens the di- gestion. 5. The flask is provided with a long neck in order that the acid fumes, which would otherwise be lost, may condense and run back into the digestion mixture. 6. After the acid solution has been diluted, it is specifically lighter than the NaOH solution used ; upon pouring the latter carefully down the neck of the inclined flask, it sinks to the bottom and leaves the surface of the liquid still acid, thus preventing the loss of ammonia at this stage of the procedure. The contents should not be mixed until after the flask has been tightly connected with the distilling head. 7. The granulated zinc is added in order to prevent bumping during the distillation ; the zinc dissolves slowly in the alkaline solution, with the evolution of hydrogen. Fragments of pumice stone or of platinum are often used for the same purpose. 8. If nitrates are present in the sample (e.g. a fertilizer), and it is desired to determine the total nitrogen, the procedure may be modified as follows : Thoroughly wet the sample in the flask with 25 cc. of concentrated sul- phuric acid, in which one gram of salicylic acid (CeH^^QQjj J has previ- VOLUMETRIC ANALYSIS 119 ously been dissolved ; this reacts with the nitric acid to form nitrosalicylic /N0 2 acid, CeHs OH . Next add slowly, with frequent shaking, 10 g. of \COOH powdered sodium thiosulphate, which reduces the nitrosalicylic acid to /NH 2 aminosalicylic acid, CeHs OH . Now add 0.5 g. of powdered copper \COOH sulphate and complete the determination as already described, but omitting the addition of the alkali sulphate (the solution contains NaHS0 4 from the thiosulphate). 9. It is evident that ammonium salts may be analyzed for ammonia by simply distilling them with an excess of alkali, absorbing the ammonia in an excess of standard acid, etc. Moreover, nitric acid and nitrates may be quantitatively reduced to ammonia and determined in this way (see Part IV, Problems 63, 64, and 89). Certain other salts, as acetates, may be analyzed in an analogous manner by distillation with phosphoric acid in excess, the distillate being collected in standard alkali and the excess of the latter titrated, with the use of phenol- phthalein. B. METHODS OF OXIDATION AND REDUCTION In most oxidation and reduction processes the standard solu- tion employed in the titration is one of an oxidizing agent, though it is often an advantage to have at hand a standard reducing solution as well. It may be stated, in general, that oxidizable substances are most often determined by direct titration, while oxidizing substances are very frequently determined by indirect titration (i.e. by adding the substance to an excess of reducing solution, and titrating either the excess of this solution or a product, as iodine, set free by the oxidizing substance). Standard Solutions. The most important oxidizing agents employed in the form of standard solutions are potassium per- manganate, potassium dichromate, iodine, potassium bromate, and ferric chloride; while the most important reducing agents are ferrous ammonium sulphate, oxalic acid, sodium thiosul- phate, arsenious acid, and titanous chloride. Other oxidizing and reducing agents are frequently used in the processes, but not usually in the form of standard solutions. 120 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS The most important combinations among the foregoing stand- ard solutions are: potassium permanganate and ferrous salts or oxalic acid ; iodine and sodium thiosulphate or arsenious acid ; potassium dichromate and ferrous salts. Indicators. With respect to the indicators employed, potas- sium permanganate, owing to its own intense coloring power, is its own indicator ; the slightest excess is indicated with great accuracy in otherwise colorless (or even in certain faintly colored) reaction mixtures. Since in the case of potassium dichromate no indicator has been found which is entirely satisfactory for use within the solution, potassium ferricyanide is employed as an outside indicator to determine the point at which the ferrous iron is completely oxidized. In the case of iodine, starch solu- tion is employed as an indicator. The use of these indicators will be discussed under the respective processes. I. DICHROMATE PROCESSES Fundamental Principles. In the presence of hydrochloric or sulphuric acid, ferrous salts are promptly and completely oxidized in the cold to ferric salts upon the addition of potassium dichromate solution. Since hydrochloric acid is by far the most suitable solvent for iron and its compounds, the titration is most often carried out in the presence of this acid : 6 FeCl 2 +K 2 Cr 2 O 7 +i4 HC1 = 6 FeCl 3 + 2 KC1+2 CrCl 3 + 7 H 2 0. As an indicator, potassium ferricyanide is used outside the solution to determine the end-point of the reaction. A drop of the iron solution is added to one of the indicator solution on a white surface, and the mixture examined for a blue coloration due to the formation of insoluble ferrous ferricyanide. The potassium ferricyanide must of course be free from ferrocyanide, and the indicator solution must be very dilute to diminish the interference of its own color ; a crystal the size of a pin- head dissolved in 25 cc. of water gives the right concentration. Since this solution is not stable, it must be freshly prepared each day. VOLUMETRIC ANALYSIS 121 THE PREPARATION AND STANDARDIZATION OF THE AP- PROXIMATELY TENTH-NORMAL DICHROMATE AND FERROUS IRON SOLUTIONS Procedure. Pulverize about 3 g. of potassium dichromate, dissolve 2.5 g. of the powder in water, and dilute to 500 cc. ; also dissolve 20 g. of ferrous ammonium sulphate and 5 g. of ammonium sulphate in water, with the addition of 5 cc. of con- centrated sulphuric acid, and dilute to 500 cc. Thoroughly mix the solutions, see that they are of the room temperature, and then fill a burette with each solution, observing the pre- cautions previously emphasized. Prepare a solution of potassium ferricyanide of the strength recommended above, and place single drops of this solution on the surface of a white porcelain tile. Run out from a burette into a 300-cc. beaker about 20 cc. of the ferrous solution, add 15-20 cc. of 6-normal hydrochloric acid, dilute to 150 cc., and run in about 18 cc. of the dichromate solution from a second burette. Test at this point by adding a small drop of the well-mixed iron solution from the end of a stirring rod to a drop of indicator on the tile. (The stirring rod which has touched the indicator should be washed off with distilled water before being returned to the iron solution.) If a blue pre- cipitate appears at once, 0.5 cc. of the dichromate solution may be added before another test is made. As soon as the blue appears to be less pronounced, add the dichromate solution in smaller amounts, finally a drop at a time, until the point is reached at which a bluish coloration fails to appear within 30 seconds after mixing a large drop of the iron solution with a drop of the indicator on the tile, the time being carefully noted. As soon as the 30 seconds have elapsed, remove another large drop of the iron solution and mix it with the indicator beside the last test; if no difference can be noted between the last mixture and this fresh one, the reaction is com- plete. Should the end-point accidentally be overstepped, more of the ferrous solution may be added and the titration pro- 122 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS ceeded with as before. Repeat this titration until good dupli- cates are obtained. From the volumes of the solutions used (after applying any necessary corrections) calculate the value of the ferrous solution in terms of the dichromate solution; the ratios found should not differ by more than two parts in one thousand. Standardize the dichromate solution as follows: Weigh out two portions of bright iron wire of about 0.15 g. each. The wire should be free from rust, and should be handled with filter paper. It should be bent so as not to interfere with the move- ment of the balance. Place 2O-cc. portions of 6-normal hydro- chloric acid in 3oo-cc. beakers, cover them with watch glasses, and heat to boiling. Remove the flames, drop in the portions of wire, and after the solution of the iron boil carefully for two or three minutes, keeping the beakers covered. Wash the sides of the beakers and the watch glasses with a very little water, and add stannous chloride solution to the hot liquid, drop by drop, until the mixture is colorless ; avoid more than a drop or two in excess. Allow to cool, dilute with 150 cc. of cold water, and add rapidly with stirring 25 cc. of mercuric chloride solu- tion. Allow the solutions to stand for two minutes, and titrate without further delay. (Calculate the volume of o.i .N dichro- mate solution which would be required by the sample of wire, and add almost this quantity before beginning to test with the indicator.) The ferrous solution may be used if the end-point is passed. From the volume of dichromate solution required to oxidize the known quantity of iron, calculate the normality factor of the solution. Repeat the standardization until duplicates are obtained which do not differ by more than two parts in a thou- sand, and from the mean of these calculate the normality factor of the ferrous ammonium sulphate solution. NOTES. i. The ionic changes which occur during oxidation and re- duction are more complicated than those of the methathetical reactions of precipitation and neutralization ; in the case of oxidation and reduction VOLUMETRIC ANALYSIS 123 an ion may change its entire character. The electrical charges on the new ion may differ in sign as well as in number from those on the original ion. In equations expressing these ionic changes the algebraic sum of the charges is always the same on the two sides ; in this case, for example, we have 6 Fe ++ +Cr 2 7 +14 H+=6 Fe +++ +2 Cr++++7 H 2 0. 2. It is possible to prepare an exactly tenth-normal solution of the dichromate by dissolving 2.4517 g. of the pure salt in water and accurately diluting the solution to 500 cc. The commercial salt, however, should not be used for this purpose ; it should be purified by recrystallization from hot water, and then dried at 130. 3. The presence of ammonium sulphate and sulphuric acid in the ferrous ammonium sulphate seems to increase the stability of the ferrous solution. 4. The iron wire offered in the market for this purpose answers well as a standard, and the iron content of each lot purchased may be ascertained by a number of gravimetric determinations. It may be preserved in a desiccator over concentrated sulphuric acid, but this must not be allowed to come in contact with the wire ; the wire should always be carefully ex- amined for rust before use. If necessary, it should be cleaned with fine emery paper. 5. The solution of the wire in hot acid and the short boiling insure the removal of gaseous hydrocarbons, due to the presence in the iron of a small amount of carbon. If not expelled, these might reduce some of the dichro- mate solution. Their complete expulsion is even more important when the wire is used as a standard in connection with potassium permanganate. 6. In the determination of iron by this method it must be wholly present in the ferrous condition. The common agents for the reduction of ferric iron are stannous chloride, zinc, sulphurous acid, and hydrogen sulphide; of these stannous chloride is the most convenient, but it should be used in very slight excess. To this end, it should be added to the hot concentrated ferric solution. The removal of the small excess of stannous chloride is necessary, and this is readily accomplished by means of a large excess of mercuric chloride; the chlorides of mercury do not react with the iron salts nor with the dichromate under the analytical conditions. The re- actions are, 2 FeCl 3 +SnCl 2 =2 FeCl 2 +SnCl 4 ; and SnCl 2 +2 HgCl 2 =SnCl 4 +Hg 2 Cl 2 . The mercurous chloride is precipitated. It is essential that stannous chloride should not be present in great excess and that the solution should be dilute and cold when the mercuric chloride is added; otherwise a secondary reaction is likely to take place with the reduction of mercurous chloride to metallic mercury (SnCl 2 +Hg 2 Q 2 =SnCl 4 + 2 Hg) , which would readily reduce the dichromate solution. The oc- currence of this secondary reaction is indicated by the darkening of the 124 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS precipitate; in such a case the solution is worthless, and should be dis- carded. If the ferric solution is hot and concentrated upon the addition of the stannous chloride, the reduction takes place with the greatest ease and only a drop or two of the stannous solution in excess need be added. 7. The solution should be allowed to stand for a minute or two after the addition of the mercuric chloride, to permit the complete precipitation of the calomel. It should then be promptly titrated to avoid any reoxidation of the iron by the air. 8. Less than 30 seconds are required for the appearance of the reaction with the indicator when the ferrous iron has nearly all been oxidized; if the mixture is left too long, the combined effect of light and dust will lead to a partial reduction of the ferricyanide, with the formation of a blue precipitate of ferric ferrocyanide. Thirty seconds is a sufficient interval. 9. The accuracy of the titration may be impaired by the removal of too much of the solution for the tests ; for that reason the tests should not be begun until most of the iron has been oxidized, but at the close of the titra- tion drops of considerable size may properly be taken (see note concerning this point under the gravimetric determination of iron). It is best never to overstep the end-point. The stirring rod should be washed each time in order not to transfer any of the indicator to the main solution. If the end-point is determined as prescribed, the dichromate method is capable of giving very exact results. THE DETERMINATION OF IRON IN SIDERITE Procedure. Weigh out two portions of about 0.23-0.25 g. of the finely powdered ore into 300-00. beakers, moisten the samples with water, cover the beakers, and add to each 20 cc. of 6-normal hydrochloric acid and about 0.2 g. of potassium chlorate. Heat at a temperature just below boiling until solvent action has ceased, and to the hot solution add stannous chloride solution, drop by drop, avoiding an excess greater than two drops. Add 150 cc. of cold water and 25 cc. of mercuric chloride solution, allow to stand for a minute, and proceed with the titra- tion as already described. Finally, calculate the percentage of iron in the ore. NOTES. i. Siderite is native ferrous carbonate; it may contain some organic matter, and, in order to destroy this, it is directed to add a little potassium chlorate to the hydrochloric acid. VOLUMETRIC ANALYSIS 125 2. Other ores can of course be analyzed by this method. Since most of them contain ferric iron, and since in the case of ferrous ores the iron is generally oxidized during the preparation of the solution, the quantity of stannous chloride required for the reduction of the iron will be much larger than that added to the solution of iron wire in the previous exercise. In no case, however, should stannous chloride solution be added in greater excess than two or three drops ; otherwise it is apt to cause the separation of metallic mercury, and spoil the determination. 3. For another method of dissolving iron ores, see the determination of iron by means of potassium permanganate, and also Note i under that method. THE DETERMINATION OF CHROMIUM IN CHROME IRON ORE Procedure. Weigh out two portions of about 0.25 g. each into iron crucibles which have been scoured inside until bright. Weigh out upon watch glasses, on the rough laboratory balance, two 4-g. portions of dry sodium peroxide, pour about three quarters of each upon the samples of ore, and mix the ore and flux by means of a dry glass rod. Remove any adhering par- ticles from the rod by stirring with it the remaining peroxide, and then pour the latter upon the surface of the mixture. (Ow- ing to the tendency of the peroxide to absorb moisture, the first portion should be mixed with one sample before the second portion is weighed out from the container.) Place the crucible upon a triangle and raise the temperature very slowly to the melting point of the flux, using a low flame and holding the burner in the hand. Maintain the fusion for five minutes, stirring with a stout iron wire; do not heat above moderate redness. Allow the crucible to cool until it can be held in the hand, and then cover it with water in a 3oo-cc. beaker, keeping the beaker covered with a watch glass. When the evolution of gas has ceased, rinse off and remove the crucible; heat the solution to boiling for 15 minutes in the covered beaker, add sufficient 6-normal sulphuric acid (calculated) to almost neutralize the liquid, and filter. To the filtrate and washings, which should 126 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS be slightly alkaline, add about 0.2 g. of sodium peroxide, boil for several minutes, acidify with 6-normal sulphuric acid, and add 10 cc. in excess. Weigh out for each solution about 0.5 g. more than enough pure ferrous ammonium sulphate to reduce the chromate, on the assumption that the ore was pure chromite, FeO . Cr 2 O3. Dissolve the salt, in a 5oo-cc. Erlenmeyer flask, in about 50 cc. of freshly boiled water and 20 cc. of 6-normal sulphuric acid, and transfer to this flask the chromate solution, diluting the whole to about 200 cc. Promptly titrate the excess of ferrous iron with the standard dichromate solution. From the data obtained, calculate the percentage of chromium in the ore. NOTES. i. Chrome iron ore consists essentially of ferrous chromite, Fe(Cr0 2 ) 2 . The ore is decomposed by the flux, which oxidizes the iron oxide to sodium ferrate and dissolves and oxidizes the chromic oxide to sodium chromate : 2 FeO . Cr 2 O3+ioNa 2 O2=4 Na 2 CrO 4 +2 Na 2 FeO 4 +4 Na 2 0. 2. Fused sodium peroxide attacks most materials; although it attacks iron and nickel, crucibles of these metals may nevertheless be used if care is taken to keep the temperature as low as possible. The peroxide must be dry, and no dust or organic matter of any kind should be present ; other- wise explosions may occur. 3. When iron crucibles are used, the fusion should be allowed to become cold before it is placed in water; otherwise magnetic oxide of iron, FeO . Fe 2 3 , is apt to scale off from the crucible. This will lead to no error, however, if the solution is only partially neutralized before filtration. Partial neutralization is to prevent the alkali from destroying the filter paper. 4. Upon treatment with water the chromate goes into solution, the sodium ferrate is decomposed into sodium hydroxide, ferric oxide, and oxygen, and the excess of sodium peroxide is decomposed with the evolution of oxygen. The subsequent boiling insures the complete decomposition of the peroxide, any of which if present would react with the chromate upon acidification. The alkaline chromate solution is always slightly reduced upon filtration through a paper filter ; it is therefore directed to add to the filtrate a small quantity of sodium peroxide to reoxidize the chromium, and to boil a second time to destroy the excess. VOLUMETRIC ANALYSIS 127 5. The addition of acid transforms the sodium chromate into dichromate, which, of course, behaves like potassium dichromate in acid solution. If any of the sodium peroxide is allowed to remain undecomposed in the solu- tion, the chromate is at least partially oxidized to a perchromate, upon acidification. 6. Instead of using Fe(NH 4 S04)2 . 6 H 2 O, the ferrous solution may be prepared from a suitable quantity of pure iron wire ; or, of course, a stand- ard ferrous ammonium sulphate solution itself may be added in excess. Perhaps an even better method for the determination of chromium consists in the addition of potassium iodide in excess to the acidified fusion extract, followed by the titration of the iodine with sodium thiosulphate solution : Cr 2 7 +6 I-+i 4 H+= 2 Cr +++ +7 H 2 0+3 I 2 ; and 2 Na 2 S 2 3 +l2= 2 NaI+Na 2 S 4 O 6 . 2. PERMANGANATE PROCESSES Fundamental Principles. In acid solution, potassium per- manganate promptly and completely oxidizes ferrous iron in the cold to ferric iron. Also, at 80-90, it reacts quantitatively with oxalic acid, which it oxidizes to carbonic acid. Though in reality the reactions are not so simple, the quantitative relation- ships are accurately represented by the following equations : 10 FeS0 4 +2 KMn0 4 +9 H 2 S0 4 = 5 Fe 2 (S0 4 ) 3 +2 KHS0 4 +2MnS0 4 +8H 2 0; and 5 H 2 C 2 O 4 -f 2 KMn0 4 +4 H 2 SO 4 = 2 KHSO 4 +2 MnS0 4 + ioC0 2 +8H 2 0. Or, more simply expressed, 5 Fe+++Mn0 4 -+8 H+ = 5 Fe++++Mn+++4 H 2 O ; and 5 C 2 O 4 + 2 Mn0 4 -+i6 H+ = 2 Mn+++io CO 2 +8 H 2 O. In a hot neutral or faintly acid solution, in the presence of zinc salts, potassium permanganate oxidizes manganous salts quantitatively in the sense of the equation, 3 Mn+++2 Mn0 4 ~-f 2 H 2 = 4 H++5 Mn0 2 . From these equations it is readily seen that for use in acid solution the normal weight of the salt is one fifth of a mol, or 31.61 g., while for use in neutral solution the normal weight is one third of a mol, or 52.68 g. 128 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS In addition to the above, potassium permanganate is capable of oxidizing stannous, cuprous, and mercurous salts, antimonious, arsenious, nitrous, and sulphurous acids, hydrogen sulphide, ferrocyanides, and many other substances. Furthermore, as a less desirable feature, the permanganate is capable under certain conditions of oxidizing free hydrochloric acid, with the liberation of chlorine; the action is rapid in hot or strongly acid solutions, especially in the presence of ferrous iron, but slow in cold dilute solutions. It is possible, however, with suitable modifications, to obtain very exact results in the presence of hydrochloric acid, even in the titration of iron ; but, other things being equal, in acid solution, it is preferable to carry out permanganate titrations in the absence of chlorides. Potassium permanganate has an intense coloring power. Even the tenth-normal solution is so deeply colored that the lower line of the meniscus is not visible in an ordinary burette ; readings must therefore be made from the upper edge. More- over, the slightest excess added to an otherwise colorless solu- tion is indicated with great accuracy; as its own indicator, it renders the titration one of the most satisfactory known. The permanganate solution should not be placed in burettes with rubber tips ; it is more or less rapidly reduced by most organic substances. THE PREPARATION AND STANDARDIZATION OF AN AP- PROXIMATELY TENTH-NORMAL SOLUTION OF POTAS- SIUM PERMANGANATE Procedure. Dissolve 3.25 g. of the permanganate crystals in 200 cc. of warm water, dilute the solution to one liter, and mix thoroughly. The value of this solution is apt to change slowly, especially just after it has been prepared. For this reason the solution should be allowed to stand for several days, and then filtered through a layer of asbestos to remove the precipitate of hydrated manganese dioxide. After thorough mixing, it is then ready for standardization. The solution should be pre- VOLUMETRIC ANALYSIS 129 served in glass-stoppered bottles, and should be protected from heat and light. Thais prepared and preserved, it will retain its oxidizing value for months. The solution is said to be still more stable if it is made very slightly alkaline with potassium hy- droxide (before standardization, of course). Weigh out accurately into 700 cc. Erlenmeyer flasks several 0.12-0.14 g. samples of pure sodium oxalate, previously dried at 110-120; dissolve each sample in 250 cc. of hot water (80- 90), with the addition of 30 cc. of 6-normal sulphuric acid, and titrate at once with the permanganate solution. At first, the permanganate should be added drop by drop, with shaking after each addition until the color disappears. After several drops have been added, the solution may be run in slowly (10-12 cc. per minute) with continuous shaking. Toward the end of the titration, particular care must be taken to allow the color due to each drop to disappear before the addition of the next, in order to avoid passing the end-point. Titrate to the first per- manent pink. The temperature at the end of the titration must not be below 60. From the data obtained, calculate the normality factor of the solution. Duplicate values should check within two parts in one thousand. NOTES. i. It is not satisfactory to prepare a standard solution by directly weighing out the calculated quantity of potassium permanganate, even after the latter has been purified by recrystallization. The best practice is to prepare the solution as described in the procedure, and then to standardize it by comparison with iron wire or with sodium oxalate. Ferrous ammonium sulphate, oxalic acid, potassium tetroxalate, acid potas- sium oxalate, and other substances have been proposed as standards, but iron wire and sodium oxalate are readily obtainable in a sufficiently pure condition, and being non-hygroscopic and free from water of crystallization, their composition is less subject to change. 2. Upon treating a given weight of pure sodium oxalate with an excess of sulphuric acid, the corresponding weight of oxalic acid is set free; so that the use of this salt as a standard merely enables us easily to measure out a specific amount of oxalic acid. The oxidation of the oxalic acid by the permanganate is at first slow, and the permanganate should be added 130 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS dropwise, with full time for decolorization between successive drops. After a certain small amount of manganous sulphate has been produced in the solution, however, the speed of the reaction is very greatly increased (by the catalytic action of this substance) and the permanganate may be run in much faster. THE ^DETERMINATION OF IRON : IN HEMATITE Principles. One of the most accurate methods for the deter- mination of iron is based upon the oxidation of a chloride-free ferrous sulphate solution, in the presence of sulphuric acid, with potassium permanganate. Under these conditions, ferrous iron is oxidized and permanganate is reduced, according to the equation : MnO 4 -+5 Fe+++8 H+ = Mn+++5 Fe++++4 H 2 0. But if chlorides are present, some of the permanganate will be reduced by these, with the liberation (and partial escape) of chlorine, and the results will be somewhat high : 2 MnO 4 -+i6 H++IO Cl- = 2 Mn+++8 H 2 O + 5 C1 2 . Upon the addition of the permanganate to a cold, dilute solu- tion of hydrochloric acid alone, or to one containing ferric iron, no chlorine is evolved ; ferrous iron, therefore, seems to accel- erate this reaction by catalysis. Nevertheless, since in dissolving iron ores it is nearly always necessary to use strong hydrochloric acid, to which it is often well to add a little stannous chloride, and since stannous chloride is a most convenient reagent for the reduction of ferric iron to the ferrous condition, it is desirable, if possible, to carry out the titration in the presence of fairly large quantities of chlorides. Now it has been shown that if, 'when chlorides are present, a small quantity of manganous salt is added to the solution, the ferrous iron alone is oxidized, and that accurate titrations can be performed (Zimmermann) . But the end-point is somewhat indistinct, owing to the yellow tint of the ferric chloride pro- duced. This difficulty can be overcome by the addition of phosphoric and sulphuric acids (Reinhardt), which have recently VOLUMETRIC ANALYSIS 131 been shown to combine with ferric iron to form colorless com- plexes such as H[Fe(S0 4 ) 2 ], H 3 [Fe(P0 4 ) 2 ], and H 6 [Fe(P0 4 ) 3 ] (Weinland and Ensgraber, Zeitschrift fur anorganische Chemie t Vol. 84, p. 349) ; the large excesses of these acids repress the dissociation of these complexes and insure a colorless solution. Procedure. Weigh out three samples of the finely ground ore, of about 0.25 g. each, into 100 cc. beakers. To each sample add 15 cc. of 6-normal hydrochloric acid and 2 cc. of stannous chloride solution, and gently heat the covered beakers for 10-15 minutes, until nothing other than a small, white, sandy residue remains undissolved. If the hot solution is at all yellow, dis- charge this color by adding stannous chloride solution, one drop at a time, with stirring ; avoid an excess of more than two drops. If, however, after the heating, the solution is colorless, stannous chloride is present in unknown excess, and must be oxidized by adding permanganate solution (not to be counted, of course, in the volume required for the titration) drop by drop with stirring, until the yellow color due to ferric iron appears; dis- charge this color as above directed, with stannous chloride solution, one drop in excess. After cooling, dilute the colorless solution with 50 cc. of cold water, and transfer, with stirring, to a yoo-cc. beaker containing 10 cc. of mercuric chloride and 50 cc. of water. (If, instead of a white precipitate of calomel, a gray precipitate of mercury is formed at this point, the solution must be discarded.) Dilute the mixture with cold water to about 500 cc., add 8-10 cc. of the Zimmermann-Reinhardt solution, 1 and titrate at once with the standard permanganate solution. Add the permanganate slowly, with constant stirring, finally in single drops, until the pink color flashes throughout the solution and persists for 15-20 seconds ; do not pass the end-point. Report the percentage of iron in the ore. 1 Made by dissolving 67 g. of MnSO 4 . 4 H 2 O in 500 cc. of water, adding 138 cc. of phosphoric acid (sp. gr., 1.7) and 130 cc. of sulphuric acid (sp. gr., 1.84), and diluting with water to one liter. 132 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS NOTES. i. Many iron ores are not completely decomposed by hydro- chloric acid, the insoluble residue containing more or less iron, as silicate, titaniferous iron, etc. Unless iron is known to be absent in the insoluble residue, the finely ground sample should be digested on the hot plate with 10 cc. of hydrochloric acid until the residue is white, or until there appears to be no further action; if the ore contains carbonaceous matter, a little potassium chlorate should be added. Finally evaporate to dryness, ex- tract with 5 cc. of hydrochloric acid, dilute with 10 cc. of water, allow to settle, and decant the clear liquid through a small filter, transferring the residue to the filter and washing with as little cold water as possible. Ignite the filter and residue in a small platinum crucible, allow to cool, and add 20-30 drops of sulphuric acid and twice as much hydrofluoric acid. Heat carefully, and, if the residue is dissolved, evaporate to white fumes, allow to cool, dissolve in water, and add to the solution at first obtained. If, however, this treatment fails to decompose the residue, drive off most of the sulphuric acid, add 0.5-0.6 g. of potassium bisulphate, and heat gradually until the bisulphate is quite liquid and fumes of sulphuric acid are given off whenever the lid of the crucible is raised. When the black specks have disappeared, allow the crucible to cool and dissolve the salt in the crucible with hot water and a few drops of hydrochloric acid. In case ferric iron has been dissolved in hydrochloric acid in contact with platinum, the solution should be oxidized with bromine water and the iron precipitated with ammonia ; i.e. if it is desired to use stannous chloride in the reduction. The ferric hydroxide can then be redissolved (after washing it with hot water) in hydrochloric acid and reduced. Otherwise the iron solution will contain a small quantity of platinum, 4 FeCls+2 HCl+Pt =4 FeCl 2 +H 2 PtCl 6 , which gives a characteristic ferric-iron color with stannous chloride, and prevents the recognition of the point at which the iron is reduced. 2. Three samples should be taken, in order that one may be used for a rapid preliminary titration. Having ascertained in a rough manner the iron content oi the sample, the final titrations are greatly facilitated. 3. Stannous chloride is a great help in the solution of many ores con- taining ferric iron. Apparently the difficultly soluble particles of hematite are continuously reduced at the surface to ferrous oxide, which is much more readily dissolved by the acid. 4. The available agents for the reduction of ferric iron are zinc, sulphurous acid, and hydrogen sulphide ; stannous chloride is excluded unless the titra- tion is to be made by the Zimmermann-Reinhardt method. In that case it should be carefully added, in very slight excess, to the hot, concentrated, acid solution (cf. the standardization of dichromate solution, Note 6). VOLUMETRIC ANALYSIS 133 5. Soluble salts of mercurous mercury are readily oxidized by potassium permanganate in acid solution. Mercurous chloride, however, is exceed- ingly insoluble, and, provided only a very small quantity is suspended in the solution, its action is so slow that the end-point of the titration can be accurately fixed. The pink color which flashes throughout the solution at the end of the titration is, however, not permanent, and for that reason the time-limit set should be closely observed. For the greatest accuracy, the permanganate should of course be standardized, under exactly the same conditions, against a known quantity of metallic iron. But the error due to the use of a solution standardized against sodium oxalate is for most purposes negligible. 6. For a rapid method for the reduction of ferric iron by means of zinc, see Notes i and 2 under the Determination of Phosphorus in Steel. It should be noted that titanium is also reduced by zinc, but not by the other agents mentioned ; with the use of zinc, therefore the presence of titanium would lead to high results. THE DETERMINATION OF CALCIUM IN LIMESTONE Procedure. Instead of igniting the precipitate of calcium oxalate, obtained from the limestone by double precipitation according to the procedure described in Part II, and weighing it as calcium oxide, the calcium may be determined volumetrically as follows : Wash the reprecipitated calcium oxalate by decanta- tion, keeping it as far as possible in the precipitation vessel, and decompose this precipitate by slowly pouring through the filter at least six 5-cc. portions of hot, 3-normal sulphuric acid, wash- ing afterwards with hot water, and receiving the acid filtrate and washings in the beaker containing the bulk of the precipitate. Dilute this mixture to 100 cc. and warm gently, with stirring, to completely decompose the calcium oxalate. Allow the mixture to cool, transfer it quantitatively to a 25o-cc. measuring flask, and dilute to the mark with water, finally mixing the solution by pouring it into a clean dry beaker and back into the flask. Measure out by means of a pipette 50.00 cc. portions of this solution, add to each 30 cc. of 6-normal sulphuric acid, dilute to 300 cc., heat to 90, and titrate as already described with the standard permanganate solution. Remembering that only one 134 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS fifth of the sample was used in each titration, calculate the per- centage of CaO in the limestone. NOTE. The reactions involved in the volumetric determination of calcium are: CaC 2 O4+H 2 S0 4 =CaS04+H 2 C 2 C>4; and 5 H 2 C 2 4 +2 KMnO 4 +4H 2 S04=2KHSO4+2MnS04+ioC0 2 +8H 2 0. It is therefore plain that the normal or equivalent weight of calcium oxide in this case is one half of a mol, or that o.i N permanganate solution has a calcium oxide value of 0.00280 g. per cubic centimeter. THE DETERMINATION OF THE OXIDIZING VALUE OF PYROLUSITE Procedure. Weigh out two portions of the very finely ground mineral, of about 0.3 g. each, into 500 cc. Erlenmeyer flasks. Calculate the weight of ferrous ammonium sulphate, Fe(NH 4 S0 4 )2 . 6 H 2 O, required to react with each sample, assum- ing it to be pure manganese dioxide (2 FeS0 4 +Mn0 2 +2 H 2 SO 4 = Fe 2 (S0 4 ) 3 +MnS0 4 -f-2 H 2 0), and weigh out accurately por- tions of the pure salt 0.15-0.20 g. in excess of the calculated amounts, into the corresponding flasks. Pour into each flask 50 cc. of water and 50 cc. of 6-normal sulphuric acid, cover the flasks, and heat to boiling until the action is complete. Finally, dilute to about 300 cc., and promptly titrate the excess of fer- rous iron with the standard permanganate solution. From the data obtained, calculate the percentage of Mn0 2 in the sample. NOTES. i. The mineral should be so finely ground that no grit what- ever can be detected when a little of the powder is placed between the teeth ; upon this the success of the analysis largely depends. If properly ground, solution will be complete in 10-15 minutes. 2. A moderate excess of ferrous iron is necessary to promote rapid solu- tion, and, also to facilitate solution, the mixture should not be diluted be- fore the solvent action has ceased. 3. A solution of iron wire in sulphuric acid may be substituted for the ferrous ammonium sulphate, but in that case there is more danger of the partial oxidation of the iron by the air. For example, if iron wire is used, it should be dissolved in sulphuric acid out of contact with air, and the air should not have access to the solution during cooling. This is best accomplished by means of a Contat-Gockel valve, which consists of a glass VOLUMETRIC ANALYSIS 135 bulb with an inner siphon, as shown in the figure. In the bulb is placed a cold saturated solution of sodium bicarbonate, through which the hydrogen (and steam) evolved in the flask bubbles. After all the iron has been dissolved, the liquid is boiled for a few minutes longer, and the flame is removed. As the flask cools off, small portions of the bicarbonate are at intervals sucked into the flask and decomposed by the acid with the evolution of carbon dioxide, whereby the entrance of more bicarbonate solution is prevented. For other methods of performing this analysis, see Part IV, Problems 23, 73 and 74. According to O. L. Barnebey (J. Ind. Eng. Chem., Vol. 9, p. 961 (1917)), the use of oxalic acid in place of the ferrous salt yields less reliable results. 4. With the substitution of very dilute nitric acid for sulphuric acid in the above procedure, the method may be used to determine the oxidizing power of red lead, or minium, Pb 3 4 , and of lead peroxide, Pb0 2 . Of these substances, samples of i.o and 0.8 g., respectively, should be taken when 3o-cc. burettes are used. It is better, however, to make use of an iodometric method. THE DETERMINATION OF PHOSPHORUS IN STEEL Principle. The molybdic anhydride contained in ammonium phosphomolybdate, (NH 4 )3P0 4 . 12 MoO 3 , may be reduced by zinc in the presence of sulphuric acid, from Mo0 3 to Mo 2 03; but molybdenum in the latter condition is not stable in the presence of air. If, however, the acidified molybdate solution is passed through a Jones reductor (see below) directly into a solution of ferric sulphate, the sensitive molybdic compound is oxidized by the ferric salt with the formation of an equivalent amount of ferrous sulphate, less sensitive to the atmospheric action. The molybdenum solution is green as it leaves the reductor, but upon mixing with the ferric salt the green color disappears; if phosphoric acid is added, the color due to the presence of ferric iron is destroyed. The decolorized solution is titrated while still hot with tenth-normal permanganate 136 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS solution, of which the quantity necessary corresponds to the equation, 5 Mo 2 O 3 +6 KMn0 4 = 3 K 2 0+6 MnO+io Mo0 3 . From this it may be seen that 5 P =c= 30 Mo 2 3 036 KMn0 4 ~ 90 0, or Poi8 0; one cubic centimeter of o.i N permanganate solu- tion represents, therefore, 0.0862 mg. of phosphorus. Procedure. Weigh out two samples of steel drillings, each sufficient to contain 1.7-2.0 mg. of phosphorus, into 250-0:. Erlenmeyer flasks. Add to each a mixture of 25 cc. of nitric acid (sp. gr., 1.42) and 75 cc. of water. Suspend in the neck of each flask a small funnel and heat until, after complete solu- tion, the oxides of nitrogen have been expelled. Dissolve 0.3- 0.4 g. of KMn0 4 crystals in 10 cc. of hot water, add one half of this solution to the contents of each flask, and boil until the permanganate color has disappeared. Remove the flame, add sulphurous acid or ammonium bisulphite solution, a few drops in excess, to dissolve the precipitated oxides of manganese, boil out the excess of sulphur dioxide, and filter the solution; re- ceiving the filtrate in a similar flask. Add ammonia to the solution with stirring until a permanent precipitate just begins to form, and then add nitric acid drop by drop to clear up the solution. Finally, at a temperature of 40, add 40 cc. of molyb- date solution, close the flask with a rubber stopper, and shake vigorously for five minutes; allow the precipitate to settle. (At this point, prepare the Jones reductor for use, as described in Note 2.) Now filter the solution, keeping the precipitate as far as pos- sible in the flask, and wash by decantation with a solution of ammonium sulphate acidified with sulphuric acid l until the washings give no test for molybdenum with ammonium sulphide and hydrochloric acid. Dissolve the precipitate by pouring through the filter a mixture of 5 cc. of 6-normal ammonia and 20 cc. of water, and collecting the filtrate and washings in the 1 Made by mixing 15 cc. of ammonia (sp. gr., 0.90) and 25 cc. of sulphuric acid (sp. gr., 1.84) with one liter of water. VOLUMETRIC ANALYSIS 137 precipitation flask. Acidify the solution, which should have a volume of about 60 cc., with 10 cc. of sulphuric acid (sp. gr., 1.84) and promptly pass the acidified solution, before it has a chance to cool off, through the reductor into the receiver (collect- ing the liquid beneath the surface of 100 cc. of a solution con- taining 25 g. of ferric alum and 40 cc. of sirupy phosphoric acid, sp. gr., 1.7, per liter), preceded by 100 cc. of hot water and followed by 200 cc. of hot dilute sulphuric acid (i : 40) and 100 cc. of hot water. See that no air enters the reductor during this entire operation. Titrate the reduced solution at once with tenth-normal permanganate, and calculate the percentage of phosphorus in the steel on the assumption that the yellow pre- cipitate contains phosphorus and molybdenum in the proportion indicated by the formula (NH^sPOi . 12 MoOa. 190mm. NOTES. i. The Jones reductor, which also is useful in the reduction of ferric iron for titration, is essentially a column of amalgamated zinc, through which the solution is passed for reduction. It is as- sembled as shown in the accom- panying figure. The tube A has an inside diameter of about 18 mm. and (for this reduction) is 400 mm. long ; the small extension tube has an inside diameter of 6 mm. and a length of 300 mm. below the stop- cock. At the base of the tube A are placed some glass beads ; these are covered by a plug of glass wool several millimeters thick, and upon this is placed a layer of asbestos, such as is used for Gooch filters, not exceeding i mm. in thickness. The tube is then filled with the amalgamated zinc to within 50 mm. of the top, and this is covered with a plug of glass wool. The reductor is connected as shown with the suction bottle F, and the bottle D is a safety vessel to prevent contamination of the solution from the suction apparatus. 138 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS The amalgamated zinc is prepared by dissolving 5-6 g. of mercuric chlo- ride in 250 cc. of water, with the addition of 5-10 cc. of dilute hydrochloric acid, adding to this solution 500 g. of (20-30 mesh) granulated zinc, in a large flask, and shaking thoroughly for two minutes ; the solution is then poured off and the zinc thoroughly washed with water. 2. To prepare the reductor for use, connect the safety bottle with the vacuum pump, fill the reductor while the stopcock is nearly closed with warm, dilute sulphuric acid (25 cc. of the concentrated acid in one liter), and then open the stopcock so that the acid runs through slowly. Continue to pour in acid until 200-300 cc. have passed through, then close the cock while some liquid is still left in the funnel. (During the whole operation, see that no air enters the reductor; if air enters, hydrogen peroxide will be formed from oxygen and nascent hydrogen, and the results will be worth- less.) Now remove the filtrate, and again pass through 200 cc. of the warm acid, followed by 100 cc. of warm water; test this liquid (300 cc.) with the standard permanganate solution, in order to determine the volume of permanganate required to color the acid alone. This amount must be subtracted from the volume required in the subsequent titration. 3. Upon dissolving the steel in nitric acid of the strength indicated, the phosphorus is oxidized, and none of it is lost by evolution as phosphine. The permanganate is subsequently added in order to insure the complete oxidation of carbonaceous matter and of the phosphorus to phosphoric acid. 4. The higher oxides of manganese, as Mn0 2 , are not soluble in nitric acid. Upon the addition of a reducing agent, however, such as hydrogen peroxide or sulphurous acid, their solution is effected : MnO 2 +H 2 S03=MnO+H 2 SO4~MnS0 4 +H 2 O. 5. In connection with the precipitation of phosphoric acid as ammonium phosphomolybdate, the student should consult the notes under the Deter- mination of Phosphoric Anhydride. 6. Since the molybdenum in the precipitate prepared from one gram of a steel containing 0.15% of phosphorus would require by this method 17.44 cc. of o.i -normal permanganate solution, it is readily seen that the process is a rapid one for arriving at very accurate results. This is es- pecially true if the permanganate has been standardized under the same conditions against a steel of accurately known phosphorus content; in such a case, it would be unnecessary to correct for the small amount of iron extracted from the (impure) amalgamated zinc, since this would be the same in both standardization and analysis. 7. It scarcely needs to be pointed out that the method is not suitable for determining phosphorus or phosphoric acid in substances containing them VOLUMETRIC ANALYSIS 139 in large amount. This would require for titration relatively enormous quantities of permanganate solution, and, what is still worse, it would be practically impossible to completely reduce the molybdenum. If small aliquot portions were taken for reduction and titration, any error of meas- urement would be multiplied by a very large factor in the calculation of the result. 8. The following method is suitable for the volumetric determination of phosphorus or phosphoric acid when these are present in larger amounts. The phosphorus or phosphoric acid is converted into ammonium phospho- molybdate ; this, after washing with KNO 3 solution, is dissolved in an excess of standard sodium hydroxide solution ; and the resulting solution is titrated with standard nitric acid, with phenolphthalein as an indicator. Needless to say, the sodium hydroxide should be standardized under identical condi- tions against a sample of accurately known phosphorus content. THE DETERMINATION OF MANGANESE IN AN ORE Fundamental Principles. When potassium permanganate is added to a hot, neutral or very faintly acid solution of manganese sulphate a reaction takes place (the Guyard reaction) in which the manganous oxide of the sulphate is oxidized at the expense of the anhydride of the permanganate, with the precipitation of hydrated intermediate oxides in varying proportions. These are manganous acid, MnO(OH) 2 , and hydrated salts of man- ganous acid. The essential changes in the state of oxidation may be represented as follows : Mn 2 7 +3 Mn O = 5 Mn0 2 ; Mn 2 7 +8 MnO = 5(Mn0 2 . MnO) ; Mn 2 7 +i3 MnO = 5(Mn0 2 . 2 MnO). It is clear, then, that in this form the Guyard reaction cannot furnish the basis for a satisfactory volumetric method. It has been found, however, that under suitable conditions, in the presence of zinc ion, a hydrated manganite of zinc is pre- cipitated, which, while variable in composition, contains all the manganese in the quadrivalent condition. Thus regulated, the reaction furnishes a valuable means for the determination of manganese (Volhard's Method). Although the composition of 140 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the precipitate varies, the course of the reaction is typically represented by the following equation : 4 KMn0 4 +5 ZnS0 4 +6 MnS0 4 +i4 H 2 = 2 K 2 S0 4 +9 H 2 S0 4 + 5 Zn(oMn/ H ), or, more simply, 2 MnO 4 ~+3 Mn+++2 H 2 0=4 H++5 Mn0 2 . Procedure. Weigh out into a 5oo-cc. Erlenmeyer flask a sufficient quantity of the very finely ground ore to contain about 0.20 g. of manganese; add 3 g. of potassium chlorate and 20 cc. of i2-normal hydrochloric acid, and boil until the ore is com- pletely decomposed and the chlorine expelled. Dilute with water to about 50 cc. ; transfer the cold solution quantitatively to a loo-cc. measuring flask; dilute to the mark with water; and mix thoroughly by pouring the contents of the flask into a clean, dry beaker, and back into the flask. Now, from a burette or pipette, measure into 500-0:. Erlen- meyer flasks four 20.00 cc. portions of this solution, and treat each as follows : Dilute with water to 100 cc., heat, and to the acid solution add with shaking an aqueous suspension of zinc oxide, 1 in small portions, until the iron is completely precipitated as ferric hydroxide ; this point may be recognized by the sudden coagulation of the precipitate, upon shaking, and the decoloriza- tion of the brownish colored solution. The precipitate should not be light yellow, but should have the characteristic brownish red color of ferric hydroxide, and the least possible excess of zinc oxide should be used. (Should the ore contain a quantity 1 Dissolve 100 g. of crystallized zinc sulphate in 300 cc. of hot water, and with stirring cautiously add to the clear solution a few drops of a solution made by dissolving 25-27 g. of pure sodium hydroxide in 150 cc. of water, until the zinc solution remains distinctly turbid ; then add a little bromine water, heat, and filter. To the nitrate add the bulk of the sodium hydroxide solution, and stir. Rinse the mixture into a one-liter bottle, and fill the latter with water. The mixture should be well shaken when used. (This suspension should not react alkaline with phenol- phthalein, and a 10 cc. portion of the mixture, when cleared up with sulphuric acid, diluted to 100 cc., and treated with one drop of o.i N KMnO 4 , should be perma- nently colored pink.) VOLUMETRIC ANALYSIS 141 of iron insufficiently in excess of that required by any phosphoric and arsenic acids present, then 5 cc. of a solution containing 20 g. of ferric chloride per liter should be added before the pre- cipitation with zinc oxide.) If too much zinc oxide is added, the solution will be milky ; in that case very dilute hydrochloric acid should be added drop by drop to the hot solution until the supernatant liquid just becomes clear. Finally dilute the solutions to 300 cc. and, at 80, treat them successively as follows : Run into the first solution the standard permanganate in 5 cc. portions, until after continued shaking the liquid retains a permanent pink tinge, say after the addi- tion of the fifth portion (i.e. 25 cc.) ; into the second solution run 5 cc. less permanganate than the volume previously used (e.g. 20 cc.), shake until the pink color disappears, and then finish the titration by the further addition of permanganate in portions of i cc. until the pink color persists after protracted shaking, say after 23.0 cc. in all have been added; to the third solution add at once i.o cc. less permanganate than the total volume used in the second case (e.g. 22.0 cc.), and continue the titration with the addition of 0.20 cc. portions, until the hot solution matches in color a solution prepared by the addition of o.io cc. of the permanganate to 300 cc. of water. With the fourth solution, repeat this titration. If, for example, 22.60 cc. of the perman- ganate have been used in each of the last two titrations, then this quantity minus the o.io cc. of the solution used for com- parison should be taken as the volume actually required. Report the percentage of manganese in the ore. NOTES. i. In case the treatment with hydrochloric acid and potas- sium chlorate should be insufficient to thoroughly decompose the ore (in- dicated by the presence of a dark-colored residue), the residue should be filtered off, washed, dried, and ignited hi a platinum crucible. It should then be fused with sodium carbonate, the melt dissolved in hydrochloric acid, and the solution evaporated to dryness in a porcelain dish, to dehy- drate the silica. The final residue should be moistened with hydrochloric acid, taken up in water, and filtered into the Erlenmeyer flask containing the acid filtrate from the original residue. The resulting solution, which 142 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS contains all the manganese, is then evaporated to a small volume, trans- ferred to the measuring flask, and treated as described in the procedure. 2. Upon the addition of zinc oxide to the acid solution of the ore, the zinc oxide first neutralizes the acid with the formation of zinc chloride, and then precipitates the iron with the further formation of zinc chloride, accord- ing to the reaction: 2 FeCl 3 +3 ZnO+3 H 2 O = 2 Fe(OH) 3 +3 ZnCl 2 . In this way sufficient zinc ion is introduced into the solution to insure the conversion of the manganese into the hydrated manganite of zinc. 3. Although it is often recommended to convert the chlorides in the solution into sulphates before the addition of zinc oxide, this treatment is not necessary. The titration of manganese in a dilute neutral solution with potassium permanganate is a very different thing from that of ferrous iron in a dilute acid solution containing chlorides. In the latter case, the ferrous iron catalyzes the reaction, 2 KMn04+i6 HC1=2 KC1+2 MnCl2 + 5 Cl 2 -|-8 H 2 0, some of the chlorine escapes, and there is a tendency to high results. In the former case, however, nothing is present in the solution to catalyze the reaction between the permanganate and the small quantity of hydrochloric acid which is formed ; and, although the solution is hot, its acid concentration is so low that there is no danger from this source. Start- ing with 0.2000 g. of an ore containing 20% of manganese, for example, the total quantity of acid formed in the titration (e.g 4 KMn0 4 + 5 ZnCl 2 +6 MnCl 2 +i4 H 2 0=4 KCl+i8 HCl+s ZnO . Mn 2 3 (OH) 2 ) weighs about TT/"M ^r Xo.04, or somewhat less than o.i g. ; and this quantity in a volume of over 300 cc. would give an acid strength of less than o.oi-normal. 4. The titration should be performed at 80-85, an d especial care should be taken not to heat the solution too hot during the titration. 5. For the greatest accuracy, in spite of all that has been said above, the permanganate solution should be standardized against a known quantity of manganese, weighed as MnS04, under conditions similar to those to be used in the analysis. 3. IODOMETRIC PROCESSES Fundamental Considerations. Analytical processes which depend upon the volumetric measurement of specific amounts of iodine are known as iodometric methods. In these processes, either iodine is used in standard solution to bring about a definite reaction, or the iodine liberated in a reaction is determined by titration with some suitable standardized reagent. VOLUMETRIC ANALYSIS 143 The titration of iodine against sodium thiosulphate, with starch as an indicator, is one of the most accurate of volumetric processes. The process may be used in neutral or slightly acid solutions to determine free iodine, and this in turn may serve as a measure of any substance capable of liberating iodine from hydriodic acid. For example, the quantity of potassium iodate in a sample of the salt may be determined on the basis of the reactions : KI0 3 +6 KI+6 HC1 = 6 KCl+KI+ 3 I 2 +3 H 2 0; and I 2 +2 Na 2 S 2 3 = 2 NaI+Na 2 S 4 6 . It should be noted that chlorine and bromine oxidize sodium thiosulphate partially to sulphate, while, under analytical con- ditions, iodine oxidizes it wholly to sodium tetrathionate. Iodine acts as an oxidizing agent either through the direct withdrawal of a positive constituent, as shown in the equations : 2Na 2 S 2 3 +I 2 = 2NaI+Na 2 S 4 6 , and H 2 S+I 2 = 2 HI+S, or through the decomposition of water in the presence of a reducing agent, as in the equations: H 2 S0 3 +I 2 +H 2 0$H 2 S0 4 +2 HI, and H 3 As0 3 +I 2 +H 2 = H 3 As0 4 +2 HI (H 2 O+I 2 :HI+HOI; and H 3 As0 3 +HOI^lH 3 As0 4 +HI). It will be seen from these equations that a one-tenth normal iodine solution contains one tenth of one gram-atom, or 12.692 g. of iodine per liter. The solubility of iodine in water is too small for the prepara- tion of even a one-tenth normal solution. In the presence of sufficient potassium iodide, however, the iodine dissolves much more readily, owing to the formation of an unstable but soluble polyiodide of the formula KI 3 : KI . I 2 , or I-+I 2 ^(I . I,)-. In the presence of reducing agents iodine is removed from this equilibrium mixture, the reaction runs to completion from right to left, and the solution can be used as though it were a simple solution of iodine. The potassium iodide used in the prepara- tion of the solution should weigh about 1.5 times as much as 144 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the iodine. Moreover, the presence of potassium iodide in the solution renders it possible to employ commercial iodine (which is apt to contain chlorine as an impurity) in the preparation of the standard solutions; the chlorine is removed according to the equation, IC1+KI = KC1+I 2 . In performing iodometric titrations in the presence of sul- phuric acid, particular attention should be given to the main- tenance of suitable analytical conditions. If, for example, it is desired to determine copper by titrating the iodine liberated in the reaction, 2 CuS0 4 +2 H 2 S0 4 +4 KI$Cu 2 I 2 +4 KHS0 4 +I 2 , it is not sufficient to simply add potassium iodide and sul- phuric acid in (unknown) excess. It must be remembered that such a mixture will contain both sulphuric and hydriodic acid, and that if the concentration of either is too great, or if the solution is allowed to become at all warm, the determination is very apt to be spoiled: H 2 S0 4 +2 HI = H 2 S0 3 +H 2 OH-I 2 ; H 2 S0 4 +6HI = S+4H 2 0-{-3l 2 ; or, in extreme cases, H 2 S0 4 +8HI = H 2 S+4H 2 0-|-4 1 2 . Other things being equal, when acid solutions are required, it is better to use acetic acid or dilute hydrochloric acid. In direct titrations with iodine, e.g. in the presence of sodium bicarbonate, it is best to work in the absence of ammonium salts. Such solutions are very faintly alkaline, especially if at all warm; and in the presence of ammonium salts ammonia is apt to be liberated, which is not entirely without influence upon the titration. Iodine solutions act upon rubber, so that burettes with glass stopcocks should be used. Determination of the End-point. A single drop of one-tenth normal iodine solution imparts a distinct tint to 200 cc. of water, and in many titrations with this solution no other indicator is required. If, however, the solution to be titrated contains colored substances, or if the greatest possible accuracy is demanded, a solution of starch should be used as an indicator. Under the proper conditions, the presence of one part of free iodine in VOLUMETRIC ANALYSIS 145 several millions of solution can be recognized with this indi- cator, but the sensitiveness of the reaction and the color pro- duced are affected by a number of factors. The test is decidedly more sentitive when the concentration of iodide ion (and of hydro- gen ion l ) is not too low, and when the quantity of starch present is sufficient to give a deep blue color. Under less favorable conditions, the starch may give a greenish or a reddish color ; or it may be very unreliable, as in solutions containing an abnormally low iodide ion concentration. How- ever, since the standard iodine solution always contains potas- sium iodide, and since an iodide is always one product of the titration, there is ordinarily not much danger from this source. Attention should be directed mainly toward the observance of uniform conditions in all related titrations : the volumes of the solutions titrated should be approximately equal, the starch solution should be properly prepared, and the same quantity of it should be added for each titration. Finally, all titrations should be made in the cold ; the iodo-starch blue is discharged by heat. Preparation of the Starch Solution. Rub i g. of potato starch with 5 cc. of cold water to a smooth paste, and slowly add this to 200 cc. of boiling water. Continue the boiling for about i min. until an almost clear solution is obtained, set this aside to settle, and finally decant the supernatant liquid through a filter. Use 5 cc. of the clear filtrate for each titration. A " soluble starch " which is in the market is more convenient, since with it filtration is unnecessary. A solution made by add- ing 200 cc. of boiling water to i.o g. of this starch, previously mixed with a little cold water, serves the purpose well. Use 5 cc. of this solution for each titration. In either case, the starch solution should be freshly prepared. If a great many titrations are to be made, however, it is ad- visable to prepare a liter of the starch solution; a number of 1 The iodo-starch blue is discharged by caustic alkalies and somewhat less readily by sodium or potassium carbonate, but not by the bicarbonates. L 146 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS small (50-100 cc.) bottles should be filled with this solution, heated for two hours in a water bath, and, while still in the bath, they should be closed with paraffined soft cork stoppers. Thus sterilized and protected from the air, the solution will retain its sensitiveness almost indefinitely. After a bottle has been opened, mould nearly always begins to form within a few days ; hence the use of small bottles. THE PREPARATION AND STANDARDIZATION OF APPROX- IMATELY ONE-TENTH NORMAL SOLUTIONS OF IODINE AND SODIUM THIOSULPHATE Procedure. Weigh out on the rough laboratory balance 6.3-6.4 g. of commercial iodine, add it to a solution of 9 g. of potassium iodide in 25 cc. of water, in an Erlenmeyer flask, and agitate the mixture until the iodine is completely dissolved. Dilute the solution to 500 cc., in a measuring flask, and mix it thoroughly. Heat 600-700 cc. of distilled water in a large flask and boil for about 5 minutes. Stopper the flask loosely, and allow the water to cool. Weigh out 1 2.5 g. of sodium thiosulphate, Na 2 S 2 3 . 5 H 2 O, introduce it into a 5oo-cc. measuring flask, and dissolve it in about 200 cc. of the cold, freshly boiled water. Finally dilute to the mark with more of the same water, and mix thoroughly. After these solutions have come to the room temperature, fill a burette with each (the iodine in a glass-stoppered burette), observing the usual precautions to prevent dilution. Run out 25 cc. of the thiosulphate solution into a beaker, dilute with 150 cc. of water, add 5 cc. of starch solution, and titrate with the iodine to the appearance of the blue of the iodo-starch. If the end-point is overstepped, titrate back with the thiosulphate solution. (All waste solutions containing iodine and potassium iodide should be poured into the vessel provided for iodine residues.) Repeat until the ratio of the two solutions is accurately estab- lished, taking into account all necessary corrections for burettes and for temperature changes. VOLUMETRIC ANALYSIS 147 Standardization of the Iodine Solution. Weigh out into 500-cc. beakers two o.i2-o.i3-g. portions of pure arsenious oxide, and in each case dissolve the arsenious oxide in 10 cc. of 6-normal sodium hydroxide solution. Dilute the solution to 100 cc., add 2 drops of methyl orange, and then cautiously add 6-normal hydrochloric acid until the solution contains 2 or 3 drops in excess. Cover the beakers, add to each a solution of 5 g. of pure sodium bicarbonate in 75 cc. of cold water, then add 5 cc. of starch solution, and titrate with the iodine to the ap- pearance of the blue color. Do not pass the end-point. From the corrected volume of the iodine solution used, calculate the normality factor of the solution. Duplicate values should agree within two parts in one thousand. Also, from the ratio pre- viously found, calculate the normality factor of the thiosulphate solution. NOTES. i. Iodine solutions are acted upon by sunlight with the formation of hydriodic acid, and a high room-temperature tends to volatilize the iodine. They require frequent standardization against pure arsenious acid, anhydrous sodium thiosulphate, or standard thiosulphate solution. 2. Sodium thiosulphate, Na 2 S 2 3 . 5 H 2 O, may be obtained pure by recrystallization. It is then possible to prepare a standard solution by dissolving the calculated weight of the salt in pure cold water, and dilution to the required volume. Such solutions are quite stable and may be kept for months without appreciable change in concentration, provided they are not allowed to absorb carbon dioxide ; but they should be protected from heat and light, both of which are likely to promote decomposition. 3. Carbonic acid causes a slow decomposition of the thiosulphate solu- tion, with the formation of free sulphur and sulphurous acid; and, since sulphurous acid acts in the same way, the decomposition once started be- comes progressive : xro Q n -LTT rn -> I Na 2 C0 3 + Na 2 S 2 3 +H 2 C0 3 ; | H2Sa03 Na 2 S0 3 + The reducing value of the solution increases gradually as the decomposition progresses ; i.e. the solution apparently becomes stronger. When it is con- sidered that in this decomposition each molecule of thiosulphate yields one 148 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS molecule of sulphite^the greater reducing value is readily understood ; for iNa&Qi+Iii-NftiSjOrHNal, while iNa 2 SO 8 +I 2 +H 2 O=Na 2 S0 4 -|-2HL 4. Solutions of the thiosulphate may be standardized against pure iodine, or, more conveniently (with the help of pure potassium iodide), against potassium bromate, potassium iodate, or potassium dichromate. These three salts are readily obtainable in a pure condition. 5. Arsenious oxide dissolves most readily in caustic alkalies, and for this reason the sodium hydroxide is used. The presence of sodium hydroxide is not admissible, however, during the titration, since it reacts readily with iodine. It is therefore removed by the addition of a slight excess of acid, and sodium bicarbonate is then added in large excess. The purpose of the bicarbonate, which under the analytical conditions is without action upon the iodine, is to neutralize the acid formed in the reversible reaction, As 2 03-f-2 I 2 +2 H 2 0^rAs 2 05-|-4 HI, and thus cause it to run to completion from left to right. The reaction may then be written : Na 2 HAs0 3 +I 2 +2 NaHC0 3 =Na 2 HAs0 4 +2 NaI+2 C0 2 +H 2 0. 6. Since the addition of iodine in excess to the weakly basic bicarbonate solution is likely to lead to a slight degree of action, it is best in this titra- tion not to overstep the end-point. 7. Iodine is a rather expensive chemical and it is well worth while to recover it from the united residues of a large class. THE DETERMINATION OF ANTIMONY IN STIBNITE The sample for analysis should be an antimony ore, practically free from arsenic and iron, and with hydrochloric acid it should leave only a siliceous residue. Procedure. Weigh out two 0.20 g. portions of the finely ground mineral into dry i5o-cc. beakers. Cover the beakers, add 5 cc. of i2-normal hydrochloric acid, and allow the acid to act in the cold for 10 minutes; then heat gently on the steam bath, for about 15 minutes, until the residue is white. Add 2 g. of powdered tartaric acid and gently warm the mixture for 10 minutes longer. Do not allow the liquid to evaporate sufficiently to expose any part of the bottom of the beaker. Dilute the solution cautiously with 5-cc. portions of water; if a red coloration appears, stop the dilution, warm until the solution is colorless, and again dilute. Continue the dilution VOLUMETRIC ANALYSIS 149 until a volume of 100 cc. is reached, and boil for a minute. Neutralize the clear, cold solution with sodium hydroxide (methyl orange), and then acidify it with dilute hydrochloric acid, a drop or two in excess. Dissolve 10 g. of sodium bicarbonate in 400 cc. of water, place 200 cc. of this solution in each of two yoo-cc. beakers, and trans- fer to these the cold solutions of the ore, avoiding loss by effer- vescence. Add 5 cc. of starch solution, and titrate each mixture with the standard iodine solution, to the appearance of the blue color. Do not overstep the end-point. From the corrected data, calculate the percentage of antimony in the stibnite. NOTES. i. Stibnite is essentially native antimony sulphide, Sb 2 S 8 , and upon treatment with hydrochloric acid hydrogen sulphide is liberated ; but this is partially absorbed by the acid. The gas should be wholly ex- pelled during the heating on the steam bath ; if it is not completely driven out, antimony sulphide will begin to separate at some point in the dilution. In that case, however, if the dilution is at once stopped and the solution heated, the hot acid will redissolve the sulphide, and the hydrogen sulphide may then be expelled. The final boiling is to insure the absence of hydro- gen sulphide, which itself reacts with iodine. 2. Antimony trichloride hi the presence of strong hydrochloric acid is some- what volatile, and for this reason the solution should not be boiled before dilution. If the solution is gently heated as described, no error need be feared from this source. 3. If, for any reason, a white precipitate of oxy chloride separates during the dilution or neutralization, it is best to reject the solution and start anew. Antimony chloride is readily hydrolyzed upon dilution, with the precipitation of basic compounds such as SbOCl; but the addition of tar- taric acid leads to the formation of stable antimonyl tartrates, which are soluble. In this way the antimony is kept in solution. 4. The reaction between the iodine and the antimonyl tartrate is not so simple, but for purposes of calculation it is accurately expressed by the equation, Sb 2 3 +2 I 2 +2 H 2 0^Sb 2 05+4 HI. The purpose of the bicar- bonate is here also to neutralize the hydriodic acid formed, and thereby drive the oxidation to completion. 5. The sodium hydroxide is added merely to neutralize most of the acid, and to make it easy to provide for the presence of a known amount of 150 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS sodium bicarbonate. The solution should be slightly, but distinctly, acid when the bicarbonate is added. 6. If the ore to be analyzed contains more than traces of iron, it is dis- solved in hydrochloric acid, the antimony is precipitated with hydrogen sulphide, and the washed precipitate is redissolved in hydrochloric acid and determined as above. In case arsenic also is present, a somewhat more complicated separation of the antimony is necessary. THE DETERMINATION OF LEAD IN AN ORE Procedure. Weigh out two samples of the finely ground ore sufficient to contain about 0.20 g. of lead (0.28-0.29 g- f a 7% ore), and treat each as follows : Moisten the sample with water, add 15 cc. of i2-normal hydrochloric acid, and evaporate on the steam bath to about 5 cc. Add 3 cc. of strong nitric acid, evaporate nearly to dryness, then add 20 cc. of 6-normal hydro- chloric acid and again heat to bring all the lead chloride into solution. Add 20 cc. of 6-normal sulphuric acid and evaporate to* white fumes. Allow to cool, add 50 cc. of water, boil, and then add 15 cc. of alcohol ; stir, allow to settle, and filter. Wash the lead sulphate and gangue six times with lo-cc. portions of o.5-normal sulphuric acid (15 cc. of 6-normal acid in 165 cc. of water), transfer the residue to a small beaker by means of a jet of water, and heat it gently for a few minutes with 20 cc. of ammonium acetate solution ; l filter the liquid through the orig- inal filter and wash the latter with small portions of the hot ammonium acetate solution. Dilute the extract to 150 cc., heat to boiling and add from a pipette 10 cc. of potassium dichro- mate solution. 2 Boil the mixture gently for 10 minutes, filter off the precipitate of lead chromate and wash the filter and precipitate about ten times with lo-cc. portions of dilute am- monium acetate solution (25 cc. of the extraction solution diluted to 250 cc.), until the excess of potassium chromate is completely removed. 1 Made by neutralizing 30% acetic acid with 6-normal ammonia, and then add- ing a slight excess of ammonia. 2 A solution containing 75 g. of K 2 Cr 2 O 7 per liter. VOLUMETRIC ANALYSIS 151 Now place a clean 5oo-cc. Erlenmeyer flask under the funnel, and with a jet of cold, acid, sodium chloride solution 1 stir up and dissolve the precipitate ; continue washing with the same liquid until every trace of color is removed from the filter. In any case, use at least 50 cc. of the liquid. Finally dilute to 150 cc., add i g. of potassium iodide, mix, and titrate at once with a solution of sodium thiosulphate (which has been stand- ardized in the same way against test lead, see Note 5) until the brown color becomes faint ; then add 5 cc. of starch solution, and continue the titration cautiously until the solution becomes pale green (CrCl 3 ) with no tinge of blue. The end-point is very sharp, but without great care it may easily be passed. It is best to have a white surface under the flask. Report the percentage of lead in the ore. NOTES. i. The ore is first heated with strong hydrochloric acid in order to expel most of the sulphur. Nitro-hydrochloric acid is then used to decompose any refractory sulphides. Upon evaporating the chloride solution to white fumes with sulphuric acid, the volatile acids in which lead sulphate is slightly soluble are completely expelled, and upon dilution with water, especially if alcohol is added, the lead is all left in the residue as lead sulphate. 2. Lead sulphate is readily dissolved by ammonium acetate solution, owing to the exceptional behavior of lead acetate with respect to ionization (see Part I), leaving the siliceous gangue, BaS0 4 , etc., as a residue. 3. While lead is not precipitated from solutions containing a large ex- cess of acetate ion by sulphates, the addition of a soluble chromate causes the precipitation of lead chromate. This behavior is due to the fact that such solutions contain Pb ++ -ion at an extremely low concentration (owing to the presence of the lead mainly in the form of intermediate or com- plex ions, as (Pb . C 2 H 3 2 ) + , [Pb(C 2 H 3 2 ) 3 ]~, etc.), and also to the fact that lead sulphate is very much more soluble than lead chromate ; the lead-ion concentration is still great enough in such solutions to cause the solubility product of lead chromate to be exceeded upon the addition of potassium chromate in excess. 4. The solubility of lead chromate in the acid chloride solution is due on the one hand to the lowered concentration of the chromate ion, owing to 1 Mix 10 cc. of i2-normal hydrochloric acid with 15 cc. of water, and add this mixture to 100 cc. of a saturated solution of sodium chloride. 152 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the formation of non-ionized H 2 Cr0 4 , HCr0 4 ~, etc., and on the other hand to the great tendency of lead ion to form soluble complexes with chloride solutions. (Cf. the solubility of silver chloride in chloride solutions.) 5. The reaction of the acid solution with potassium iodide is most simply represented by the equation, Cr 2 Of-+6 I-+I4 H+= 2Ci++N-7 HzO+3 I 2 ; from this it may be seen that one atom of lead (as PbCr0 4 ) leads to the liberation of 3 atoms of iodine. But the results vary slightly with the conditions, and for that reason the thiosulphate solution must be standardized under identical conditions against a known amount of lead. In this case, 0.20 g. of test lead should be dissolved in 5 cc. of 6-normal nitric acid, the solution evaporated to white fumes with 20 cc. of 6-normal sulphuric acid, and the subsequent operations carried out as described in the procedure. THE DETERMINATION OF COPPER IN AN ORE Principle. This method is based upon the reaction which takes place upon the addition of potassium iodide to a slightly acid copper salt solution; cuprous iodide is precipitated as a cream-colored powder, and iodine is set free : 2 CuS0 4 +4 KI = 2 K 2 SO 4 +Cu 2 l2+l2. The iodine is promptly titrated with a standard thiosulphate solution. Standardization of the Thiosulphate Solution. Weigh ac- curately two portions of pure bright copper wire or foil, of 0.15- 0.16 g. each, and, in 25o-cc. Erlenmeyer flasks, dissolve these in 5-cc. portions of 6-normal nitric acid. Dilute each solution to 15 cc. and boil to expel the red fumes; then dilute to 25 cc. and add ammonia (sp. gr., 0.90) in slight excess. Again boil until the ammonia odor is faint, add 80% acetic acid, 2-3 cc. in excess, and boil for a moment longer, agitating the flask in a holder to prevent bumping. Cool to room temperature, dilute to 40 cc., add a solution of 3 g. of potassium iodide in 10 cc. of water, and titrate at once with the approximately tenth-normal thiosul- phate solution to a faint brown tinge ; add 5 cc. of starch solu- VOLUMETRIC ANALYSIS 153 tion, and continue the titration until the last faint lilac tint is removed by a single drop. Do not overstep the end-point. From the data obtained, calculate the value of the solution per cubic centimeter in terms of copper. $> Analytical Procedure. Weigh out into 3oo-cc. beakers samples of the ore sufficient to furnish about 0.15 g. of copper, and treat each as follows: Add 10 cc. of hydrochloric acid (sp. gr., 1.19) and 5 cc. of nitric acid (sp. gr., 1.42) and heat in the covered beaker on the hot plate until decomposition is complete, adding more of the acids if necessary, and enough water at the end to hold all soluble salts in solution. Then add 15 cc. of 6-normal sulphuric acid, and continue the heating until abundant white fumes begin to come ofl. Cool, add 50-60 cc. of water, boil for a moment, and allow to stand, hot, until any anhydrous ferric sulphate has dissolved. Finally, filter off from any lead sulphate, gangue, and sulphur, receiving the filtrate and washings in a 3Oo-cc. beaker. Now add a solution of 5 g. of sodium thio- sulphate in 25 cc. of water, boil to coagulate the precipitate, and filter, transferring the precipitate quantitatively to the filter by means of hot water. Dry the precipitate on the filter. Place the precipitate, together with the filter, in a porcelain crucible, ignite gently until the filter is consumed, and allow to cool. Transfer the bulk of the precipitate to a 2 5o-cc. Erlenmeyer flask, and set aside. To dissolve the last portions of the pre- cipitate from the crucible, add 3 cc. of concentrated nitric acid and 2 cc. of water, and warm gently on the hot plate, finally pouring the acid solution into the flask containing the bulk of the precipitate, and washing out the crucible with a few small portions of 6-normal nitric acid. Heat the mixture in the flask until the decomposition is complete, dilute to 25 cc., boil, add ammonia in slight excess, and heat until the odor is fault. Add 80% acetic acid, 2-3 cc. in excess, and boil for a moment, vigor- ously agitating the flask to prevent bumping. Cool to room temperature, dilute to 40 cc., add 3 g. of potassium iodide dis- solved in 10 cc. of water, and titrate at once with the thio- 154 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS sulphate solution, as previously described. Report the per- centage of copper in the ore. NOTES. i. Since iron and other elements likely to be present inter- fere with the process, the copper must be separated from these. Lead is first removed by means of sulphuric acid, after which the copper is pre- cipitated from the hot, acid solution by means of sodium thiosulphate ; this gives a flocculent precipitate of cuprous sulphide mixed with sulphur, which filters readily and can be washed with hot water without fear of oxidation. Arsenic and antimony, if present, are also precipitated, but under the treatment prescribed the usual quantities of these elements are without influence. They are mostly volatilized during the ignition. If antimony is present in appreciable quantity, it is perhaps better to filter the solution before the addition of the ammonia. 2. In order to obtain the best results it is necessary to standardize the thiosulphate solution against pure metallic copper. When this is done the method is very accurate ; otherwise the results are not so good. For ex- ample, a thiosulphate solution which (titrated against a freshly stand- ardized iodine solution) had a calculated copper value of 0.00608 g. per cubic centimeter, was found upon standardization against pure copper to have a value of 0.006 n g. per cubic centimeter. 3. Since nitrous fumes liberate iodine from potassium iodide, they must be completely expelled by boiling before the addition of the salt. The expulsion of the last traces of these fumes is insured by boiling the solu- tion after it has been acidified with acetic acid. 4. The return of the blue tinge in the liquid after long standing is of no significance, but a quick return which is not prevented from recurring by the addition of a single drop of the thiosulphate solution is usually an evidence of faulty work. 5. In such a case, or if the end-point has accidentally been passed, the same sample may be prepared anew for titration : Add 10 cc. of concen- trated nitric acid, and heat very cautiously, with great care not to allow the mixture to foam over. After most of the iodine has been expelled, manipulate the flask (in a holder) over a free flame and boil the solution down rapidly to a volume of 5-10 cc. Dilute to 25 cc. with water, boil, add ammonia in slight excess, and finish as described in the procedure. 6. In the electrolytic determination of copper in ores containing arsenic and other interfering substances, a satisfactory copper solution is most readily prepared by dissolving the ignited thiosulphate precipitate in a suitable quantity of strong nitric acid, with subsequent dilution to the required volume. VOLUMETRIC ANALYSIS 155 C. PRECIPITATION METHODS General Discussion. The completion of neutralization re- actions depends upon the very slight degree of ionization of one of the products, water. The completion of reactions of oxida- tion and reduction most often depends upon the relative poten- tials of oxidizing and reducing agents under specific experi- mental conditions. Certain other reversible reactions which serve as the basis of volumetric processes run to completion in consequence of the formation of very slightly soluble precipi- tates. In most cases an indicator is used, but in some the cessa- tion of precipitation with the further addition of the standard solution indicates the completion of the reaction. An example of the latter kind is found in Gay-Lussac's method for silver, which dates from 1832, and which is still widely used in determining the fineness of silver bullion. When silver chloride first separates it is finely divided, and a very minute quantity can easily be recognized; if the solution is shaken vigorously the precipitate coagulates and settles, leaving a supernatant liquid which is perfectly bright and clear. Hence, if silver nitrate is titrated with a solution of sodium chloride and the mixture well shaken in a stoppered bottle after each addition, the point at which the addition of a further quantity of the standard solution fails to produce a precipitate can readily be determined. Near the end-point it is customary to use a standard solution ten times as dilute as that used at the start. In the case of this reaction, this method of determining the end-point admits of a very high degree of accuracy, and it is the method in use at the government mints. Since, however, it is rather tedious and demands considerable skill and experi- ence, a slightly less accurate but much more convenient method is generally employed. Silver thiocyanate is even less soluble than silver chloride, and it is therefore possible to titrate silver very accurately with a standard solution of potassium or ammonium thiocyanate. If a 156 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS solution of a ferric salt, acidified to suppress hydrolysis, is present as an indicator, the first drop of thiocyanate solution in excess will impart to the mixture a pink tint. This method (Volhard's) is also suitable for the determination of the halogens (except fluorine) and of certain other ions which give silver compounds insoluble in dilute nitric acid. A measured volume of standard silver nitrate solution is added in excess, and the excess is then determined by means of the standard thiocyanate solution. THE PREPARATION AND STANDARDIZATION OF APPROX- IMATELY ONE-TENTH NORMAL SOLUTIONS OF SILVER NITRATE AND POTASSIUM THIOCYANATE Procedure. Dissolve 5.0 g. of potassium thiocyanate (or the equivalent quantity of the ammonium salt) in water and dilute the solution to 500 cc. Also dissolve 8.5 g. of silver nitrate in water and dilute the solution to 500 cc. Further, mix 10 cc. of 6-normal nitric acid with 40 cc. of water, heat the solution to boiling, and dissolve in the hot liquid 5 g. of pure ferric alum. Allow the solution to cool, and keep it for use as an indicator. Now fill the burettes with the respective solutions, placing the silver nitrate solution in a glass-stoppered burette. (Ob- serve the usual precautions to prevent dilution, and place all solutions and precipitates containing silver in the receptacle for silver residues.) Run out 20 cc. of the silver nitrate into a beaker, dilute to 150 cc., add 10 cc. of 6-normal nitric acid which has been recently boiled, and 5 cc. of the indicator solution. Run in the thiocyanate solution until, after vigorous stirring, a faint pink tinge can be detected in the solution. If the end-point is overstepped, titrate back with the silver nitrate solution. From the corrected volumes used, calculate the ratio of the thiocyanate to the silver nitrate solution. Repeat until the results do not differ by more than two parts in one thousand. Finally, standardize the silver nitrate solution, as follows: Weigh out portions of pure sodium chloride, of 0.12-0.14 g. each, dissolve these in 75-cc. portions of water, heat to boiling, VOLUMETRIC ANALYSIS 157 and with stirring run into each from a burette 25.00 cc. of the silver nitrate solution. Add 10 cc. of freshly boiled 6-normal nitric acid, stir, and filter, washing the precipitate by decantation with several small portions of hot distilled water, and pouring these slowly over the filter; the united filtrate and washings should have a volume of about 150 cc. To this solution add 5 cc. of the indicator, and titrate the excess of silver with the thiocyanate solution, as already described. From the data obtained, calculate the normality factor of the silver nitrate solution; and from the mean of the duplicate values, which should agree within two parts in a thousand, calculate the nor- mality factor of the thiocyanate solution. NOTES. i. The reactions between the thiocyanate and the indicator are essentially as follows : Fe +++ +6 CNS-3*Fe(CNS).+3 It will be recalled that in testing for ferric iron with potassium thiocyanate, it is necessary to add a large excess of the latter in order to detect the smallest possible quantity of iron. In the same way, when using ferric iron as an indicator for thiocyanate, it is necessary to provide a high concentration of the former in order to detect the slightest possible excess of the thio- cyanate in the solution. The reactions which give rise to the colored sub- stances are reversible, but in the presence of a large excess of one of the colorless constituents the dissociation of the colored substances is prevented by mass action. 2. Nitric acid is added to the solution to be titrated in order to prevent the hydrolysis of the ferric salt, which would impart a brownish red color to the mixture. It is boiled to free it from nitrous fumes, though this is of less importance here than in testing for iron in the presence of nitric acid ; nitrou? fames color the thiocyanate pink. 3. Sodium chloride may easily be obtained pure by filtering a concen- trated solution of the commercial salt, saturating it with hydrogen chloride gas, and filtering off the precipitate. The latter is washed with strong hydrochloric acid and dried at 150, or higher. 4. Standard solutions of silver nitrate can of course be prepared by the solution of the calculated amount of pure metallic silver in nitric acid, and dilution to the required volume; or by means of the calculated weight of pure silver nitrate. 158 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS THE DETERMINATION OF CHLORINE IN A SOLUBLE CHLORIDE The sample may be an artificial mixture of the chloride and carbonate of sodium. Procedure. Weigh out into 3oo-cc. beakers, two portions, each sufficient to contain about 0.12 g. of sodium chloride, and treat each as follows : Dissolve the sample in 50 cc. of water, run in from a burette 30.00 cc. of the standard silver nitrate solution, and carefully acidify the mixture with dilute nitric acid. Heat to boiling, see that the liquid is distinctly acid, and filter. Receive the filtrate and washings in a 3oo-cc. Erlenmeyer flask. To the united filtrate and washings, which should have a volume of about 150 cc., add 10 cc. of 6-normal nitric acid and 5 cc. of the indicator solution, and titrate with the standard thiocyanate solution, as already described. Calculate the per- centage of chlorine in the sample. NOTES. i. Since silver chloride is several times as soluble as silver thiocyanate, the former must be filtered off before the titration of the excess of silver nitrate ; otherwise the silver chloride would react with the thiocyanate solution and render the end-point uncertain. This behavior is best represented by the following system of equilibria : JC1-+ lAg + (Solid) (Diss'd) lAg + (+CNS-^AgCNS^AgCNS). (Diss'd) (Solid) That is, if the silver chloride were left in the mixture during the titration, owing to the slow conversion of the soluble (colored) thiocyanate com- pounds into insoluble silver thiocyanate, there would be no permanent end-point. 2. Silver bromide and silver iodide are less soluble than silver thio- cyanate, so that in the determination of bromine and iodine by this method it is not necessary to filter. 3. Soluble chlorates, etc., may be determined by this method by first reducing them to the corresponding halides (e.g. with sulphurous acid), and then determining the latter. 4. For other uses of precipitation methods, see Part IV, Problems, 90 91, 92, and 93. PART IV STOICHIOMETRY Preliminary Discussion. The stoichiometrical problems met with in analytical work are, as a rule, neither hard to compre- hend nor difficult to solve. The beginner will find that a mod- erate amount of time devoted to the intelligent study of these problems will enable him rapidly to make the calculations necessary for the interpretation of analytical data; the ability to do this is at least as important as the manipulative skill by which the data are obtained. It cannot be too strongly emphasized that, in making such cal- culationSj the beginner should from the outset strive to take the shortest and most direct route to the result. With a little practice, the student who is not unacquainted with the reactions of an- alytical chemistry should soon acquire the ability to recognize at once, upon the inspection of a problem, the factors which will lead most directly to its solution, 1 as well as the equivalent 1 Of course most analytical problems can be solved in stages, by means of a series of proportions, and it is perhaps only natural that most beginners should have a predilection for this method. In the examples given, however, the common factors have been eliminated, and the problems solved hi a single opera- tion. The beginning student will better appreciate the advantages of the shorter method upon comparing the solutions given of problems IV and V with the follow- ing roundabout method of arriving at the same results : iv. (a) 10 Fe : 2 KMnO 4 = 0.005 : w. ^ = 316/558X0.005 = 0.00283 g- KMnO 4 per cc. (6) 2 KMnO 4 : 5 H 2 C 2 O 4 = 0.00283 : x. #=450/316X0.00283 = 0.00403 g. H 2 C2O4 per cc. (c) 5 H 2 C 2 O 4 : 5 CaC 2 O 4 = 0.00403 : y. y = 640/450X0.00403 = 0.005 73 g- CaC 2 O 4 per cc. (rf) 5 CaC 2 O 4 : 5 CaO = 0.005 73 : * 2 = 280.5/640.5X0.00573=0.00251 g. CaO per cc. Ans. 160 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS relationships of the substances involved. To do this, however, he should understand and bear in mind the relationships and dif- ferences which exist between chemical and physical units, such as atoms, molecules, and equivalents on the one hand, and grams and cubic centimeters on the other. Detailed solutions of a few typical problems are given below. The student should study these careiully until they are fully understood. i. A sample of a soluble chloride weighing 0.200 7 g. yields on analysis 0.4920 g. of silver chloride; what percentage of chlorine does it contain? { From the proportion, Wt. of chlorine in sample : Wt. of sample = % of chlorine : 100, ... , . , , , Wt. of chlorine in sample ~ r , , . it is obvious that ^ - 100 = % of chlorine. Wt. of sample Also, since the chlorine contained in the sample is identical with that which is later contained in tlie silver chloride precipitate, we have the proportion, Cl : AgCl= Wt. of chlorine . Wt. of silver chloride, Cl or, Wt. of silver chloride =Wt. of chlorine. Substituting this value in the preceding equation, we get, -^- Wt. of silver chloride 354(5 - 0.4020 AgC1 _ f = 100= H^34 Wt. of sample 0.2007 = 60.50% of Cl. v. (a) AgCl:HCl=o.i527:*. x = 36.46/143.34X0.1527 = 0.0388 g. HC1 in 20.50 cc. (6) 20.50: 1000 = 0.0388 : y. y= 1000/20.5X0.0388= 1.893 g- HC1 in one liter. (c) 36.46:1.893 = 1:2. z= 1.893/36.46=0.0519 N. Ans. STOICHIOMETRY 161 A chemical factor expresses the quantity by weight of an ele- ment or compound which is equivalent to one part by weight of some other substance. For example, the ratio or factor Ag 107.88 -T ^r= - ' - = 0.7526 AgCl 143.34 tells us that one gram of silver chloride contains 0.7526 g. of silver, and if we wish to calculate what weight of silver there is in a specific weight of silver chloride, we simply multiply the latter by this factor; e.g. 10.15 g. of silver chloride contain 10.15 Xo.7526 = 7.64 g. of silver. Again, if the weight of FeO which corresponds to a specific weight of Fe 2 C>3 is desired, the factor is 2 FeO _ 143.68 _ Fe 2 3 " 159.68" And, similarly, if it is wished to find the weight of K 2 which corresponds to a specific weight of KC1, the factor is In the calculation of these (physical-unit) factors, the equiva- lent relations of the two substances must be kept clearly in mind ; thus it is plainly incorrect to express the ratio of potassium TT O oxide to potassium chloride by the fraction , since each JxC/1 molecule of K 2 must yield upon treatment with HC1 two mole- cules of KC1. Similarly, the factor for the conversion of Mn 2 P 2 07 into Mn 3 04 is - 3 4 ; for two molecules of MnsO^ contain- 3 Mn 2 P 2 O 7 ing six atoms of manganese, will yield three molecules of Mn 2 P 2 7 , also containing six atoms of manganese. Carelessness in this respect is one of the most frequent sources of error. ii. How many cubic centimeters of a solution containing 25 grams of BaClz 2 H Z per liter will be required to precipitate the sulphur from 0.1073 S ram f P ure stibnite, Sb^S^ as BaSOt? 162 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Each molecule of 80283 will yield upon treatment three mole- cules of H 2 S0 4 , and these will require three molecules of barium chloride for precipitation. We therefore arrive at the propor- tion, 3 (BaCl 2 2 H 2 0) : 80283=3; : 0.1073, where x represents the weight of the crystalline salt required. That is, _ 3 (BaCl 2 .2H 2 Q) Sb 2 S 3 in which the factor 3 (Ba ^ 2 g 2 H20) =2.177 indicates the quan- 00203 tity by weight of BaCl 2 . 2 H 2 which is required to precipitate the sulphur from one gram of Sb 2 S 3 ; this factor therefore does not differ essentially from the chemical factors previously discussed. Finally, since each cubic centimeter of the solution contains 0.025 g. of BaCl 2 . 2 H 2 0, we have, 9-34 cc. 0.025 0.025 Hi. What volume of aqueous ammonia of sp. gr. 0.960, contain- ing 9.91% of NH$, will be required to precipitate, as Fe(OH)$, the iron contained in 1.475 of Fe(NH4$Ot)2.6HzO?- Since the iron is to be precipitated, after oxidation, as Fe(OH)s, it is plain that each atom of iron will require three molecules of NH 4 OH, which in turn are furnished by three molecules of NH 3 . Therefore, , -' I -^7= wt. of NH 3 required; Fe(NH 4 S0 4 ) 2 and, since each cubic centimeter of the aqueous ammonia weighs 0.960 g., and contains 9.91% of NHs, the solution contains 0.960 Xo.0991 g. of NH 3 per cubic centimeter. That is, 51.10 3O2.I6' 1 * 475 - - 2.02 cc. 0.960 Xo.0991 0.960 Xo.0991 STOICHIOMETRY 163 iv. A solution of potassium permanganate is equivalent to 0.00500 g. of ferrous iron per cubic centimeter; what is its value in terms of calcium oxide ? The reactions involved in the volumetric determinations of iron and calcium are : 10 FeSO 4 +2 KMnO 4 +9 H 2 SO 4 = 2 KHS0 4 +2 MnSO 4 +5Fe 2 (S0 4 ) 3 +8H 2 0, and 5 CaC 2 O 4 +2 KMn0 4 +9 H 2 SO 4 = 5 CaSO 4 +2 KHSO 4 + 2MnSO 4 +ioCO 2 +8H 2 0; and from these equations it is clear that, in this case, 2 Fe ++ =c=CaO. That is, CaO 56. i 0.00500 = - 0.00500 = 0.00251 g. CaO. 2 Fe 1 1 1. 6 v. If 20.50 cc. of hydrochloric acid yield 0.1527 g. of silver chloride, what is the normality factor of the solution? Although silver chloride is insoluble, the normality factor of the acid may nevertheless be calculated directly from the weight of the precipitate obtained. One liter of normal hydrochloric acid, containing one mol of HCl, would yield 143.34 g. (i.e. one mol) of silver chloride, whence 20.50 cc. would yield 0.14334 X 20.50 g. We obtain, therefore, the equation, o.: 1 0.14334X20.50 vi. A sample of stibnite weighing 0.1793 g. is heated with strong HCl, and the H 2 S evolved absorbed by means of sodium hydroxide solution; the resulting mixture (containing the sulphur as Na^S) being introduced under the surface of a solution made by adding 25 cc. of 6-normal HCl and 50.00 cc. of o.n6o-normal iodine to 500 cc. of water. The excess of iodine is titrated with 0.0957- 1 64 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS normal sodium thiosulphate solution, of which 28.57 cc - are re- quired. Calculate the percentage of (evolved) sulphur in the stibnite. (H 2 S+I 2 = 2HI+S.) cc. N. F. 50.00X0.1160 = 5.800 cc. of normal iodine. 28.57 Xo.OQ57 = 2.734 cc. of normal thiosulphate. I.e. the H 2 S required 3.066 cc. of normal iodine. Since normal iodine has a sulphur value of '^ 20 ^= 0.01604 5 g- per cubic centimeter, we have, 0.016035 X 3 .o66 XioQ Qf ^ The normality factor of a solution expresses the value of the solution per cubic centimeter in terms of a normal solution. For example, if a solution is known to be one half normal (i.e. N. F.= 0.500), it is obvious that i cc. of it is equivalent to 1.000X0.500=0.500 cc. of a normal solution; or that 27.31 cc. of it are equivalent to 27.31X0.500 = 13.655 cc. of a normal solution. Knowing the normality factors of a series of solutions, therefore, we can readily reduce the different volumes of the solutions used in a determination to a common standard ; and in this way the calculations are rendered almost as simple as if the tune and labor had been expended to make the solutions all of exactly the same strength, say one-tenth normal. mi. Indirect methods of analysis depend upon the fact that when two or more substances are made to undergo the same chemical treatment they either experience a relatively different change of weight, or unit weights of each require unequal volumes of a standard solution. For example, suppose we wish to determine the weight of NaCl and of KCl in a mixture of the two salts. The mixture, weighing a grams, may be converted into silver chloride, of which there is formed, say, p grams. STOICHIOMETRY 165 Let x represent the weight of the sodium chloride, and y that of the potassium chloride, and we have, and AgCL AgCl NaCl* h KCl y P If we designate by m the factor ^ and by n the factor % we obtain, and mx+ny=p, from which we find that or Indirect analyses may in general be calculated by means of this or a similar general equation. In the above example, and 1^^ = 0.5297. If these values are substituted in the general equation, we obtain, x = 1.888 #-3.628 a. Consequently, in order to determine the weight of sodium chloride in the mixed sample it is only necessary to determine the values of a and p, multiply them by 3.628 and 1.888, re- spectively and subtract the first product from the second. The sa me analysis might be performed by weighing the mixed chlorides in a platinum crucible, then changing them to sulphates (by evaporation with H 2 SO 4 ) , and again weighing. In this case also, mn mn 21. 547 #-25.181 a. i66 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS In the first case, the coefficients are relatively small, and consequently good results might be expected, since the experi- mental errors made in the determination of a and p are multiplied by only 3.63 and 1.89, respectively. In the latter case, however, the coefficients are very large, and the unavoidable analytical errors would have to be multiplied enormously in the calculation ; the method is therefore worthless. Although some indirect methods may appear simple and attractive on paper, they frequently lead to impossible values in practice ; so that extreme caution should be exercised regard- ing the use of an indirect method. In general, if accurate and reliable results are desired, indirect methods of analysis should be avoided. PROBLEMS GRAVIMETRIC ANALYSIS 1. Calculate the chemical factors for the following : KC1 from K 2 PtCl 6 ; K 2 from K 2 PtCl 6 ; P from Mg 2 P 2 O 7 ; Fe 3 O 4 from Fe 2 O 3 ; MnO 2 from Mn 3 O 4 . 2. What weight of Mn 3 4 corresponds to 0.5785 g. of Mn 2 P 2 O 7 ? To 0.4327 g. of MnSO 4 ? 3. A sample of an impure ammonium salt weighing 0.4988 g. is con- verted into (NH 4 ) 2 PtCl 6 , and this upon ignition yields 0.3258 g. of platinum. Calculate directly from the weight of platinum the percentage of NH 3 in the sample. 4. A sample of phosphorus pentoxide weighing 0.2018 g. yields 0.3132 g. of Mg 2 P 2 0?. Calculate the percentage of P 2 O5 in the sample. 5. What weight of a silver nitrate solution known to contain 2.31% of Ag will be required to precipitate the chlorine from 25.0 cc. of a solution containing 25.0 g. of BaCl 2 . 2 H 2 O in one liter? 6. If 25.0 cc. of sodium chloride solution yield 0.1434 g. of silver chlo- ride, what is the strength of the solution in grams of the salt per liter ? In mols per liter? 7. How many cubic centimeters of a solution containing 25.0 g. of BaCl 2 . 2 H 2 per liter will be required to precipitate, as BaSO 4 , the sul- phuric acid formed upon oxidizing 0.2543 g. of FeS 2 with fuming nitric acid? 8. How many cubic centimeters of hydrochloric acid of sp. gr. 1.050, containing 10.17% of HC1, will it take to precipitate the silver from a solu- tion containing 0.8430 of silver sulphate? STOICHIOMETRY 167 9. How "many cubic centimeters of hydrochloric acid of sp. gr. 1.040, containing 8.16% of HC1, will be required to dissolve one gram of calcium carbonate ? 10. What weight of Mn 2 P 2 O7 is it possible to prepare from 50.0 cc. of a permanganate solution which contains 4.500 g. of KMnO 4 per liter? 11. How many cubic centimeters of a solution of sp. gr. 1.116, con- taining 10.06% of NaOH- will it take to neutralize a solution containing 5.00 g. of NaHS0 4 ? 5.00 g. of KHSO 4 ? 12. A sample of impure potassium sulphide weighing 0.4320 g. is treated with hydrochloric acid, and by means of ammoniacal hydrogen peroxide solution the hydrogen sulphide evolved is converted into ammonium sul- phate. This yields 0.8034 g. of BaSO 4 . Calculate the percentage of K 2 S in the sample. 13. A sample of stibnite weighing 1.078 g., upon being analyzed by the method indicated in Problem 12, yields 0.6750 g. of BaSO 4 . Assuming the sulphur to be present wholly as Sb 2 S 3 , calculate the percentage of the latter in the mineral. 14. How many cubic centimeters of aqueous ammonia of sp. gr. 0.96, containing 9.91 % of NH 3 , will be required to precipitate the aluminum in 0.8674 g. of KA1(SO 4 ) 2 . 12 H 2 ? How many cubic centimeters of 6-normal ammonia ? 15. What volume of the ammonia water first referred to in Problem 14 will it take to neutralize 10.0 cc. of hydrochloric acid of sp. gr. 1.12, con- taining 23.81% of HC1? To neutralize 10.0 cc. of 6-normal hydrochloric acid? 16. If 15.0 cc. of a solution of barium chloride yield, upon evaporation with hydrochloric acid and gentle ignition, 1.563 g. of the anhydrous salt, what is the strength of the solution in mols per liter? In equivalents per liter? 17. A solution contains 2.25% by weight of BaCl 2 . What volume of a solution containing 5.000 g. of Ag 2 S0 4 per liter will be required to pre- cipitate the chlorine in 5.000 g. of the barium chloride solution ? What will be the weight of the precipitate formed ? 18. A sample of pyrite weighing 0.2500 g. yields upon analysis 0.8020 g. of BaS0 4 . Upon the assumption that the sulphur is wholly present as FeS 2 , calculate the percentage of the latter in the sample. 19. What volume of bromine water containing 30 g. of bromine per liter will be required to oxidize the iron in 1.75 g. of FeSO 4 . 7 H 2 O? 20. What volume of aqueous ammonia (sp. gr., 0.96, containing 9.91% of NH 3 ) will be required to precipitate the iron, after oxidation with hydro- gen peroxide, from a solution containing 0.750 g. of Fe(NH 4 S0 4 ) 2 . 6 H 2 O 168 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS and 12.0 cc. of hydrochloric acid (sp. gr., 1.12, containing 23.8% of HC1)? 21. A mixed sample of CaO, Ca(OH) 2 , and CaC0 3 weighing 0.5896 g. is evaporated with excess sulphuric acid, and gently ignited ; the residue weighs 0.8651 g. What volume of 6-normal hydrochloric acid will be re- quired to convert 5.00 g. of the sample into calcium chloride? 22. The ignited precipitate of ferric and aluminum oxides from 1.497 g- of a mineral weighs 0.4196 g. ; after ignition in a current of hydrogen the product weighs 0.3311 g., the ferric oxide being reduced to metallic iron. Calculate the percentage of Fe 2 C>3 and of AUOs in the mineral. 23. A sample of pyrolusite weighing 0.5124 g. is heated in the presence of dilute sulphuric acid with an excess of oxalic acid, and the gas evolved is absorbed in a weighed bulb containing potassium hydroxide. The gain in weight of the bulb is found to be 0.4789 g. Calculate the percentage of Mn0 2 in the pyrolusite. 24. What volume of o.5-normal ammonium oxalate solution will be required to precipitate, as CaC 2 O 4 . H 2 0, the calcium from one gram of apatite, [(^(POOtJi - CaF 2 ? 25. What volume of a solution containing 66 g. of (NH 4 ) 2 HP0 4 per liter will be required to precipitate, as ZnNH 4 P0 4 , the zinc from 0.9786 g. of a brass which contains 30.15% of zinc? What is the normality of this solu- tion as a precipitant for zinc ? 26. How many grams per liter of K 2 Cr 2 0: must a solution contain in order that, by reduction of the chromium (with HC1 and S0 2 ), precipitar tion with ammonia, and ignition of the precipitate in a current of hydro- gen, a 50.0 cc. portion shall yield 0.3752 g. of Cr 2 8 ? 27. What volume of a solution containing 25.0 g. of KH 3 (C 2 4 ) 2 . 2 H 2 O per liter will be required to precipitate the calcium from 0.976 g. of a lime- stone which yields, besides a small quantity of an insoluble residue, 2.14% of Fe 2 0a, 9.56% of MgO, and 45.36% of C0 2 , assuming iron, magnesium, and calcium to be present wholly as carbonates (the iron of course as ferrous carbonate) ? 28. What volume of 6-normal sulphuric acid will be required to replace the nitric acid in the salts obtained upon evaporating to dryness with nitric acid 4.984 g. of brass, if the brass contains 65.98% of Cu, 31.42% of Zn, 1.84% of Sn, and 0.76% of Pb? 29. A sample of silicate mineral weighing 1.0245 g. yields 0.2602 g. of potassium and sodium chlorides; and the mixed chlorides yield 0.4304 g. of K 2 PtCl 6 . Calculate the percentage of Na 2 in the sample. 30. A solution of chloroplatinic acid contains 0.050 g. of Pt per cubic centimeter. What is the minimum volume with which 0.2602 g. of mixed sodium and potassium chlorides must be evaporated in order to insure the STOICHIOMETRY 169 complete conversion of the alkali metals into chloroplatinates, no matter in what proportions the two chlorides may exist in the mixture ? 31. A sample of phosphate rock contains 0.87% of moisture and 91.92% of calcium phosphate. Calculate the percentage of Caa(P04)2 which is present on the dry basis. 32. If 2.497 g. of a fertilizer containing 4.45% of moisture yields 0.3150 g. of Mg 2 P207j what is the percentage of P 2 05 on the dry basis? 33. Upon treatment with sulphuric acid, 1.430 g. of a salt yields 0.5952 g. of Na2S04 and 101.5 cc - of COz, measured moist at 17 C. and 757 mm. Calculate the percentages of Na 2 and C0 2 in the salt. (Tension of aqueous vapor at 17= 14.45 mm.) 34. What weight of water is present in one liter of air which is 50% saturated with moisture at 17 and 748 mm.? (See problem 33.) 35. If in the analysis of a substance an error of 0.1% is unavoidable, how accurately is it necessary to weigh a sample of 200 mg. ? A sample of five grams? (See p. 4.) 36. 1.3250 g. of pure Na 2 COs is dissolved in water and the solution made up accurately to 250.0 cc. A portion is carefully transferred without loss to a platinum dish by means of a pipette supposed to deliver 50.00 cc. of liquid. After evaporation with hydrochloric acid, and ignition, the sodium chloride residue is found to weigh 0.2927 g. What volume of this solution does the pipette actually deliver? 37. 0.7500 g. of a substance containing chlorine and bromine yields 0.5000 g. of Ag(Cl, Br). This mixture is heated in a current of chlorine, which converts the bromide of silver into the chloride, and the loss in weight due to this change is found to be 0.0683 g- Calculate the percentages of chlorine and bromine in the sample. 38. From the following data, calculate the percentage of each salt present in a mixture of sodium chloride, bromide, and iodide : Weight of sample, 0.1500 g. ; weight of precipitate obtained by distilling the solution with nitrous acid and converting the iodine into silver iodide, 0.1056 g. ; weight of silver chloride and bromide from the solution after removal of the iodine, 0.1784 g. ; weight of this precipitate after conversion of the whole into silver chloride, 0.1623 g. 39. A sample of baking powder, which is known to contain only NaHCOa and KHC^-jOe, in equivalent proportions, and starch, yields upon treat- ment with water 12.0% by weight of C0 2 . Calculate the percentage of each salt in the sample. 40. A salt containing barium, chlorine, and water of hydration gave upon analysis the following data: Weight of sample, i.oooo g. ; weight after heating (water driven off), 0.8522 g. ; weight of silver chloride ob- 170 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS tained, i . 1 73 5 g. ; weight of barium sulphate obtained, 0.9594 g. Calculate : (a) the percentage of each constituent ; (6) the formula of the compound. 41. If two i.oooo-g. samples of a substance containing 10.00% of MgO are weighed out, and the precipitate of MgNH 4 PO 4 . 6 H 2 is in one case contaminated with 0.0250 g. of Mg 3 (PO 4 )2, and in the other case with 0.0250 g. of Mg[(NH 4 ) 2 P04]2, what percentages of MgO will be found if the calculations are based upon the assumption that the ignited precipi- tate in each case consists entirely of Mg 2 P2O7? 42. The carbonates of calcium, strontium and barium obtained from a lo-liter sample of mineral water are converted into the anhydrous nitrates, and the calcium nitrate is extracted with absolute alcohol-ether mixture. The residue is dissolved in water, the barium separated from the strontium, as BaCr0 4 , and the strontium precipitated from the nitrate with sulphuric acid and alcohol. There are finally obtained 0.8507 g. of CaO, 0.1324 g. of SrS0 4 , and 0.1072 g. of BaCrO 4 ; calculate the content of the water in milligrams per liter (i.e. parts per million) of Ca, of Sr, and of Ba. 43. How many cubic centimeters of sulphuric acid of sp. gr. 1.840, con- taining 95.6% of H 2 SO 4 , must be added to i liter of sulphuric acid of sp. gr. 1.560, containing 65.1% of H 2 SO 4 , to obtain a solution containing 75.0% of H 2 SO 4 ? 44. A fuming sulphuric acid contains 25.5% of non-hydrated SO 3 . How many grams of 98.2% H 2 SO 4 must be added to 100 g. of the fuming acid to give a product containing 100% of H 2 SO 4 ? 45. A limestone contains 90.0% of CaCO 3 , 3.50% of MgC0 3 , 3.00% of CaSO 4 -2 H 2 0, 1.25% of FeCO 3 , and 2.25% of anhydrous siliceous material. What numerical difference would you expect to find between the loss on ignition and the true percentage of CO 2 ? 46. An ore contains 28.15% f nickel, and 0.5000 g. samples are taken for analysis. In one sample the element is determined by electrolysis, as metallic nickel, while in a second sample it is determined by means of dimethylglyoxime, as Ni(C 4 H 7 N 2 O 2 ) 2 . If the algebraic sum of the errors involved in each determination were equivalent to a negative error of 1.7 mg. of the substance finally weighed, how much greater would the percentage error be in the first determination than in the second? 47. An electric current is passed simultaneously through a series of three electrolytic cells which contain water acidified with sulphuric acid, an ammoniacal solution of nickel sulphate, and molten silver chloride. What is deposited upon the cathode in each of the other cells, and how many grams, in the time in which one liter of hydrogen, measured moist at 17 and 746 mm., is liberated from the water? (Tension of aqueous vapor at 17= 14.45 mm.) STOICHIOMETRY 171 48. From the following data, calculate the percentages of nickel and cobalt in the steel: Weight of sample, 1.124 g. ; weight of nickel and cobalt obtained upon electrolysis, 0.1246 g. ; weight of nickel dimethyl- glyoximine, Ni (C 4 H 7 N 2 O2)2, obtained from the electrolytic deposit, 0.4382 g. 49. A mass of platinum weighs 12.145 g- m au "j 11.580 g. in water, and 11.115 g. in sulphuric acid. What is the specific gravity of the platinum? Of the sulphuric acid? 50. A quantity of pure metallic silver weighing 1.0788 g. is dissolved in nitric acid and the solution made up to the mark in a measuring flask gradu- ated to contain 100.0 cc. Three portions are carefully transferred without loss to three separate beakers by means of a pipette known to deliver 25.00 cc. If the solution remaining in the flask, together with the liquid finally washed from the pipette, yields on analysis 0.3560 g. of AgCl, what volume of liquid does the flask actually contain? VOLUMETRIC ANALYSIS 51. If 25.00 cc. of hydrochloric acid yield 0.1435 g. of AgCl, what is the normality of the solution ? 52. If a 25.00 cc. portion of acid requires 21.50 cc. of 0.526 N alkali for neutralization, what is the normality of the acid? Supposing the acid to be HC1, what weight of silver chloride will 10.00 cc. of it yield with silver nitrate ? 53. If a 2.453 g. sample of pure anhydrous sodium carbonate requires 45.72 cc. of an acid for neutralization, and if 41.90 cc. of the acid requires 44.35 cc. of an alkali, what is the normality factor of each solution? 54. A sample of pure calcite, CaCOs, weighing 2.150 g. is dissolved in 50.00 cc. of an acid, and the excess of acid is neutralized with 29.12 cc. of an alkali of which 28.40 cc. require 7.10 cc. of the acid for neutralization. To what volume must one liter of the acid be diluted in order to make it exactly normal ? 55. How many cubic centimeters of 0.526 N acid will it take to neu- tralize the ammonia set free upon distilling 1.0378 g. of MgNH 4 PO4 . 6 H 2 O with an excess of caustic alkali? 56. If 15.25 cc. of alkali will neutralize 20.00 cc. of a solution containing 6.000 g. of KH 3 (C 2 O 4 )2 2 H 2 O in 250.0 cc., what is the normality factor of the alkali? 57. In the analysis of a feeding stuff, a Kjeldahl determination is carried out with a sample weighing 1.500 g. The ammonia is received in 25.00 cc. of 0.500 N acid, and the excess of acid is found to require 12.50 cc. of a standard alkali, of which 21.20 cc. will neutralize 18.00 cc. of the acid. Calculate the percentage of nitrogen in the sample. 172 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS 58. What weight of crude cream of tartar must be taken for titration in order that twice the number of cubic centimeters of 0.2000 N alkali required may numerically equal the percentage content of KH^H^Oe? 59. A sample of soda ash weighing 25.05 g. is dissolved and made up to 250.0 cc., and one fifth of this solution is taken for titration. What must be the normality of the standard acid (assuming the alkalinity to be due only to Na 2 C0 3 ) in order that the burette reading multiplied by two may indicate the percentage of Na 2 C0 3 in the sample? 60. A sample of caustic soda weighing 4.000 g. is dissolved in water and made up to one liter. A 100.0 cc. portion of this solution requires for neutralization 47.50 cc. of 0.2000 N acid. A second 100.0 cc. portion, after treatment with barium chloride in slight excess, is diluted to 200.0 cc. and allowed to settle, and 50.0 cc. of the clear solution require 11.50 cc. of the acid. Calculate the percentages of NaOH and Na 2 C0 3 in the sample. (Neglect the volume occupied by the solid precipitate.) 61. A sample of Solvay soda weighing 3.750 g. is dissolved in water and made up to one liter. A 100.0 cc. portion of this solution, titrated in the cold with o.i ooo N acid, with the use of phenolphthalein, is found to re- quire 29.95 cc - of the acid; the burette is then refilled and the titration completed at the boiling temperature of the solution, 35.15 cc. more of the acid being required. Calculate the percentages of Na 2 C0 3 and NaHC0 3 in the sample. (Under suitable experimental conditions, phenolphthalein becomes colorless in the cold as soon as the carbonate has been wholly converted into bicarbonate.) 62. A sample of sirupy phosphoric acid weighing 5.767 g. is dissolved in water and made up to one liter. A 100.0 cc. portion of the solution is treated with sodium acetate and silver nitrate in excess, whereby the phosphate is quan- titatively precipitated as AggPC^. Phenolphthalein is added to the filtrate and washings, and the solution titrated with 0.500 N alkali, of which 27.25 cc. are required. Calculate the percentage of H 3 P0 4 in the original sample. 63. A sample of Chili saltpeter weighing 1.025 g. is treated in sodium hydroxide solution with pulverized Devarda's alloy (50% Cu, 45% Al, 5% Zn), which reduces the nitrogen to ammonia; the ammonia is distilled into 25.00 cc. of 0.463 ^V acid, and the excess of acid requires 5.01 cc. of 0.212 N alkali for neutralization. Assuming that nitrogen was wholly present as NaN0 3 , calculate the percentage of the latter in the sample. 64. A sample of strontium nitrate weighing 10.53 g- is dissolved in water and made up to one liter. One tenth of this solution is distilled, in the presence of alkali, with an excess of titanous chloride, which reduces the nitrate to ammonia (KN0 3 +8 Ti(OH) 3 +6 H 2 0=KOH+8 Ti(OH) 4 +NH 3 ). The ammonia is received in 25.00 cc. of 0.500 N acid, and the excess of acid STOICHIOMETRY 173 requires 10.16 cc. of 0.250 N alkali. Assuming that the nitrogen was wholly present as Sr(N03)2, calculate the percentage of the latter in the sample. 65. How many grams of KH 3 (C20 4 )2 . 2 H 2 will it take to prepare one liter of a 0.500 N solution, to be used as a standard acid? How many to prepare one liter of a o.iooo N solution, to be used as a reducing agent in connection with potassium permanganate ? 66. How many grams of K 2 Cr 2 07 per liter will be required to prepare a solution of such strength that each cubic centimeter shall indicate 2.00% of iron, when a sample weighing 0.2792 g. is used for analysis? What is the normality factor of this solution ? 67. From the following data, calculate the percentage of iron in the ore: Weight of sample, 0.2186 g. ; the reduced iron solution requires for oxidation 25.14 cc. of 0.0996 N permanganate solution. 68. What is the maximum weight of an ore containing 70.00% of iron which can be taken for analysis without having to refill a 3o-cc. burette, if the permanganate solution is 0.1025 iV? 69. A calcium oxalate precipitate, obtained from 0.8432 g. of a rock, is decomposed with dilute sulphuric acid and made up to 250.0 cc. ; of this a 50.0 cc. portion is titrated with 0.1012 N permanganate solution. If 27.35 cc - of the latter are required, what percentage of CaO does the rock contain? 70. 11.56 cc. of nitric acid of sp. gr. 1.19 are diluted to 250.0 cc. ; 20.00 cc. of this solution are found to require 12.92 cc. of 0.410 N alkali. Calculate the percentage of HN0 3 in the acid of sp. gr. 1.19. 71. The Sb 2 S 3 precipitate obtained from the solution of an ore is dis- solved in sodium sulphide solution, and this is evaporated and fumed with an excess of sulphuric acid. The residue is then dissolved in dilute hydro- chloric acid, and the antimony oxidized from the trivalent to the penta- valent condition by means of a standard solution of permanganate. The sample of ore weighs 0.2749 g. and 24.17 cc. of 0.1025 N permanganate solution are used in the titration ; what is the percentage of antimony? 72. What is the normality factor of an acid, of which 25.37 cc - are equiva- lent to 1.263 g. of KN0 3 when the nitrogen of the latter is reduced in alkaline solution to ammonia and this is distilled off. and received in the acid solution? 73. From the following data, calculate the percentage of Mn0 2 in the ore: Weight of sample, 0.2000 g. ; after heating this in the presence of sulphuric acid with 50.00 cc. of o.iooo N oxalic acid, the excess of oxalic acid requires 8.50 cc. of o.iooo N permanganate solution. 74. From the following data, calculate the percentage of Mn0 2 in the ore: Weight of sample, 0.2400 g. ; this is boiled with hydrochloric acid and the distillate received in an excess of potassium iodide solution, 174 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS the liberated iodine requiring 25.51 cc. of 0.2000 ^V sodium thiosulphate solution. 75. A sample of mineral substance weighing i.ooo g. is taken for analysis. In the determination of the iron, the ferric solution is completely reduced by means of sulphurous acid, and the excess of the latter removed by the passage of carbon dioxide through the boiling solution; the iron then re- quires 28.17 cc - f o.iooo N permanganate solution. Calculate the per- centage of iron in the substance. 76. What weight of pyrolusite containing 86.50% of MnO 2 will oxidize the same amount of oxalic acid as 50.0 cc. of a permanganate solution, if 10.00 cc. of the latter will liberate 0.1905 g. of iodine from an acidified solu- tion of potassium iodide ? 77. What weight of Fe(NH 4 SO 4 ) 2 . 6 H 2 O will reduce 50.0 cc. of a permanganate solution, of which 10.00 cc. will liberate from an acidified solution of potassium iodide a quantity of iodine sufficient to react with 15.25 cc. of 0.1025 N thiosulphate solution? 78. A standard solution of permanganate will oxidize 0.00730 g. of ferrous iron per cubic centimeter ; what is the value of the same solution in terms of (a) H 2 C 2 O 4 . 2 H 2 ; (6) KNO 2 ; (c) H 2 O 2 ; (4 are required for one liter of tenth-normal solution, to be used as an oxidizing agent in acid solutions ? To be used in neutral solutions for the oxidation of manganese? 2. Why should a permanganate solution be allowed to stand for several days, and then be filtered through asbestos, before it is standardized? Why should it not be placed in burettes having rubber outlet tubes? 3. Name at least four substances which can be used to standardize permanganate solutions. Write an equation in each case. 4. What is the maximum weight of Na 2 C 2 04 which can be titrated with tenth-normal permanganate solution without having to refill a 30-cc. burette? 5. In the titration of oxalic acid, why is it that the oxidation proceeds so much more slowly at first than later on? Explain in full. 6. Name ten substances which can be quantitatively determined by means of potassium permanganate. 7. What effect does potassium permanganate have upon hydrochloric acid in the presence of ferrous salts, even in very dilute solution? How may this action be prevented ? 8. What are the components of the Zimmermann-Reinhardt solution? Explain the purpose of each. 9. Under what conditions can ferrous iron be accurately determined with potassium permanganate without the use of the Zimmermann-Rein- hardt solution? 10. What is the maximum weight of a sample of ore containing 40.00% of iron which can be taken for titration with tenth-normal oxidizing agent without having to refill a 3o-cc. burette ? 11. Discuss the preparation of a solution for analysis from a refractory iron ore. 12. Describe a method for the determination of calcium by means of potassium permanganate. 13. Outline a method for the determination of the oxidizing power of pyrolusite by means of oxalic acid and potassium permanganate. 14. How can the determination referred to in the preceding question be made by a gravimetric process ? The Determination of Phosphorus in Steel. 1. Assuming the presence of the phosphorus as FesP2, show by means of an equation the action of nitric acid upon this compound. Why is the nitric acid solution heated with potassium permanganate? 2. In order to cause the higher oxides of manganese, such as MnOgj to QUESTIONS 189 go into solution in nitric acid, what kind of a reagent should be added? Illustrate and explain. 3. What is the purpose of precipitating the phosphorus as ammonium phosphomolybdate ? 4. Why is the yellow precipitate dissolved in ammonia and the solution acidified with sulphuric acid, rather than to dissolve it directly in sulphuric acid? Why is Mo0 3 not precipitated when the ammoniacal solution is a cidified with sulphuric acid ? Would hydrochloric or nitric acid do as well here, and why? 5. Describe the construction and use of the Jones reductor. Can it be used for the reduction of substances other than molybdenum? 6. Why is it best to receive the reduced molybdenum solution below the surface of a solution containing ferric alum? What is the purpose of the phosphoric acid in this solution? 7. How should the permanganate solution be standardized hi order to obtain the most reliable results? 8. Would you recommend the determination of phosphoric anhydride in apatite by this method ? Why ? Is there a suitable volumetric method ? If so, describe it. The Determination of Manganese in an Ore. 1. What is the Guyard reaction ? What role does it play in the titration of oxalic acid or of iron with potassium permanganate? 2. Discuss the preparation of the solution for analysis from a refractory ore containing manganese. 3. What happens when zinc oxide is added to the acid solution of the ore (equations) ? 4. Explain fully why a zinc salt should be present in the solution during the titration. 5. Explain why the presence of chlorides does not interfere with the accuracy of this titration. 6. If the permanganate solution used in this titration is o.iooo-N for use with iron or oxalic acid, what is its normality factor for this reaction? 7. How is it best to standardize the permanganate solution used hi this determination ? lodometric Methods : The Preparation and Standardization of Iodine and Thiosulphate Solutions. 1. How many grams of iodine are required for one liter of the tenth- normal solution? Of sodium thiosulphate, Na 2 S 2 3 . 5 H 2 0? 2. How is the iodine solution made, and what is the purpose of the potassium iodide ? Show what equilibria exist in the iodine solution. Why IQO INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS should the water be freshly boiled and allowed to cool out of contact with the air, in the preparation of the thiosulphate solution ? 3. Write the reaction between iodine and sodium thiosulphate. How do chlorine and bromine differ from iodine in their behavior towards sodium thiosulphate? Explain why this is so. 4. What is the effect of free carbonic acid upon sodium thiosulphate solution? Does the decomposition cease as soon as all of the carbonic acid has reacted? Write equations to illustrate your answers. 5. Does the solution resulting from the partial decomposition of the thiosulphate have a greater or lower reducing value than the original solu- tion? Explain why. 6. Give equations to show two ways in which iodine may act as an oxidizing agent. 7. What is the maximum weight of As 2 0a which can be taken for reaction with tenth-normal iodine solution without having to refill a 30-cc. burette? 8. Can the standardization of iodine against arsenious oxide be per- formed in a strongly alkaline solution ? Can it be done in an acid solution ? Give reasons for your answers. 9. What is the purpose of the sodium bicarbonate ? Is the bicarbonate solution acid, alkaline, or neutral? Explain your answer. 10. Discuss the determination of the end-point. Explain why the indi- cator is added hi such large quantity. 11. Discuss the use of iodine solutions in the presence of sulphuric acid. In the presence of ammonium salts. The Determination of Antimony in Stibnite. 1. Write the reaction between pure stibnite and hydrochloric acid. 2. Why must the hydrochloric acid solution be heated on the steam bath ? Why must it not be boiled until after dilution ? 3. What is the purpose of adding tartaric acid to the solution ? Explain. 4. Explain why the solution may possibly turn red during gradual dilution. What is the correct procedure in such a case? 5. If a white precipitate forms upon dilution, what error has been made ? What is the white precipitate, and what should be done with the mixture? 6. What is the purpose of almost neutralizing the solution with sodium hydroxide, and how is this accomplished? Why is sodium bicarbonate then added in large excess ? 7. What elements would, if present, interfere with this determination, and why ? The Determination of Lead in an Ore. i. After the decomposition of the ore and the addition of sulphuric acid, why is it necessary to evaporate to white fumes? QUESTIONS 191 2. Explain the solubility of lead sulphate in ammonium acetate solution. 3. Explain why it is possible to quantitatively precipitate the lead from the ammonium acetate solution by means of an excess of potassium dichro- mate. 4. Explain the solubility of lead chromate in the acidified solution of sodium chloride. 5. Write an equation to show the reaction of the acid chromate solution with potassium iodide. 6. Why should the thiosulphate solution used in this determination be standardized under identical conditions against test lead ? The Determination of Copper in an Ore. 1. In this determination, why is the copper separated from the other metals present in the ore? Explain how iron, arsenic, or antimony would interfere with the accuracy of the titration. 2. What would you expect the composition of the precipitate to be which is formed upon the addition of sodium thiosulphate to a solution of copper sulphate? How does it happen, then, that we obtain cuprous sulphide ? 3. Can any other metals be precipitated from their salt solutions by means of sodium thiosulphate? (Try, for example, silver nitrate and sodium thiosulphate, in the cold, and explain what takes place.) 4. What is the object of igniting the precipitate of cuprous sulphide? What becomes of any antimony which is present ? 5. Why is it so important to standardize the thiosulphate solution against pure metallic copper? 6. Why must the nitrous fumes be completely expelled before the ad- dition of the potassium iodide ? 7. Why is it preferable to titrate the free iodine in the presence of acetic acid, rather than in the presence of sulphuric acid? Explain fully. 8. Write an equation to show the action of nitric acid upon cuprous iodide. Precipitation Methods : The Determination of Chlorine. 1. Briefly outline the procedure for the standardization of the silver nitrate and potassium thiocyanate solutions against pure sodium chloride. Could these solutions be standardized against pure metallic silver, and if so how ? 2. How may pure sodium chloride be prepared from the commercial salt ? 3. What indicator is used in connection with thiocyanate solutions? Why must nitric acid be present? Explain fully why the indicator should be added in such large quantity. INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS 4. At 18-20, the solubility product of silver chloride is about o.6Xio~ 10 and that of silver thiocyanate is about o.6Xio~ 12 ; what is the relative con- centration of the chloride and thiocyanate ions in a solution which is satu- rated with both salts? Assume the equal (practically complete) ioniza- tion of both salts. 5. Why is it necessary to filter off the silver chloride before making the titration with the thiocyanate solution? Base your explanation upon the data given in the preceding question. 6. In general, what anions may be determined by this method without first filtering off the silver salt? 7. How may the halogens in alkali chlorates, bromates, and iodates be determined by this method ? 8. Outline a procedure for the determination by this method of (a) the chlorine in horn silver, AgCl ; (6) the silver in the same mineral. APPENDIX THE PREPARATION OF THE REAGENTS MANY of the reagents used in quantitative analysis are pre- pared for one specific purpose only, and directions for the prepa- ration of such reagents will be found in the treatment- of the individual determinations. Certain reagents, however, are used at approximately fixed concentrations in a variety of processes, and it is especially these which are included in this section. There are many advantages in making the concentrations of the reagents used in analytical work follow a definite system. The most convenient system is to use multiples or submulti- ples of the equivalent weight employed in volumetric analysis, though of course the concentration of the solution need not be known with such exactitude as in volumetric work. With this system, equal volumes of the solutions bear fixed relations to one another, it is easy to calculate the volume of a reagent which is required for a specific purpose, and the addi- tion of an unnecessary excess may readily be avoided. This means a saving in time, labor, and material; and it leads to more accurate and reliable work. Thus, if it is directed to fuse one gram of a silicate, say KAlSi 3 8 , with 7.5 g. of Na 2 C0 3 and to take up the cooled mass with water and an excess of hydrochloric acid, we know that 24 cc. of 6-normal acid will suffice to neutralize the mixture (since 7.5 g. of Na 2 C0 3 corre- spond to about one seventh of an equivalent of this salt), and that 30 cc. in all will furnish a sufficient excess. If the mixture is evaporated to dryness in the regular manner and taken up with 2 cc. of i2-normal hydrochloric acid and a little water, it o 193 194 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS will be obvious that 4 cc. of 6-normal ammonium hydroxide will suffice to neutralize the acid ; and that 8 cc. in all will neutralize the acid, precipitate the aluminum, and in addition furnish a sufficient excess. Measuring cylinders and measuring pipettes are useful for delivering specific volumes of such reagents. ACIDS Acetic, 6-normal : Mix 350 cc. of glacial acetic acid with 650 cc. of water. Hydrochloric, i2-normal: Use the C. P. acid of commerce of sp. gr., 1.19. Hydrochloric, 6-normal : Mix 12 -normal acid with an equal vol- ume of water. The specific gravity of this acid is about i.io. Nitric, 1 6-normal: Use the C. P. acid of commerce of sp. gr., 1.42. Nitric, 6-normal : Mix 380 cc. of the 1 6-normal acid with 650 cc. of water. The specific gravity of this acid is about 1.195. Sulphuric, 3 6-normal : Use the C. P. acid of commerce of sp. gr., 1.84. Sulphuric, 6-normal: Pour 200 cc. of the 3 6-normal acid into 1045 cc. of water. The specific gravity of this acid is 1.18. BASES Ammonium hydroxide, i5-normal: Use the C. P. ammonia water of commerce of sp. gr., 0.90. Ammonium hydroxide, 6-normal: Mix 400 cc. of the 15- normal solution with 600 cc. of water. The specific gravity of this solution is about 0.958. Sodium hydroxide, 6-normal: Dissolve 250 g. of stick sodium hydroxide in water and dilute to one liter. SALTS Ammonium carbonate: Dissolve 250 g. of freshly powdered ammonium carbonate in one liter of 6-normal ammonium hy- droxide, and filter if there is a residue. APPENDIX 195 Ammonium mandate? Dissolve 100 g. of MoOs in 80 cc. of ammonia (sp. gr., 0.90) with the addition of 400 cc. of water ; with cooling and constant stirring, allow the clear solution to run slowly into a mixture of 400 cc. of nitric acid (sp. gr., 1.42) with 600 cc. of water, add 0.05 g. of rmcrocosmic salt, NaNH 4 HP0 4 . 4 H 2 O, and keep the mixture in a warm place for several days, or until a portion heated to 40 deposits no yellow precipitate. Decant from any sediment, and preserve in glass- stoppered bottles. This solution contains 68 g. of Mo0 3 per liter. Ammonium oxalate, o.5-normal: Dissolve 35.5 g. of (NH 4 )2C 2 O4 . H 2 in 1000 cc. of water. Barium chloride, i -normal: Dissolve 122 g. of BaCl 2 . 2 H 2 in 1000 cc. of water. Magnesia mixture, o.5-normal as a precipitant for phosphoric or arsenic acid: Dissolve 51 g. of MgCl 2 . 6 H 2 and 130 g. of NH 4 C1 in water, add 121 cc. of ammonia (sp. gr., 0.90), and dilute to one liter. Mercuric chloride, o.2-normal for oxidizing stannous chloride : Dissolve 54 g. of HgCl 2 in loco cc. of hot water. Silver nitrate, o.2-normal : Dissolve 34 g. of AgN0 3 in 1000 cc. of water. Sodium phosphate, o.5-normal as a precipitant for mag- nesium: Dissolve 90 g. of Na 2 HP0 4 . 12 H 2 (or 52 g. of NaNH 4 HP0 4 . 4 H 2 0) in 1000 cc. of water. Stannous chloride, i -normal as a reducing agent: Dissolve 113 g. of SnCl 2 . 2 H 2 in 150 cc. of i2-normal hydrochloric acid, with the gradual addition of water, finally diluting to one liter. Keep in bottles containing granulated tin. 1 Recovery of the Molybdk Acid. To the liquid molybdate residues, acidified if necessary with nitric acid, add sodium phosphate solution in excess. Collect the yellow precipitate, wash it with water containing sodium sulphate, and then dry it in the air. Dissolve i Kg. of the dried precipitate in ammonia, add a strong solution of 60 g. of NH 4 C1 and 120 g. of MgClz . 6 H 2 O in water, allow to stand for 6 hours, and filter off the precipitate. To the filtrate, decolorized if necessary with a little H 2 O 2 , add HC1 just to acid reaction, to precipitate the MoO 3 . Collect this precipitate, wash with water, and dry at 110. 196 INTRODUCTORY COURSE IN QUANTITATIVE ANALYSIS Sulphuric Acid-Dichr ornate Cleaning Solution With stirring, cautiously pour 200 cc. of sulphuric acid (sp. gr., 1.84) into 150 cc. of cold water, and saturate the hot solu- tion, without further heating, with powdered sodium (or potas- sium) dichromate. When cleaning measuring vessels with this liquid, they should be filled with the cold solution and allowed to stand overnight, or longer. Analytical Samples for the Use of Students The analyzed samples indicated in the text for the use of beginners in quantitative analysis may in some cases be ob- tained in the market. Otherwise, they may be prepared by mixing together the component materials in the proportions decided upon. This mixing is best accomplished by long con- tinued grinding in a ball mill, the material being finally passed through a fine-meshed sieve, and bottled. These samples should be carefully analyzed by members of the quantitative staff, so that the student's work may be judged according to its accuracy. Most of the mixtures indicated can be kept from year to year without change. It is desirable to have in each case a continuous series of at least ten samples, varying in content from sample to sample by about 0.4-0.5%. APPENDIX 197 APPARATUS IN THE STUDENT'S DESK 1 QUANTITATIVE CHEMICAL LABORATORIES Above in the drawers i Brush, camel's hair, i Burette, g. s., 30 cc. 1 Burette, 30 cc., for pinchcock. 4 Crucibles, porcelain, o. 2 Crucibles, porcelain, Gooch, extra disc. 2 Cylinders, graduated, 50 cc. and 10 cc. i File. 1 Forceps, steel. 2 Funnels, diam. 25 mm., stem 40 mm. 2 Glasses, watch, 140 mm. 2 Glasses, watch, 70 mm. 2 Glasses, watch, 50 mm. i Vial litmus paper, blue. 1 Vial litmus paper, red. 2 Boxes matches, safety, i Pinchcock. i Pipette, 25 cc. 1 Pipette, 10 cc. 2 Policemen, rubber tip. 3 Rods, glass, 200 mm. 6 Test tubes. i Thermometer, 100 C. i Tongs, brass, nickel plated. 1 Tube, connecting, 3-way. 2 Tubes, rubber, for Gooch crucibles. 3 Tubes, weighing, with corks. i Tube, rubber, pressure, length 300 mm. i Tube, rubber, small, length 300 mm. Tubing, soft glass, 900 mm. 1 The articles listed above represent the apparatus with which it is desirable to provide each student at the outset ; the list can of course be modified in many particulars without jeopardizing the success of the work. Any additional appa- ratus which may be required can be obtained as needed from the store room. Below in the cupboard 12 Beakers, 2 nests, 1-6, with win- dow pane. 2 Bottles, g. s., 2500 cc. i Bottle, g. s., 250 cc., for cleaning solution, i Bottle, g. s., 125 cc., for silver nitrate, i Bottle, weighing. 1 Burette holder, Lincoln's. 2 Burners, adjustable. 2 Burner tubes, rubber. 2 Casseroles, porcelain, 500 cc. 1 Desiccator for 4 crucibles. 2 Flasks, Erlenmeyer, 500 cc. 2 Flasks, Erlenmeyer, 250 cc. 2 Flasks, Erlenmeyer, 150 cc. 2 Flasks, filter, 500 cc. 2 Flasks, Florence, 500 cc. 2 Flasks, Florence, 250 cc. 2 Flasks, Florence, 50 cc., for in- dicators, i Flask, volumetric, 1000 cc. 1 Flask, volumetric, 500 cc. 2 Flasks, volumetric, 250 cc. 4 Funnels, diam. 70 mm., stem 200 mm. 2 Funnels, for Gooch crucibles, i Sponge. 1 Stand, filter, wooden. 2 Stands, iron, i ring each. 2 Triangles, pipe stem, new form. 2 Tripods, iron. 2 Wire gauzes. 198 Table A Four Place Logarithms N 1 2 3 4 5 6 7 8 9 123 456 789 10 0000 0043 0086 0128 0170 0212 0253 0294 0334 0374 4 812 17 21 25 29 33 37 11 12 13 0414 0792 1139 0453 0828 1173 0492 0864 1206 0531 0899 1239 0569 0934 1271 0607 0969 1303 0645 1004 1335 0682 1038 1367 0719 1072 1399 0755 1106 1430 4 811 3 710 3 610 15 19 23 14 17 21 13 16 19 263034 24 28 31 23 26 29 14 15 16 1461 1761 2041 1492 1790 2068 1523 1818 2095 1553 1847 2122 1584 1875 2148 1614 1903 2175 1644 1931 2201 1673 1959 2227 1703 1987 2253 1732 2014 2279 369 368 358 12 15 18 11 14 17 11 13 16 21 24 27 20 22 25 18 21 24 17 18 19 20| 2304 2553 2788 2330 2577 2810 2&55 2G01 2833 2380 2625 2856 2405 2648 2878 2430 2672 2900 2455 2695 2923 2480 2718 2945 2504 2742 2967 2529 2765 2989 257 257 247 10 12 15 9 1214 9 11 13 17 20 22 161921 101820 3010 3032 3243 3444 3636 3054 3263 3464 3655 3075 3096 3118 3139 3160 3181 3201 246 8 11 13 15 17 19 21 22 23 3222 3424 3617 3284 3483 3674 3304 3502 3692 3324 3522 3711 3345 3541 3729 3365 3560 3747 3385 3579 3766 3404 3598 3784 246 246 246 8 1012 8 10 U 7 9 11 14 16 18 141617 13 15 17 24 25 26 3802 3979 4150 3820 3997 4166 3838 4014 4183 3856 4031 4200 3874 4048 4216 3892 4065 4232 3909 4082 4249 3927 4099 4265 3945 4116 4281 3962 4133 4298 245 246 235 7 9 11 7 9 10 7 8 10 12 14 16 12 14 10 11 13 15 27 28 29 4314 4472 4624 4330 4487 4639 4786 4928 5065 5198 4346 4502 4654 4800 4942 5079 5211 4362 4518 4669 4378 4533 4683 4393 4548 4698 4409 4504 4713 4425 4579 4728 4440 4594 4742 4456 4609 4757 235 235 134 689 689 679 11 12 14 11 12 14 10 12 13 30 4771 4914 5051 5185 4814 4829 4969 5105 5237 4843 4983 5119 5250 4857 4997 5132 5263 4871 4886 4900 134 679 10 11 13 31 32 33 4955 5092 5224 5011 5145 5276 5024 5159 5289 5038 5172 6302 134 3 4 3 4 678 578 578 10 11 12 91112 91112 34 35 36 5315 5441 5563 5328 5453 5575 5340 5465 5587 5353 5478 5599 5366 5490 5611 5378 5502 5623 5391 5514 5635 5403 5527 5647 6416 5539 5658 5428 5551 6670 2 4 2 4 2 4 568 567 667 91011 91011 81011 37 38 39 5682 5798 5911 5694 5809 5922 5705 5821 5933 5717 5832 5944 5729 5843 5955 5740 5855 5966 5752 08C6 5977 5763 5877 50C8 5775 5888 5999 5786 5899 6010 2 4 2 3 123 567 667 457 8 911 8 910 8 910 40 6021 6031 6042 6053 6064 6075 6085 C096 6107 6117 1 2 3 456 8 910 41 42 43 6128 6232 6335 6138 62 13 6345 6149 6253 6355 6160 6263 6365 6170 6274 6375 6180 6284 C3S5 6191 6294 6395 6201 6304 6405 6212 6314 6415 6222 6325 6425 123 123 123 456 466 466 789 789 789 44 45 46 6435 6532 6628 6444 6542 6637 6454 6551 6646 6464 6561 6656 6474 6571 6665 6484 6580 6675 6493 6590 6684 6503 6599 6693 6513 6609 6702 6522 6618 6711: 123 123 123 456 456 456 789 789 778 47 48 49 6721 6812 6902 6730 6821 6911 6739 6830 6920 6749 6839 6928 6758 6848 6937 6767 6857 6946 6776 6866 6955 6785 6875 6964 6794 6884 6972 6803 6893 6981 123 123 123 456 456 445 778 778 678 j>P_ 51 52 53 6990 7076 7160 7243 6998 7084 7168 7251 7007 7016 7024 7033 7118 7202 7284 7042 7050 7059 7143 7226 7308 7067 7152 7235 7316 123 345 678 7093 7177 7259 7101 7185 7267 7110 7193 7275 7126 7210 7292 7135 7218 7300 123 123 122 346 345 345 678 677 667 54 7324 7332 7340 7348 7356 7364 7372 7380 7388 7396 122 345 667 N 1 2 3 4 5 6 7 8 9 122 456 789 The proportional parts are stated in full lor every tenth at the right-hand side. The logarithm of any number of four significant figures can be read directly by add- A] Table A Four Place Logarithms 199 IT 1 2 3 4 5 6 7 8 9 123 456 789 55 56 7404 7482 7412 7490 7419 7497 7427 7505 7435 7513 7443 7520 7451 7528 7459 7536 7466 7543 7474 7551 122 122 345 345 567 567 57 58 59 7559 7634 7709 7782 7853 7924 7993 7566 7642 7716 7789 7860 7931 8000 7574 7649 7723 7582 7657 7731 7803 7875 7945 8014 7589 7664 7738 7810 7882 7952 8021 7597 7672 7745 7604 7679 7752 7612 7686 7760 7619 7694 7767 7627 7701 7774 112 112 112 345 344 344 567 567 567 60 61 62 63 7796 7818 7825 7832 7839 7846 112 344 566 7868 7938 8007 7889 7959 8028 7896 7966 8035 7903 7973 8041 7910 7980 8048 7917 7987 8055 112 112 112 334 334 334 566 556 556 64 65 66 8062 8129 8195 8069 8136 8202 8075 8142 8209 8082 8149 8215 8089 8156 8222 8096 8162 8228 8102 8169 8235 8109 8176 8241 8116 8182 8248 8122 8189 8254 112 112 112 334 334 334 556 556 556 67 68 69 8261 8325 8388 8267 8331 8395 8274 8338 8401 8280 8344 8407 8287 8351 8414 8293 8357 8420 8299 8363 8426 8306 8370 8432 8312 8376 8439 8319 8382 8445 112 112 112 334 334 334 556 456 456 70 8451 8457 8463 8470 8476 8482 8488 8494 8500 8506 112 334 456 71 72 73 8513 8573 8633 8519 8579 8639 8525 8585 8645 8531 8591 8651 8537 8597 8657 8543 8603 8663 8549 8609 8669 8555 8615 8675 8561 8621 8681 8567 8H27 8686 112 112 112 334 334 2*4 456 456 455 74 75 76 8692 8751 8808 8698 8756 8814 8704 8762 8820 8710 8768 8825 8716 8774 8831 8722 8779 8837 8727 8785 8842 8733 8791 8848 8739 8797 8854 8745 8802 8859 112 112 112 234 233 233 465 455 445 77 78 79 80 8865 8921 8976 8871 8927 8982 8876 8932 8987 8882 8938 8993 8887 8<)43 8998 8893 8949 9004 8899 8954 9009 8904 89(50 9015 8910 8965 9020 8915 8971 9025 1 1 2 112 112 233 233 233 445 445 445 9031 9036 9042 9047 9053 9058 9063 9069 9074 9079 1 1 2 233 445 81 82 83 9085 9138 9191 9090 9143 9196 9096 9149 9201 9101 9154 9206 9106 9159 9212 9112 9165 9217 9117 9170 9222 9122 9175 9227 9128 9180 9232 9133 9186 9238 1 1 2 112 112 233 233 233 445 445 445 84 85 86 9243 9294 9345 9248 92!)9 9350 9253 9304 9355 9258 9<09 9360 9263 9315 9365 9269 9320 9370 9274 9325 9375 9279 9330 9380 9284 9335 9385 9289 9340 9390 112 112 112 233 233 233 445 445 445 87 88 89 9395 9445 9494 9400 9450 949!) 9405 9455 9504 9410 9460 9509 9415 9465 9513 9420 9469 9518 9425 9474 9523 9430 9479 9528 9435 9484 9533 9440 9489 9538 112 Oil 1 1 233 223 223 445 344 344 90 9542 9547 9552 9557 9562 9566 9571 9576 9581 9586 Oil 223 344 91 92 93 9590 9638 9685 9595 9643 9689 9600 9647 9694 9605 9652 9699 9609 9657 9703 9614 9661 9708 9619 9666 9713 9624 9671 9717 9628 9675 9722 9633 9680 9727 Oil Oil Oil 223 223 223 3 4 3 4 3 4 94 95 96 9731 9777 9823 9736 9782 9827 9741 9786 9832 9745 9791 9836 9750 9795 9841 9754 9800 9845 9759 9805 9850 9763 9809 9854 9768 9814 9859 9773 9818 9863 Oil Oil Oil 223 223 223 3 4 3 4 3 4 97 98 99 9868 9912 9956 9872 9917 9961 9877 9921 9965 9881 9926 9969 9886 9930 9974 9890 9934 9978 9894 9939 9983 9899 9943 9987 9903 9948 9991 9908 9952 9996 Oil Oil Oil 223 223 223 3 4 4 334 334 N 1 2 3 4 5 6 7 8 9 123 456 789 ing the proportional part corresponding to the fourth figure to the tabular numbei corresponding to the first three figures. There may be an error of 1 in the last place. 200 Table B Antilogarithms to Four Places [B 1 2 8 4 5 6 7 8 9 123 456 789 .00 1000 1002 1005 1007 1009 1012 1014 1016 1019 1021 001 111 222 .01 .02 .03 1023 1047 1072 1026 1050 1074 1028 1052 1076 1030 1054 1079 1033 1057 1081 1035 1059 1084 1038 1062 1086 1040 1064 1089 1042 1067 1091 1045 1069 1094 001 001 001 111 1 1 1 1 222 222 222 .04 .05 .06 .07 .08 .09 .10 .11 .12 .13 1096 1122 1148 1175 1202 1230 1099 1125 1151 1178 1205 1233 1102 1127 1153 1180 1208 1236 1104 1130 1156 1183 1211 1239 1107 1132 1159 1186 1213 1242 1109 1135 1161 1189 1216 1245 1112 1138 1164 1191 1219 1247 1114 1140 1167 1194 1222 1250 1117 1143 1169 1197 1225 1253 1119 1146 1172 1199 1227 1256 Oil Oil Oil 1 2 1 2 1 2 222 222 222 222 223 223 Oil Oil 112 112 1259 1262 1291 1321 1352 1265 1268 1297 1327 135S 1271 1274 1276 1279 1282 1285 Oil 112 223 1288 1318 1349 1294 1324 1355 1300 1330 1361 1303 1334 1365 1306 1337 1368 1309 1340 1371 1312 1343 1374 1315 1346 1377 Oil Oil Oil 122 122 122 223 223 233 .14 .15 .16 1380 1413 1445 1384 1416 1449 1387 1419 1452 1390 1422 1455 1393 1423 1439 1396 1429 1462 1400 1432 1466 1403 1435 1469 1406 1439 1472 1409 1442 1476 Oil 1 1 Oil 122 122 122 233 233 233 .17 .18 .19 1479 1514 1549 1483 1517 1552 1486 1521 1556 1489 1524 1560 1493 1528 1563 1496 1531 1567 1500 1535 1570 1503 1538 1574 1507 1542 1578 1510 1545 1581 Oil Oil 1 1 122 122 1 2 2 233 233 233 .20 .21 .22 .23 1585 1622 1660 1698 1589 1626 1663 1702 1592 1629 1667 1706 1590 1GOD 1003 1607 1611 1614 1618 1 1 1 2 2 333 1633 1071 1710 1637 1675 1714 1641 1679 1718 1644 1683 1722 1648 1687 1726 1052 1690 1730 1656 1694 1734 Oil Oil Oil 122 222 222 333 333 333 .24 .25 .26 1738 1778 1820 1742 1782 1824 1740 1780 1828 1759 1791 1832 1754 1795 1837 1758 1799 1841 1762 1803 1845 1766 1807 1849 1770 1811 1854 1774 1816 1858 Oil Oil Oil 222 223 223 334 334 334 .27 .28 .29 1862 1905 1950 1866 1910 1954 1871 1914 1959 1875 1919 1963 1879 1923 1968 1881 1928 1972 1888 1932 1977 1892 1936 1982 1897 1941 1986 1901 1945 1991 Oil Oil Oil 223 223 223 334 344 344 .30 .31 .32 .33 1995 2000 2004 2009 2014 2018 2023 2028 2032 2037 Oil 223 344 2042 2089 2138 2046 2094 2143 2051 2099 2148 2056 2104 2153 2061 2109 2158 2065 2U3 2163 2070 2118 2168 2075 2123 2173 2080 2128 2176 2084 2133 2183 Oil Oil Oil 223 223 223 344 344 344 .34 .35 .36 2188 223!) 2291 2193 2244 2296 2198 22i9 2301 2203 2254 2307 2208 2259 2312 2213 2235 2317 2218 2270 2323 2223 2275 2328 2228 2280 2333 2234 2286 2339 112 112 1 1 2 233 233 233 445 445 445 .37 .38 .39 .40 2344 2399 2455 2350 2404 24(50 2355 2410 2466 2360 2415 2472 2366 2421 2477 2371 2427 2483 2377 2432 2489 2382 2438 2495 2388 2443 2500 2393 2449 2506 112 112 112 233 233 233 445 455 455 2')12 2518 2523 2529 2535 2594 2655 2716 2541 2547 2553 2559 2564 112 234 455 .41 .42 .43 2570 2630 2692 2576 2636 2698 2582 2612 2704 2588 2649 2710 2600 2661 2723 2606 2667 2729 2612 2673 2735 2618 2679 2742 2624 2685 2748 112 112 112 234 234 234 456 456 456 .44 .45 .46 2754 2818 2884 2761 2825 2891 2767 2831 2897 2773 2838 2904 2780 2844 2911 2786 2851 2917 2793 2858 2924 2799 2864 2931 2805 2871 2938 2812 2877 2944 112 112 112 334 334 334 456 556 556 .47 .48 .49 2951 3020 3090 2958 3027 3097 2965 3034 3105 2972 3041 3112 2979 3048 3119 2985 3055 3126 2992 3062 3133 2999 3069 3141 3006 3076 3148 3013 3083 3155 112 112 112 334 334 344 566 566 566 B] Table B Antilogarithms to Four Places 20 1 1 2 3 4 5 6 7 8 9 123 456 789 .50 3162 3170 3177 3184 3192 3199 3206 3214 3221 3228 567 .51 .52 .53 3236 3311 3388 3243 3319 3396 3251 3327 3404 3258 33M 3412 3266 3342 3420 3273 3350 3428 3281 3357 3436 3289 3365 3443 3296 3373 3451 3304 3381 3459 112 1 1 2 122 344 345 345 567 567 667 .54 .55 .56 3467 3548 3631 3475 3556 3639 3483 3565 3648 3491 3573 3656 3499 3581 3664 3508 3589 3673 3516 3597 3681 3524 3606 3690 #532 3614 3698 3540 3622 3707 122 122 122 345 345 345 667 677 678 .57 .58 .59 3715 3802 3890 3724 3811 3899 3733 3819 3908 3741 3828 3917 3750 3837 3926 3758 3846 3936 3767 3855 3945 3776 3364 3954 3784 3873 3963 3793 3882 3972 123 123 123 345 345 455 678 678 678 .60 .61 .62 .63 3981 3990 3999 4009 4018 4027 4036 4046 4055 4064 123 456 788 4074 4169 426(i 4083 4178 4276 4093 4188 4285 4102 4198 4295 4111 4207 4305 4121 4217 4315 4130 4227 4325 4140 4236 4335 4150 4246 4345 4159 42f>6 4355 123 123 123 456 456 456 789 789 789 .64 .65 .66 43fi5 4467 4571 4375 4477 4581 4385 4487 4592 4395 4498 4603 4406 4508 4613 4416 4519 4624 4426 452!) 4634 4436 4539 4645 4446 4550 4656 4457 4560 4667 123 123 123 456 456 456 789 789 7 910 .67 .68 .69 .70 4677 4786 4898 4688 4797 4909 5023 4699 4808 4920 4710 4819 4932 4721 4831 4943 4732 4842 4955 4742 4853 4966 4753 4864 4977 4764 4875 4989 4775 4887 5000 123 123 123 457 567 567 8 910 8 910 8 910 5012 5035 5047 5058 5070 5082 5093 5105 5117 123 567 8 910 .71 .72 .73 5129 5248 5370 5140 5260 5383 5152 5272 5395 5164 5284 5408 5176 5297 5420 5188 5309 5433 5200 5321 5445 5212 5333 5458 5224 5346 5470 5236 5358 5483 124 1 2 4 134 567 567 567 81011 91011 91011 .74 .75 .76 5495 5623 5754 5508 5636 5768 5521 5649 5781 5534 5662 5794 5546 5675 5808 5559 5689 5821 5572 5702 5834 5585 5715 5848 5598 5728 5861 5610 5741 5875 134 134 134 568 578 578 91012 911 12 91112 .77 .78 .79 5888 6026 6166 5902 6039 6180 5916 6053 6194 5929 6067 6209 5943 6081 6223 5957 60! )5 6237 5970 6109 6252 5984 6124 626(5 5998 6138 6281 6012 6152 6295 134 134 134 578 678 679 10 11 12 10 11 13 10 11 13 .80 6310 6324 6339 6353 6368 6383 6397 6412 6427 6442 134 679 10 12 13 .81 .82 .83 6457 6607 6761 6471 6622 6776 6486 6637 6792 6501 6653 6808 6516 6668 6823 6531 6683 6839 6546 6699 6855 6561 6714 6871 6577 6730 6887 6592 6745 6902 235 235 235 689 689 689 11 1214 11 12 14 11 13 14 .84 .85 .86 6918 7079 7244 6934 7096 7261 6950 7112 7278 6966 7129 7295 6982 7145 7311 6998 7161 7328 7015 7178 7345 7031 7194 7362 7047 7211 7379 7063 7228 7396 235 235 235 7 810 7 810 7 810 11 13 15 12 13 15 12 14 15 .87 .88 .89 .90 .91 .92 .93 7413 7586 7762 7943 8128 8318 8511 7430 7603 7780 7962 8147 8337 8531 7447 7621 7798 7980 8166 8356 8551 7464 7(538 7816 7482 7656 7834 8017 7499 7674 7852 8035 7516 7691 7870 80.54 7534 7709 7889 8072 7551 7727 7907 8091 7568 7745 7925 8110 245 245 246 7 910 7 911 7 911 12 14 16 12 14 16 13 15 1(5 7998 8185 8375 8570 246 7 911 13 15 17 8204 8395 8590 8222 8414 8610 8241 8433 8630 8260 8453 8650 8279 8472 8670 8299 8492 8690 246 246 246 8 911 81012 81012 13 15 17 14 15 17 14 16 18 .94 .95 .96 8710 8913 9120 8730 8933 9141 8750 8954 9162 8770 8974 9183 8790 8995 9204 8810 9010 9226 8831 90136 9247 8851 9057 9268 8872 9078 9290 8892 9099 9311 246 246 246 81012 8 10 12 911 13 14 16 18 15 17 19 15 17 19 .97 .98 .99 9333 9550 9772 9354 9572 9795 9376 9594 9817 9397 9616 9840 9419 96.38 9863 9441 9661 9886 9462 9683 9908 9484 9705 9931 9506 9727 9954 9528 9750 9977 246 247 257 911 13 91113 91114 15 17 19 16 18 20 16 18 21 INDEX Acidimetry 105, 115 Acids, degree of ionization of . . 24 determination of the neutraliza- tion value of 115 standard solutions of . . . . 105 titration of 108 Accuracy 3> 4 Adsorption 22, 30 Affinity constant 25 Afterflow, error from 45 Alkalimeter, Mohr's 78 Alkalimetry 105, 108, 113 Alkali solutions, standard ... 105 Aluminum, determination of . . 62 Ammonium thiocyanate, standard solutions of 156 Ampere, definition of 90 Analyzed chemicals 6 Antilogarithms 200 Antimony, determination of in stibnite 148 Apparatus, list of for quantitative work 197 Arsenic, removal of from copper 153 Arsenious oxide, primary standard 147 Asbestos, preparation of for niters 33 use of in nitration . . Ashless filter papers . . Atomic weights, table of - 33, 34 . - 30 Back cover sheet Baking powder, determination of carbon dioxide in .... 79 Balance, analytical 7 adjustment of 9 conditions to be fulfilled by . . 8 exercises with S3 location of 9 relative length of arms of . . 17 sensitivity of 13 use and care of 9 zero-point of 1 1 Barium sulphate, properties of 63, 64 Bases, degree cf ionization of . . 24 standard solutions of . . . . 105 titration of 108 Bumping 42 Buoyancy, correction for ... 18 Burettes 44 calibration of 49 cleaning of 46 reading of 46 Burning filter papers 38 Calcium, determination cf in lime- stone 70, 133 oxalate, properties cf . . . 73, 74 Calibration of measuring vessels . 49 of weights 14 Carbon dioxide, determination of in limestone 76 determination of in baking powders 79 Chemical equilibrium ... 23, 24 Chemical factors .... 161, 162 Chlorine, determination of 54, 59, 158 Chrome iron ore, determination of chromium in 125 Chromic oxide, ignition of ... 63 Chromium, determination of . 62, 125 Cleaning solution, preparation and use of 196 Colloidal precipitates . .... 21 Common-ion effect 26 Contamination of precipitates . . 22 Contat-Gockel valve 135 Copper, determination of . . 86, 152 Counterpoise, use of in weighing bulky objects 18, 80 Crucibles, materials of .... 40 Current density 94 Current, production of for electro- analysis 90 Decantation 32 Decimal, number of places to report 5 203 204 INDEX Deposition voltages of elements, table of 92 Desiccators , ;'. . 39 Dichromate processes .... 120 solutions, standard 121 Dichromate-sulphuric acid clean- ing solution 196 Digestion of precipitates ... 21 Distilled water, testing of ... 6 Double precipitation 23 Drainage, error from 45 Drying ovens 37 Economy of time 4 Electro-analysis 86 Electrode potentials, table of . . 92 Electrodes, material and form of . 95 Electrolytes, influence of composi- tion of upon electro-analysis . 94 Electrolytic separations .... 92 Electrolytic solution tension . . 91 End-point in titration, determina- tion of loo Equilibrium, chemical . . . 23, 24 Equilibrium constant .... 25 Evaporation of liquids .... 41 Factor, chemical .... 161, 162 normality 164 Faraday's laws . ....... 90 Ferric alum, indicator . . 156, 157 Ferric oxide, ignition of .... 62 Ferrous ammonium sulphate, standard solutions of . . . 121 Fertilizers, determination of nitro- gen in 116 Filters, selection and use of . . 30 Filtrates, testing of 32 Filtration 30 Fine-grained precipitates, enlarge- ment of the particles of . . 21 Flasks, volumetric ..... 44 calibration of 49 Flocculation of colloids .... 22 Funnels, selection of 31 Fusions, removal of from crucibles 81 Gelatinous precipitates .... 32 Gooch filters, preparation and use of 33, 34 sources of error with .... 35 Graduated cylinders 45 Gravimetric analysis i, 53 Guyard's reaction 139 Halogens, determination of ... 59 Hematite, determination of iron in 130 Hydrochloric acid, standard solu- tions of 109 Ignition of precipitates .... 37 Indicators 100, 107, 120 general theory of 101 sensitiveness of in alkalimetry and acidimetry .... 105, 107 Indirect methods of analysis . . 164 Insoluble matter, determination of in limestone 71, 72 International atomic weights, table of ... Back cover sheet Iodine, standard solutions of . 146, 147 lodometric processes 142 lonization, degree of 24 repression of 24 Ions, complex 24 composition of the 24 Iron, determination of .... 59 oxidation of ferrous to ferric 61, 122 reduction of ferric to ferrous 62, 123, 132, 137 Iron ores, decomposition of 124, 131, 132 Iron wire, primary standard . . 123 Jones reductor, assembly and use of 137 Kjeldahl, determination of pro- tein nitrogen 116 Labels 4 Lead, determination of in an ore 150 Limestone, analysis of . . 70, 76, 133 determination of carbon dioxide in 76 Liquids, evaporation of .... 41 transference of . . . fc . . 42 volumetric measurement of . . 43 Liter, Mohr's ....... 48 normal 48 true 48 INDEX 205 Limits of error in experimental work 4 Logarithms 198 Magnesium, determination of in limestone 70, 75 Magnesium ammonium phos- phate, ignition of ... 68, 70 Manganese, determination of in an ore . 139 Manganese ores, decomposition of 141 Methyl orange solution .... 109 Neatness 3 Neutral solution, definition of . . 106 Neutralization methods . . . 105, 108 Nickel, electrolytic determination of 89 Nitrogen, Kjeldahl determination of . . 116 Normality factor, definition of .164 Normal solutions 98 Normal System of Reagents, ad- vantages of 193 Notebooks 4 Ohm, definition of 90 Ohm's law 90 Ovens, drying 37 Overvoltage 92 Oxidation and reduction methods 119 Oxidizing agents, standard solu- tions of 119 Parallax, error from 46 Permanganate processes . . . . 127 Permanganate solutions, standard 128 Phenolphthalein solution . . . 109 Phosphoric anhydride, determina- tion of 66 Phosphorus, determination of in steel 135 Pipettes, transfer 44 calibration of 49 Platinum ware, defects of mod- ern 41 specifications for 41 use and care of 41, 87 Polarity of terminals, determina- tion of 87 Polarization 92 Policeman, definition of .... 32 Potash, determination of ... 84 Potassium bitartrate, primary standard 113 dichromate, standard solutions of 121 ferricyanide, indicator . .120,121 iodate, primary standard 143, 148 permanganate, standard solu- tions of 128 thiocyanate, standard solutions of .156 Precipitates, colloidal .... 21 contamination of 22 digestion of 21 drying of 37 enlargement of the particles of . 21 fine-grained 21 flocculation of 22 for use in gravimetric analysis 20 ignition of 37 purification of 23 washing of 22, 30 Precipitation 20 theory of 23 volumetric methods of ... 155 Problems 166 Pyrolusite, oxidizing value of . . 134 Questions 178 Reaction, Guyard's 139 Reactions suitable for use in volu- metric analysis 99 Reagents, analyzed 6 preparation of 193 quality of 6 testing of 6 Records 4, 5 Reducing agents, standard solu- tions of 119 Reduction, methods of oxidation and 119 Reductor, Jones . . . . . . 137 Reversible reactions ..... 23 Salts, degree of ionization of . . 24 Samples, Preparation of, for An- alysis 51, 196 Saturated solution, definition of . 28 Sensitivity, of balance .... 13 of indicators, table of the . . 107 Siderite, determination of iron in 124 Silica, determination of .... 80 2O6 INDEX Silicates, determination of silica in refractory 80 Silicic acid, dehydration of ... 83 Silver, determination of . 59, 155, 158 Silver, primary standard . . . 157 Silver chloride, properties of . 57, 58 Silver ion, properties of .... 58 Silver nitrate, standard solutions of 156 Soda ash, alkaline value of . . . 113 Sodium carbonate, primary stan- dard 109, 113 Sodium chloride, determination of chlorine in 54 primary standard 157 purification of 157 Sodium hydroxide, standard solu- tions of 109 Sodium oxalate, primary standard 129 Sodium thiosulphate, standard solutions of ... 146, 151, 152 Solubility, effect of size of par- ticles on 21 Solubility product 27 Solution tension 27 electrolytic 91 Solution of iron ores . . 124, 131, 132 Solution of manganese ores . . 141 Standardization, definition of . . 97 of hydrochloric acid . . . . 109 of sodium hydroxide solution . 109 of dichromate solution . . . 121 of ferrous ammonium sulphate solution 121 of permanganate solution . . 128 of iodine solution 146 of sodium thiosulphate solu- tion 146, 151, 152 of silver nitrate solution . . . 156 of thiocyanate solution . . . 156 Standard solution, definition of . 97 Search, indicator . . . . 144, 145 Stibnite, determination of anti- mony in 148 determination of sulphur in . . 163 Stoichiometry 159 Suction, Mse of 31, 33, 34 Sulphur, determination of . . 63, 65 Temperature, correction for dif- ferences in 46 Tension, solution . . .... . 27 electrolytic solution .... 91 Testing for complete precipitation 32 of washings 32 Thiocyanate solutions, standard . 156 Titration, definition of .... 2 Transference of liquids .... 42 Transfer pipettes 44 Triangles . 38, 41 Vacuum, use of .... 31, 33, 34 Valve, Contat-Gockel . . . . 135 Volt, definition of 90 Volume, units of ...... 48 Volumetric analysis, general dis- cussion 97 neutralization methods of . . 105 oxidation and reduction methods of 119 precipitation methods of ... 155 reactions suitable for .... 99 Volumetric apparatus .... 44 calibration of 49 cleaning of 46 necessary precautions in the use of 45 Volumetric System, Advantages of ^ ^. 43, 103 Volumetric Work, General Direc- tions .... 103 Wash bottles 33 Washing of precipitates . . 22, 30, 32 theory of 35 Washings, testing of . . .' . . 32 Water, ionization of . . . . . 106 Weighing , . 7, 53 limits of error in 1 1 methods of 12 summary of 20 Weights, calibration of .... 14 use and care of 10 Zero-point of balance, determina- tion of ii Zimmermann-Reinhardt solution . 131 Printed in the United States of America. UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. ENGINEERING LIBRA MAR 24 LD 21-100m-7,'52(A2528sl6)476 YC o UNIVERSITY OF CALIFORNIA DEPARTMENT OF CIVIL ENGINEERING f PRi SLEY. C* IFORNIA INTERNATIONAL ATOMIC WEIGHTS, 1917 Aluminum Al 27.1 Molybdenum Mo 96.0 Antimony Sb 1 20. 2 Neodymium Nd 144-3 Argon A 39-88 Neon Ne 20.2 Arsenic As 74.96 Nickel Ni 58.68 Barium Ba 137-37 Niton Nt 222-4 Bismuth Bi 208.0 Nitrogen N I4.OI Boron B II.O Osmium Os I9I.9 Bromine Br 79.92 Oxygen I6.OOO Cadmium Cd 112.40 Palladium Pd 106.7 Caesium Cs 132.81 Phosphorus P 31.04 Calcium Ca 40.07 Platinum Pt 195.2 Carbon C 12.005 Potassium K 39.10 Cerium Ce 140.25 Praseodymium Pr 140.9 Chlorine Cl 35.46 Radium Ra 226.0 Chromium Cr 52.0 Rhodium Rh IO2-9 Cobalt Co 58.97 R. oidium Rb 8545 Columbium Cb 93-i Rii ' henium Ru IOI-7 Copper Cu 63-57 S;. iiiarium Sa 150.4 Dysprosium Dy 162.5 Scandium Sc 44.1 Erbium Er 167.7 Selenium . Se 79.2 Europium Eu 152.0 Silicon Si 28.3 Fluorine F 19.0 Silver Ag IO/.88 Gadolinium Gd 157.3 Sodium Na 23.00 Gallium Ga 69.9 Strontium Sr 87.63 Germanium Ge 72-5 Sulphur S 32.06 Glucinum Gl 9.1 Tantalum Ta l8l.5 Gold Au 197.2 Tellurium Te 127-5 Helium He 4.00 Terbium Tb 159.2 Holmium Ho 163-5 Thallium Tl 204.0 Hydrogen H 1.008 Thorium Th 232.4 Indium In 114.8 Thulium Tm 168.5 Iodine I 126.92 Tin Sn II8.7 Iridium Ir I93-I Titanium Ti 48.1 Iron Fe 55.84 Tungsten W 184.0 Krypton Kr 82.92 Uranium U 238.2 Lanthanum La 139.0 Vanadium V 51.0 Lead Pb 207.20 Xenon Xe 130.2 Lithium Li 6.94 Ytterbium Yb 173-5 Lutetium Lu 175-0 Yttrium Yt 88.7 Magnesium Mg 24.32 Zinc Zn 65.37 Manganese Mn 54-93 Zirconium Zr 9O.6 Mercury Hg 200.6