EXCHANGE BOBl'12 W1W SILICIC ACID, ITS INFLUENCE AND REMOVAL IN WATER PURIFICATION BY OTTO MITCHELL SMITH B. S. Drury College, 1907 M. S. University of Illinois, 1918 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS 1919 [Printed by authority of the State of Illinois.] SILICIC ACID, ITS INFLUENCE AND REMOVAL IN WATER PURIFICATION BY OTTO MITCHELL SMITH B. S. Drury College, 1907 M. S. University of Illinois, 1918 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of DOCTOR OF PHILOSOPHY IN CHEMISTRY IN . THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS 1919 [Printed by authority of the State of Illinois.] -n ACKNOWLEDGMENT. ^ The writer wishes to express his gratitude to Professor W. F. Mon- jflrt and Lieutenant-colonel Edward Bartow, under whose direction this investigation was made, for their interest and many helpful suggestions during the progress of this work and the preparation of the manuscript of this thesis. He also desires to thank Mr. G. C. Habermeyer, mem- bers of the University faculty of the Department of Chemistry and the Water Survey staff for their kind assistance. SPRINGFIELD, ILL. ILLINOIS STATE JOURNAL Co., STATE PRINTERS. 1921 34613100 CONTENTS. PAGE. Acknowledgment 2 Introduction 5 Theoretical and Historical Part 6 The distinguishing characteristic of colloids 6 Properties of silicic acid 7 The properties of concentrated clay suspensions '8 Ionic reactions 1 / 8 Effect of electrolytes 9 Methods of Experimentation, Materials and Apparatus 10 Reagents 11 Determination of the hydrogen ion concentration 1 Measurement of turbidity 1.. Experimental Part 12 Measurement of colloidal substance 12 The influence of electrolytes on colloidal silicic acid 13 Optimum conditions for precipitation 17 Influence of hydrogen ion concentration 20 Coagulation of clay suspensions 21 The effects of electrolytes on the stability of clay suspensions. ... 21 The effect of electrolytes on the coagulation of clay suspension by aluminium sulf ate 21 The influence of silicic acid in the coagulation of clay suspension. 21 Applications to the Purification of Water Removal of silicic acid 43 The coagulation of clay-bearing surface water 43 Summary . . ; 44 References ' 45 Vita . 47 444300 ILLUSTRATIONS. Comparison between the pH, the precipitation of silicic acid and the amount of A1(OH) 3 in the solid phase 19 Comparison between the pH value, the precipitation of silicic acid and the amount of A1(OH) 3 in the solid phase *. 19 Effect of NaOH on the coagulation of a clay suspension by A1 2 (S0 4 ) 3 . . . . 26 Comparison of A1 2 (SO 4 ) 3 , colloidal Fe, Ca(OH) 2 , Ba(OH) 2 as coagulants.. 26 Effect of H 2 SO 4 on the coagulation of clay suspensions by AL(SO 4 ), 27 Effect of electrolytes on the coagulation and settling of clay suspension. . 27 Effect of electrolytes on the coagulation and settling of clay suspension. . 27 Effect of the addition of one millequivalent of electrolytes on the coagula- tion of clay suspension by A1 2 (SO 4 ) 3 28 Effect of Na 2 CO 3 on the coagulation of clay suspensions by A1 2 (SO 4 ) 3 28 Effect of NaHC0 3 on the coagulation of clay suspension by AL(SO 4 ) 3 29 Effect of Na 2 SO 4 and NaCl on the coagulation of clay suspension by A1 2 (SO 4 ) 3 30 Effect of Mg(HCO 3 ) 2 and of CaCL on the coagulation of clay suspension by AL(S0 4 ) 8 31 Effect of NaHCO 3 on the coagulation of clay suspension by A1 2 (S0 4 ) 3 - 32 The effect of Mg(HCO 3 ) 2 on the coagulation of clay suspensions by A1 2 (S0 4 ) 3 33 The effect of CaCl 2 on the coagulation of clay suspensions by A1 2 (SO 4 ) 3 . . . 34 Effect of Ca(HCO 3 ) 2 on the coagulation of clay suspension by A1,(SO 4 ) 3 . . . 35 Amount of aluminium necessary in the presence of silicic acid to produce a definite clarification and effect of silicic acid on coagulation of clay suspension by AL(S0 4 ) 3 39 The effect of silicic acid on the coagulation of clay suspensions by A1 2 (S0 4 ) 3 39 Al necessary to produce a definite clarification in presence of silicic acid. . 40 The effect of sodium salts on the coagulation of clay suspension by A1 2 (S0 4 ) 3 in the presence of silicic acid 40 Effect of silicic acid on the coagulation of clay suspensions in the presence of electrolytes ' v 41 The effect of silicic acid on the coagulation of clay suspensions by A1 2 (S0 4 ) 3 ; 42 The effect of silicic acid on the coagulation of clay suspensions by A1 2 (S0 4 ) 3 43 SILICIC ACID, ITS INFLUENCE AND REMOVAL IN WATER PURIFICATION. By Otto Mitchell Smith. INTRODUCTION. Little attention has been paid to the presence of silicic acid in natural waters except in cases where its content is quite high. It is uni- versally present but usually in amounts less than 100 parts per million. It is more prevalent in the surface waters of the Mississippi Valley 16 especially on the western watershed. It is found in as large amounts as 1,163 parts per million at Cesenate, Italy, 46 1,230 parts per million at Deep Rock Springs, Oswego, N. Y., 13 and 923 parts per million at the Yellowstone National Park. 37 Silica in solution usually occurs as colloidal silicic acid and 32 in combination with the basic elements. Its presence in water used for steam purposes has always been considered detrimental and conducive to the formation of a hard flinty scale. Besides forming an undesirable scale, A. Goldberg 21 believes silicic acid is responsible for many boiler disturbances. When the acid is distilled with water solutions of nitrates and chlorides, nitric and hydrochloric acid are liberated. Turbidity in water is generally caused by the suspensions of very finely divide^ mineral matter, mainly clay. Clay may be defined as a mixture of minerals, of which the most representative members are the silicates of aluminium, iron, the alkalies and alkaline earths. The hydrated aluminium silicate, kaolin. (Al 2 3 ,2Si0 2 ,2H 2 0) is the most abundant compound. There are many cases in the literature emphasizing the difficulties of clarifying water containing finely dispersed clay particles. Fuller 20 in 1898 found at times a turbidity which was difficult to coagulate with an abnormal consumption of alum. Ellms, 17 Black and Veatch, 7 and Catlett 12 show the difficulties in the treatment of such water. THEORETICAL AND HISTORICAL PART. As silicic Mcid and clay suspension are considered as sol and suspen- soid respectively a brief discussion of the general properties of colloids is pertinent. Burton 11 defines a colloid solution "as a suspension, in a liquid medium, of fine particles which may be graded down from those of microscopic to those of molecular dimensions ; these particles may be either homogeneous matter, solid or liquid, or solutions of a small per- centage of the medium in an otherwise homogeneous complex. The one property common to all such solutions is that the suspended matter will remain almost indefinitely in suspension in the liquid, generally in spite of rather wide variations in temperature and pressure; the natural tendency to settle, due to the attraction of gravitation, is overbalanced by some other force tending to keep the small masses in suspension." With the development of the ultra microscope and the investigations of Zsigmondy and Siedentopf 53 it is possible to show that there is a con- tinuous gradation in the size of particles of the disperse phase from 1.70 x 10~ 7 cm. in diameter to that of visible organisms and this leads to the belief that there is a gradation in size from the smallest of these particles to the molecules. Tyndall 47 (1869) found that smaH particles could be revealed by the lateral diffusion of a beam of light traversing a solution. Applying this method Linder and Picton 29 were able to grade the sizes of par- ticles of colloidal arsenous sulfide. Wiedemann 51 (1852) and Quincke 39 (1861) confirmed the discovery of Reuss that a liquid would move across a diaphragm or through a capillary tube towards one of the electrodes when a current is passing. The migration of sols in the electric field was first observed by Linder and Picton. As suspended particles carry electrostatic charges, it seems logical to conclude that as result of these charges, suspended particles whose masses are small enough are equally distributed throughout the liquid, and prevented from ever settling because of the mutual repulsion of the charges. Such a solution is called a sol in contradistinction to the jelly-like form called a gel. When the dispersion of relatively in- soluble particles is not so great as to become macroscopically invisible such a system is termed a suspension. The change of an irreversible hydrosol to an amorphous precipitate wherein there is no Brownian movement is known as coagulation. Schultz/ 3 and Linder and Picton 29 showed that the coagulating power depends upon the valency of the metal ion and according to later workers, equivalent solutions containing monovalent, bivalent and triva- lent metallic ions would possess, whatever the nature of the anion, coagulating powers in the ratio of 1 :35 :1023 which is nearly represented by the formulae 1 :x :x 2 . In 1899 Hardy 22 found that the concentration of the acid necessary for coagulation of electronegative particles and of alkali for the coagulation of electropositive particles is determined by the laws of ordinary chemical equilibrium. Burton/ adding graduated amounts of aluminium sulfate A1 2 (S0 4 ) 3 to a negative sol, found that there was a decrease and even a reversal of the charge. Thus if one is able to pass from a negative sol to a positive sol there must 'be a point of zero potential "isoelectric point." Hardy 23 suggests that the coagulation of colloids by electrolytes takes place when the particles have their charges neutralized by the adsorption of oppositely charged ions of an electrolytic solution and at the isoelectric point. His previously published conclusions were 24 that the conditions which determine coagulation are: (1) concentration of colloid, (2) temperature, (3) the nature of the ion, and (4) that the action is additive if the ions are of the same valency and "subtractive" if of different valencies, as the one inhibits the other. Crum, 15 Linder and Picton, 29 and Whitney and Ober, 50 have estab- lished the fact that during the process of coagulation a portion of the electrolyte is always adsorbed by the coagulum and that amount is pro- portional to the electrochemical equivalents of the anion. Lottermoser, 28 Blitz, 9 and Billitzer 5 demonstrated that colloids of opposite sign precipitate each other; that there is an optimum of pre- cipitating action shown for certain proportions of colloid and that, if in any case these favorable proportions are exceeded on either side, no precipitation occurs ; and that the direction of migration of the whole under the influence of the electric current is the same as that of the colloid in excess. This leads to the subject of protective colloids. Many organic colloids when added in comparatively minute quantities to suspensoids have the power of preventing the coagulation of the system. PROPERTIES OF SILICIC ACID AND CLAY SUSPENSIONS. Silicic Acid. This colloid possesses the properties common to its class. In con- centrated solutions it is unstable, but below 1 per cent it is stable for years. It has not been prepared free from electrolytes; its molecular weight is about 49,000. Billitzer 4 and Fleming 19 found it amphoteric, carrying a negative charge in akaline and a positive charge in acid solu- tions. Fleming's data, converted into terms of hydrogen ion concentra- tion, indicate that the isoelectric point lies between a P h value of 13.6 and 13.9, which is not in accordance with facts. It exhibits the Tyndal effect and is precipitated by electrolytes and positive colloids. The 8 coagulation from dilute solution is irreversible. Hardy 25 determined that the coagulating power of electrolytes, which precipitate a solution of about 1/120 normal silicic acid, varies with the cation, the anion having little effect. Of these aluminium sulfate was the most active and sodium salts least. It is also coagulated within a short time by copper sulfate, cadmium nitrate, calcium, barium and strontium chlorides and carbonates, barium hydroxide, concentrated solutions of ammonium sul- fate, dilute solutions of egg albumin, glue, basic dyestuffs, carbon diox- ide gas and graphite. PapjJada 35 found that neutral salts only act at great concentrations but accelerate gelation (coagulation) according to the lyotope series. S0 4 > Cl > M) 3 ; and Ca > Kb > K > Na > Li Silicic acid hydrosols do not act as a protective colloid for colloidal gold, but F. Kuspert 34 found that real protective action occurs at the moment the silicic acid is being precipitated. Silicic acid is precipi- tated in insoluble compounds by many chemical reagents, but this phase of the subject is not here discussed. Properties of Clay Suspensions. Clay suspensions are impure suspensoids. That they carry a nega- tive charge is shown by Ellms 17 and Count Schwerin. 44 Very dilute suspensions exhibit the Tyndal effect. Ultra-microscopic observations have shown that the parties probably have a diameter of 1.7 x 10~ 7 cm. or less. According to Mayer, Schaffer and Terroine, 80 the addition of a trace of alkali decreases the size of negative suspended particles. Ashley 1 reviewing the work of many investigators came to the conclusion that the colloids in clay are non-crystaline, hydrated, gelatinous alum- inium silicates; organic colloids; gelatinous silicic acid and hydrated ferric oxide; that aluminium hydrate, A1(OH) 3 , is rarely present; and that the colloids of clay may carry into suspension solid particles that are wholly non-colloidal, according to the ordinary ideas, which particles may stabilize the clay sol. The work of Van Bemmelen 48 and Parmelee 36 indicates that the longer clay substance is washed the more colloidal it becomes; and that bodies are gradually hydrolyzed from crystaline compounds to soluble colloids. Ionic Reactions. A. Lottermoser 28 considers that "the hydrosol condition is only possible if one of the reacting ions (I~ + Ag + , Fe +++ + 30H-, Si0 2 + 2H + ) remains up to a certain minimum amount in excess; than on exceeding this limit the gel formation begins and becomes complete if equivalent amounts of the reacting ions are brought together. The hy- drosol condition is bound up with the presence of certain ions in the colloid." If sodium carbonate or an acid be added to a clay which has just enough calcium, Ca, to keep the colloid matter in the gel form, it may react according to the following equations: Ca Gel + Na 2 C0 3 = CaC0 3 + Na 2 Gel or Ca Gel + H 2 S0 4 = CaS0 4 + H 2 Gel. Foerster 18 first perceived the nature of the action of sodium car- bonate on the clay gel. The chemical reactions of the colloidal matter in clay are remarkably similar to those of fats and soaps, the conditions of solubility and insolubility are parallel, but the colloidal matter of clay is more readily acted upon. Thus in a general way the salts of the fatty acid and the sodium, and ammonium sols of clay colloids are soluble in water. The free fatty acid and the acid gels are insoluble in water. The bivalent and trivalent bases form insoluble salts with soaps and insoluble gels with clays. Ashley 1 says "The proof of these con- ceptions is that they are not inconsistent with known facts about clays ; that they have proved a most helpful and suggestive guide in planning investigations, and that they have never misled/' Effect of Electrolytes in Clay Suspensions. Much valuable information is available in the ceramic research on the action of salts on clay suspensions. The effect of the salt varies with (1) the clay; (2) its previous mechanical treatment; (3) its age; (4) concentration; (5) degree of dispersion; (6) and the presence of col- loids. The same electrolyte will have widely different effects, coagulat- ing at one concentration and dispersing at another. The work on slips by Eohland, 40 Mellor and Green and Baugh, 81 Weber, 49 Audley, 3 Kerr and Fulton, 83 Back, 6 Ashley, 2 Thomas, 45 and Bleininger, 8 shows that sodium, potassium and lithium hydroxides, car- bonates, silicates and sulfides generally have a high dispersive or stabil- izing power while the most active coagulating agencies 'are the salts of bivalent and trivalent metals. In Table 1 substances have been classi- fied according to their action on clay slips: 10 TABLE 1 ACTION OF VARIOUS SUBSTANCES ON CLAY SLIPS. Dispersing. Irregular. Neutral. Usually coagulating. Coagulating. Univalent ions in the presence of high hydroxyl ion concentra- NauSOs Na 2 SO 4 - alcohol dilute Grape sugar Humic acid Bivalent and tri- valent ions in tion HgSO-4 solutions of Borax the presence of MgSO 4 NaCl NH 4 C1 high concentra- Alkali salts of weak acids NaaPOV CaCl 2 tion of hydroxyl Na(C 2 HsO 2 ) CaSO-4 ions. Bivalent Strongly dissociated salts in small concentration (?) Ammonium gallate Ammonium urate Aniline ions in the pres- ence of mono NaCl Methyl amine and bivalent HC1 Ethyl amine ions K 2 SO 4 , KHSO4, KNOa Watei glass Tannin Effect is to make the Gallic acid Water glass slip thinner CuS0 4 NH 4 OH Increasing addition tends to K 2 S0 4 A1 2 (SO 4 ) 3 .24 coagulate the slip HiO Small amounts "^ thicken and in- creasing amounts thin the slip - This arrangement seems to confirm Hardy's conclusions that the dispersion or coagulation of a negative sol varies inversely with the valency of the cation, and that the action of anion obeys the regular ionic laws. The above data are obtained from concentrated clay suspensions where the ultimate end is the formation of a stable fluid slip of the highest clay content. In water purification on the other hand, the aim is the removal of a very small amount of clay from a large amount of water. It is clearly evident from a study of dilute clay suspensions and colloidal silicic acid, that the physical state of the substances and their chemical properties must be taken into consideration, together with the factors which influence them, i. e. (1) degree of dispersion; (2) the presence of the protective colloids and absorbed substances; (3) magni- tude and kind of electric charge; (4) the liquid or dispersing medium; (5) the ionic content of the liquid; (6) the concentration; (7) the temperature, and (8) the speed of reaction of added substances. METHODS OF EXPERIMENTATION, MATERIAL AND APPARATUS. Briefly, the experimental work has developed along three lines: (1) the removal of silicic acid from solutions by electrolytes and col- loids; (2) the effect of electrolytes on dilute clay suspensions, and (3) the effect of silicic acid on the coagulation of clay suspensions by alum in the presence of electrolytes. 11 Regeants. Washed potters' clays were used in making the suspensions. No attempt was made to obtain the clay in a high state of purity. Those clays which remained in longest suspension were Tennessee Ball Nos. 3 and 1, number three being the best. Ashley 1 rates Tennessee No. 3 as having a relative colloid content of 95 to 100 per cent. The clay was freed from large particles and soluble salts by washing with distilled water and running the suspension through a Sharpies super-centrifuge ; the desired end was a suspension nearly like that in surface waters but as free as possible from electrolytes. Characterization of Suspensions. Coefficient of By turbidimeter. By weight, fineness. Tennessee No. 1 Ball ' 400* 325* 0.75 Tennessee No. 3 Ball 400 330 0.82 The No. 3 clay stood in contact with water in ordinary glass bottles over one year before being used. Kahlbaum's, Merck's, or analyzed chemicals were used without further purification. Colloidal silicic acid was prepared as directed by Graham and dialyzed with parchment paper until there remained only a trace of chloride. Determination of hydro- gen ion concentrations was by the colorimetric method of Clark and Lubs. 14 Standards and indicators were made as directed using ordinary distilled water. The accuracy of the solutions was determined by check- ing against fresh permanent standards prepared by Clark and Lubs. The accuracy of the readings is within -f- 0-1 P^ value. Measurements of Turbidity. Standards for determining turbidities were prepared from the original clay suspensions and checked with a standardized turbidimeter. Turbidities were not read closer than 7 per cent. Fine particles of various sizes in suspension do not settle uniformly leaving a clear supernant liquid but are deposited in layers or zones. According to Wiley's formulae, which is determined for particles between 0.0001 and 0.02 millimeters in diameter, the rate of fall is proportional to the square root of the diameter of the particle that is D = 0.0255V. 2 Those particles of approximately the same diameter will settle to- gether leaving a turbid suspension of finer particles above. In a sus- pension one can observe two or three of these zones of widely different turbidities and rates of sedimentation. Within a zone the turbidity is * Expressed in parts per million. 12 fairly uniform, but different zones vary as much as 100 per cent within a vertical distance of % inch. If the turbidity of an undisturbed suspension be taken at regular intervals at a fixed point, it will be noticed that the numerical values will be nearly constant for some time, then within a brief interval of time will change markedly as particles from a higher zone traverse the field of view. This phenomenon is well shown in many of the figures. As it was necessary to determine the turbidity of a suspension Without disturbing the liquid, the following method was devised. A stereopticon equipped with a 500 watt lamp was adjusted to yield a slightly diverging cone of light. Two screens each having four holes 3/16 inches in diameter were inserted in the path of this beam. The holes in each screen were in a vertical row arranged in pairs. The centers of the holes were %, %, and 2% inches below the surface of the liquid in the sample bottles. One screen was placed l 1 /^ inches in front of and iy 2 inches to the side of the other screen so that the standard and the unknown could be observed together. The observations were made at right angles to the beam of light. The 100 cc. samples were contained in 4-ounce French square flint bottles, 1% inches square; the depth of the liquid was 2% inches. For turbidities below 100 parts per million the ordinary method of compar- ing with standards in similar bottles was used. In the coagulation of the suspensions, the reagents were added in sufficient concentrations to give a reaction before the natural occurrence of sedimentation of par- ticles. All results are expressed as turbidities in parts per million, and milligram equivalents per liter, except in the case of colloidal silicic acid which is in parts of Si0 2 per million. EXPERIMENTAL PART. In dealing with the action of electrolytes on colloids a method of determining the amount of colloids present is greatly "to be desired. A search of the literature revealed only one method that seemed applicable. This was devised by Eohland 41 and improved upon by Ashley. 1 It de- pends upon the amount of dye adsorbed by the colloid and is fairly successful in estimating the adsorptive or colloidal power of concentrated clay suspension, but is not reliable when applied to dilute suspensions of clay. Further work along this line is needed. The refractive index and viscosity were investigated, but variations were too slight to be serviceable. There appeared to be no correlation between the hydrogen ion concentration and the degree of coagulation. Kataphoresis experi- ments only measure the sign and not the magnitude of the charge. 13 Later in the work it was found that silicic acid exerted a marked in- fluence and an endeavor was made to find a method of separating it from the clay. Several 38 ' 52 have been given but none of sufficient selec- tive action to be of value at these dilutions. An attempt was made to filter the silicic acid from a suspension by a Berkefeld army filter No. 3 with the following results: Silicic Acid. added. recovered. 298* 155 129 84 This confirms the work of Linder and Picton. 29 The silicic acid is evidently coagulated by contact with the walls of the filter. INFLUENCE OF ELECTROLYTES ON THE PRECIPITATION OF SILICIC ACID FROM DILUTE SOLUTIONS. The amount of colloidal silicic acid in a solution containing no suspensoids is easily obtained by the usual method of analysis. In the determinations of silica, Si0 2 , one evaporation was made, as on a second evaporation only a few tenths of a milligram additional was obtained. 27 The errors due to the manner of adding the reagents and their different concentration was reduced to a minimum by making the methods of manipulation as uniform as possible. TABLE 2 SILICIC ACID is NOT PRECIPITATED FROM 5 cc. OF A SOLUTION- CONTAINING 625 PARTS PER MILLION Si0 2 BY THE FOLLOWING SALTS I Reagent. Temperature ca 23 C. Reagent. Time 6 hours. No. cc. used. Highest concen- tration. No. cc. used. Highest concen- tration. 1 N NaCl 12 12 12 12 8 12 12 .07N .07N .07N .07N .06N .07N .07N 2 N BaCb 8 4.2 12 11 12 10 12 .06N .05N .07N .07N ,07N .07N .07N 1 N Na 2 COs 1.0 N CaCl 2 1 N NaHCOs 5 N Ah(SO4)3 0.1 N Na 2 S0 4 andK 2 SO 4 1 N NasPO4 0.1 N Feds... 0.1 N FeSO4(NH4) 2 SO4 0.1 N Mg(HCO 3 ) 2 0.1 N FeCl 2 ... 0. 1 N MgSO4 0. 1 N Aids From a solution containing not less than 184 nor more than 625 parts per million of Si0 2 , silicic acid is not precipitated by the reagents given in Table 1, but is precipitated from 5 cc. of a solution containing 625 parts per million of Si0 2 , by the following : * Expressed as parts per million. 14 Temperature ca 23 C. Time: 5 minutes. 5 -5 cc. 0.18 N" NaOH .099 N concentration 6 cc 0.018 N Ca(OH) 2 .0019 N concentration 4 cc 0.26 N Ba(OH) 2 .0019 N concentration 2-6 cc (.6 mg) colloidal Fe .0045 N concentration This experiment indicates that bivalent ions have a precipitating value fifty times that of monovalent ions. Trivalent ions according to the formulae l:x; x 2 should have a coagulating value of 2,500 times that of the univalent, or fifty times that of the bivalent ions. Qualitative experiments with more dilute solutions indicate that the ratio between bivalent and trivalent to be about one to four. Colloidal iron is only one-half as efficient as calcium hydroxide. The more dilute a solution the less marked is this precipitating effect of the cations. Very dilute solutions 30 or 40 parts per million of Si0 2 are not precipitated by sodium hydroxide, and rather high concentration of calcium hydroxide to silicic acid are necessary to obtain a precipitate within six hours. At these concentrations the reactions no doubt are ionic, and precipitates of calcium silicates are thrown down. Since the hydroxides of calcium and barium precipitate silicic acid, it is desirable to know the effect of added univalent and bivalent ions upon the amount of calcium hydroxide needed. To 5 cc. of a dialyzed silicic acid solution, containing 625 parts per million of Si0 2 there was added 0.5 cc. of a 0.1 normal solution of the electrolytes and the silicic acid was precipitated by the addition of calcium hydroxide. The results are given in Table 3. TABLE 3 THE EFFECT OF THE ADDITION OF VARIOUS ELECTROLYTES ON THE PRECIPITATION OF COLLOIDAL SILICIC ACID* BY CALCIUM HYDROXIDE. Temperature ca 23C. Time 5 minutes. Reagent. Ca(OH) 2 required. 0.0 cc. 0.2 N Ca(OH)2 0. 6 cc. 0.5 cc. 0.1 N BaCl 2 0.3 cc. 0.5 cc. 0.1 N MgSO 4 0.3 cc. 0.5 cc. 0.1 N CfcCb 0.3 cc. 0. 5 cc. 0.1 N NaHSO 4 0.3 cc. 1 in excess re- 0.5 cc. 0.1 N H 2 SO4 0. 35 [quired to 0.5 cc. 0.1 N KHSO4 0. 3 cc. J neutralize 0.5 cc. 0.1 N NaCl 0.6 cc. 0.5 cc. 0.1 N Na 2 SO4 0. 7 cc. 0.5 cc. 0.1 N K 2 SO 4 0.7 cc. 0.5 cc. 0.1 N Na 2 SO 4 0. 6-0. 7 cc. * 5 cc. of silicic acid containing 625 parts per million of SiO 2 were used. 15 The following compounds complicate the precipitation by reacting with the Ca. ions: V, Na 2 C0 3 Na 3 P0 4 Ca(HC0 3 ) 2 NaHC0 3 Mg(HC0 2 ) 3 This experiment proved, as expected, that the precipitating value of the cation depends upon its valence and that hydroxyl had some in- fluence at these concentrations. Neutral salts of the univalent cations were not present in sufficient amount to influence the results. In cases where the solution was acid or there was formed an insoluble calcium precipitate, the concentration of calcium ion is that necessary to com- bine with the anion plus an amount sufficient to produce precipitation. This is shown in the reactions of sulfuric acid and calcium hydroxide. To measure the effect of the hydroxylion on the reaction, the pH value was determined, the results of which are given in Table 4. TABLE 4 THE CONCENTRATION OF HYDROGEN ION NECESSARY BEFORE SILICIC ACID* IS PRECIPITATED BY Ca(OH) 2 IN THE PRESENCE OF SALTS. Salts. pH value. 0.18 N NaOH 9.5 0.01 N Ca(OH) 2 ....' 9.2 0.02 N NaOH 9.2 colloidal Fe off color Precipitation of silicic acid by Ca(OH) 2 in presence of 0.5 cc. of 0.1 normal salt solutions : Salts. pH value. 0.1 N CaCl 2 ...: -. . 9.2 0.1 N MgS0 4 8.6 0.1 N H 2 S0 4 9.2 0.1 N NaCl 9.3 0.1 N NaHS0 4 9.0 0.1 N NaHC0 3 9.2 0.1 N Na,S0 4 9.4 0.1 N Na 2 C0 3 9.4 0.1 N Na 3 P0 4 9.3 10 cc. CaS0 4 saturated soln . 9.2 0.1 N Mg(HC0 3 ) 2 8.8 0.1 N MgS0 4 8.6 0.1 N A1C1 3 . below 7.5 when precipitation occurred 0.1 N A1 2 (S0 4 ) 3 .below 7.5 when precipitation occurred Average 9.0 5 cc. of a solution containing 625 parts per million of SiO 2 were used. 16 The different systems of salt, silicic acid and calcium ions have different normalities but approximately a pH value of 9.0. This points to the fact that silicic acid may act as a buffer in the manner of large complex organic molecules. Magnesium salts show a lower value than calcium which is probably due to the precipitation of magnesium hydrox- ide which obscured the true end point. It was not possible to determine hydrogen ion concentration of iron salts with indicators. The quanti- tative removal of silicic acid by calcium hydroxide, barium hydroxide and chloride, aluminium sulfate and colloidal iron was tried on a water from Albuquerque, N. M., the analysis of which is given in Table 5. TABLE 5 ANALYSIS OF SAMPLES OF WATER FROM ALBUQUERQUE, NEW MEXICO. Residue.* 479.1 p. p. m. N equivalents 1000 Silica SiO2 82.6 Non volatile. 0.3 AhOs Fe2um m a valm cons* vof tent /&. 3 W 7 2k A/ 100 / 7 X -?nn :/ 2 ? c^ / \ 5 -50 / /W ( 5 7 t P H V< \ J i/ue > / I 4 6. A10 + + HO + H 2 and that above a pH value of 11.0 so far to the left that the solid aluminium hydroxide phase is unstable and disappears. In the region below these curves the tendency to form colloidal solutions is quite marked, and the gelatinous nature of the precipitate and the magnitude of the Tyndal effect generally varies inversely with the amount of aluminium ions added to the system, and directly as the hydrogen ion concentration departs in either direction from a pH value of 8.25. The precipitation of silicic acid by aluminium sulfate may be ex- plained : ( 1 ) by the neutralization of the charge on the silicic acid com- plex by the aluminium ion, resulting in a precipitation of silicic acid, (2) by the neutralization of the negatively charged silicic acid by posi- tively charged aluminium hydroxide, (3) by the solid aluminium hy- droxide adsorbing silicic acid, and (4) by the formation of an insoluble chemical compound. THE INFLUENCE OF ELECTROLYTES AND SILICIC ACID ON THE COAGULATION OF CLAY SUSPENSIONS. Jhe next step was to determine the effect of silicic acid and the commonly occurring electrolytes on the coagulation of a clay suspension with aluminium sulfate. Reagents were added to one hundred cubic centimeter portions of a clay suspension and thoroughly mixed by shaking for one minute. The sample was then allowed to remain perfectly quiet and the turbidity of the liquid determined at appropriate intervals at a point 1/^-inch below its surface. 22 The electrolytes used were sodium hydroxide, acid carbonate, car- bonate and chloride; magnesium bicarbonate, and sulfate; calcium hy- droxide, chloride and bicarbonate; barium hydroxide, Merck's dialyzed iron and sulfuric acid. The effect of electrolytes on the stability of a dilute clay suspension is similar to that observed with clay slips. The results of adding in- creasing amounts of sodium hydroxide is first dispersion followed by coagulation. 2 In Tables 10 and 11 and Figures 3 and 4 it is shown that the coagulative powers of calcium and barium hydroxide are practically the same, and that the ratio of aluminium to calcium and barium ions is about five to one. Data from many of the tables have been collected and shown in graphic form in Figures 6 and 7. The salts arranged according to their efficiencies as coagulants are : aluminium sulfate, cal- cium and barium hydroxides, calcium chloride, magnesium sulfate and magnesium bicarbonate. Sodium chloride has little effect until its concentration becomes so great as to salt out the clay. Sulfuric acid has no apparent effect up to concentration of 0.35 milliequivalents but higher concentrations coagulate. (See Table 16 and Figure 5.) Sodium hydroxide, carbonate, acid carbonate and sulfate have at first a stabil- izing influence followed by a coagulating effect. The coagulating effect of anions seems to be an inverse function of their valencies. The effect of added salts on the coagulation of clay suspensions by aluminium -sulfate is shown in Tables 12 to 20, Figures 8 to 15. TABLE 10 THE EFFECT OF VARYING STRENGTHS OF Naon ON THE RATE OF COAGULATION OF CLAY BY A1 2 (S0 4 ) 3 . Tenn. No. 3 Ball Clay Temperature ca 23 ( Milliequivalents of NaOH added. C. Milliequivalents of Al. 0.00 0.049 0.09 0.18 0.36 Turbidity. Ihr. It hrs. Ihr. 2 hrs. Ihr. 2 hrs. Ihr. 2 hrs. Ihr. IJhrs^ 0.0 .02 .06 .09, .13 .17 .18 .37 .55 .74 .92 1.11 1.29 1.47 1.66 2.03 420 420 400 150 125 125 100 100 125 420 400 400 75 75 50 50 420 400 420 420 400 350 175 275 375 450 300 400 360 360 375 50 150 175 350 250 450 400 450 400 450 420 300 200 200 400 420 420 420 240 75 40 400 375 420 150 100 100 100 100 400 100 90 85 85 80 400 400 350 350 50 75 400 300 75 300 250 190 125 60 275 200 75 125 90 90 50 150 60 100 75 240 100 420 125 150 75 90 50 200 50 350 100 150 50 23 TABLE 11 COMPAKISON OF ca(oH) 2 , Ba(oH) 2 , COLLOIDAL FE AND Al. (S0 4 ) 3 AS COAGULANTS. Tenn. No. 3 Ball Clay Temperature ca 23 C. Milligram equivalents. A1 2 (SO 4 ) 3 . Turbidity. Ca(OH) 2 . Turbidity. Ba(OH) 2 . Turbidity. Fe. Turbidity. .02 .06 .09 .13 .17 .37 400 400 400 375 50 40 .26 .43 .60 .69 .86 1.03 400 390 350 350 150 100 .13 .66 .79 1.05 1.32 400 400 375 100 60 .38* 3.9 4.6 6.1 6.9 8.1 400 400 350 200 100 40 1.38 50 9.2 40 1.72 50 'Parts per million. TABLE 12 EFFECT OF Na 2 co 3 ON THE COAGULATION OF A CLAY SUSPENSION BY A1 2 (S0 4 ) 2 . Tenn. No. 1 Ball Clay Temperature ca 24' Milliequivalents of Na 2 CO 3 added. C. Milliequivalents of Al. 0.5 1.0 Turbidity. 14 hrs. 14 hrs. 14 hrs. 14 hrs. .30 .45 .60 .75 .90 400+ 125 125 175 400 400 150 10 400+ 400 125 65 125 100 150 150 10 TABLE 13 EFFECT OF Nanco 3 ON THE COAGULATION OF A CLAY SUSPENSION BY A1 2 (S0 4 ) 3 . Tenn. No. 1 Ball Clay Temperature ca 24 C. Milliequivalents of NaHCOs added. Milliequiva- lents of Al. .0 3 0.6 1. 1. * li hrs. 13 hrs. li hrs. 13 hrs. 14 hrs. 13 hrs. 14 hrs. 13 hrs. .03 400+ 400+ 325 275 400 400 400 375 400 300 400+ 300 .06 .09 400+ 400 150 150 400 400 300 275 400 200 400+ 250 .12 15 350 275 75 75 300 225 100 65 350 200 5 400 150 .18 .24 .30 .36 125 60 60 50 10 10 175 180 75 60 50 10 10 5 125 100 60 60 40 10 10 5 250 75+ 75+ 60 150 25 10 5 .42 60 60 5 75 5 .48 65 60 50 .54 65 60 60 .60 60 60 .66 60 o :? 60 .90 50 TABLE 14 ACCELERATING EFFECT OF Na 2 so 4 'oN THE COAGULATION OF CLAY SUSPENSION BY A1 2 (S0 4 ) 3 . Tenn. No. 3 Ball Clay Temperature ca 24 C. Milliequivalents of Na,SO 4 added. Milli- equivalents of Al. 1 2 3 4 Turbidity. 1* 24 1 24 24 1 24 .0 .03 .06 .09 .12 .15 .18 400 375 350 350 25 350 150 35 10 400 350 400 150 400 400 200 60 50 40 25 125 60 50 40 * Time expressed in hours. TABLE 15 EFFECT OF NaCl ON THE COAGULATION OF * CLAY SUSPENSIONS BY ALUM. Tenn. No. 3 Ball Clay Temperature ca 24 C. Tenn. No. 1 Ball Clay. Milliequivalents of NaCl added. Mini- equiva- lents of Al. 1 3 10 1 2 Turbidity. 1* 24 1 24 1 24 li 12 Ij 12 .0 .03 .06 .09 .12 200 150 15 150 10 150 400 400 100 90 80 300 200 100 400 400 400 80 250 200 60 400 350 100 25 350 175 50 40 50 50 50 50 * Time expressed in hours. TABLE 16 THE EFFECT OF VARYING STRENGTHS OF H 2 so 4 ON THE RATE OF CLAY BY A1 2 (S0 4 ) 3 . Tenn. No. 3 Ball Clay Turbidity 420 Temperature Ca 23 C. Milliequivalents of H SO 4 added. Milli- equiva- lents of Al. 0.0 0.049 0.09 0.18 " 0.36 Turbidity. 1 hr. 2 hrs. 1 hr. 2 hrs. 1 hr. 1| hr. 1 hr. IJ hr. 1 hr. li hr. 0.000 .02 .06 .09 .13 .17 .18 .37 .55 .74 .92 420 420 400 150 125 125 100 420 400 400 75 ' 75 75 50 420 420 420 400 370 200 150 400 375 310 60 60 420 400 290 240 140 400 400 1 150 100 60 420 250 375 200 360 150 100 350 90 60 150 50 150 100 50 60 140 125 100 50 50 50 75 75 50 50 50 50 150 200 200 ( 50 50 50 125 50 150 60 TABLE 17 -EFFECT OF Mgso 4 AND Mg(HC0 3 ) 2 ON THE COAGULATION OF CLAY SUSPENSIONS BY A1 2 (S0 4 ) 3 . Tenn. No. 1 Ball Clay Temperature ca 24 C. Milliequjvalents of MgSO 4 added. Milliequiva- lents of A). 1 2 3 Turbidity. 11 hrs. 14 hrs. 11 hrs. 14 hrs. 11 hrs. 14 hrs. .009 .018 .030 .036 .060 .090 .120 400 100 350 50 325 400 175 150 125 100 15 15 10 200 25 150 10 125 125 125 100 10 150 125 125 TABUE 18 MILLIEQUIVALENTS OF Mg(HC0 3 ) 2 ADDED. Milliequiva- lents of Al. 1 2 3 Turbidity. 11 hrs. 14 hrs. 11 hrs. 14 hrs. 11 hrs. 14 hrs. .009 .018 .030 .045 .060 .090 .120 400 150 400 85 350 350 300 275 250 125 65 30 25 20 15 15 80 35 350 125 350 250 175 125 25 15 10 10 150 1 TABLE 19 EFFECT OF CaCl 2 THE COAGULATION OF CLAY SUSPENSIONS BY A1 2 (S0 4 ) 3 . Tenn. No. 1 Ball Clay Temperature ca 24 C. Milliequivalents of CaCl,. Milliequiva- lents of Al. 1 2 3* Turbidity. 11 hrs. 14 hrs. 11 hrs. - 14 hrs. 11 hrs. 14 hrs. .009 .010 .030 .045 .060 .090 .120 400 125 185 10 175 175 175 200 175 125 10 10 5 185 180 5 300 25 175 125 175 (?) 10 5 175 175 175 TABLE 20 EFFECT OF ca(Hco s ) 2 ON THE COAGULATION OF CLAY SUSPENSIONS BY A1 2 (S0 4 ) 3 . Tenn. No. 1 Ball Clay Temperature ca 23 C. Milliequivalents of Ca(HCO 3 ),. Milliequiva- lent of Al. 1 2 Time in hours and turbidity. H 13 li 13 ii 13 u 13 .009 .018 .030 .045 .060 .090 .120 325 325 125 200 200 150 150 125 200 75 50 40 25 25 10 5 175 10 325 125 200 15 325 75 50 70 10 75 50 50 10 5 i .06 -/ *I8 -55 -9Z Scale I div. - .006_1*_ Scale I div. - .037 : + /. 29 /. 67 2.03 Milliequivalents of Aluminium. Figure 3. Effect of NaOH on the coagulation of a clay suspension by A1.,(SO 4 ), Time of seff//nj = one hour. Milliequivalents .of Aluminium. Figure 4. Comparison of A1 2 (SO 4 ) 3 , colloidal Fe, Ca(OH) 2 , lia(OH), as coagulants. 27 32&A 3/V) Tlme of sett/tnq > ijfoZ hours Legend no millieyuiva/enfs of Hf5Q^ added. .09 mjlliequ/m/cnts ofH^^O^added- . 36rni//ieqw vafarfc tf/jJQf added. 5 A- o- x- ZOO I 100- \ k f * * ^ -o- e Mmjeouiva/enfe of Aluminum Figure 5. Effect of H 2 SO 4 on the coagulation of clay suspension by Al 2 (SO 4 ) r o .2 .4 Scate I dry.' .020 1.0 2.0 3.0 4.0 Scale 1 div. * .100 Milliequivalents of Electrolytes. Figure 6. Effect of electrolytes on the coagulation and settling of clay suspensions. a 3 Dd .03 '06 .09 Turbidity of Suspension *toa< 300 'mil/i i 'ec/uiv. of 'NaGI. s \ \ V/7/7 ^\ \ s k^ > N > O .03 .06 .09 .12. 1 ,03 .06 .09 7Z Milliequivalents of Aluminium. Figure 10. Effect of Na,SO 4 and NaCl on the coagulation of clay suspension by A1 2 (S0 4 ),. 31 ZT/^:/ of CdC/? Milliequivalents of Aluminium. Figure 11. Effect of Mg(HCO 3 ) 2 and of CaCl, on the coagulation of clay suspension by A1,(SO 4 ) 3 . ilJ O o p II 33 III! O g C O 34 ox w <& ^"3 < o 35 400- Time of sett// nq /j? hours Co^fro/. A- No o - CaflCOj)t present .03 .06 .07 ./2 Milliequivalents of Aluminium. Figure 15. Effect of Ca(HCO 3 ) on the coagulation of clay suspensions by A1 3 (S0 4 ) 8 . As would be expected the presence of the trivalent and bivalent ions aid in the coagulation, sodium chloride and sulfuric acid have little effect at low concentrations, while the addition of the sodium salts causes a behavior similar to that produced by the action of sodium hy- droxide (Figure 3). As the content of sodium hydroxide is increased the amount of aluminium sulfate must be increased in order to produce coagulation and to combat the dispersive power of the sodium compound. This same effect is quite noticeable with sodium carbonate but less so with sodium acid carbonate and sulfate. The point of coagulation is not dependent on the alkalinity of the solution. The results of these experiments are partly in accordance with work and theories of Rohland. 40 ' 42 There seems to be no question but that coagulation is a function of the concentration of the hydroxyl ion and alkali metal ions as well as the valencies of the cation. The monovalent ion of the alkalies is intimately connected with the dispersive or protective action of sodium salts in the coagulation of clay suspensoids and with the peculiar (Protective) effect of silicic acid in preventing the formation of aluminium hydroxide by the action of aluminium chloride and sodium hydroxide. If .calcium is substituted for sodium these peculiar effects are not produced. The effect of silicic acid on the coagulation of clay suspensions by aluminium sulfate is shown in Tables 21, 22, and 23, and Figures 15 to 19. A suspension containing 62 parts per million of dialyzed silicic acid and appropriate amounts of electrolytes were coagulated with alum and the rate of reaction compared with a suspension containing no added silicic acid. In all cases the effect of the added silicic acid was to retard the reaction, and more aluminium sulfate was required to pro- 36 duce coagulation than before regardless of the presence or absence of electrolytes. Photographs (Figures 20 and 21) were taken five "days after the addition of the aluminium sulfate. Silicic acid was added to the samples on the right. The data given in Tables 19 and 20 was ob- tained from this experiment. The aluminium consumed is a function of the silicic acid added, but the mathematical relationship is not a simple one and varies with the clay used. This relationship is shown in Figures 15 and 17, which have been plotted from data in Tables 21 and 22. In general the amount of aluminium required to coagulate, per unit amount of silicic acid added, is larger at low than at high concentrations. Silicic acid does not seem to stabilize or disperse the clay particles, nor does its presence influence the rate of sedimentation. In this respect it differs from the alkali salts. TABLE 21 RETARDING EFFECT OF SILICIC ACID ON THE COAGULATION OF CLAY SUSPENSION BY A1 2 (S0 4 ) 3 . Tenn. No. 3 Ball Clay Temperature ca 24C. Silicic Acid (SiO 2 ) added. Milligram equiva- lents. ppm.* 12.4 ppm. 24.8 ppm. 37.2 ppm. 49.6 ppm. Turbidity. I*** 13 11 13 11 13 tl 13 li 13 .0 03 06 09 12 15 400 400 400 125 75 225 150 125 10 5 400 225 400 250 400 275 400 275 400 400 50 50 175 150 10 10 400 250 400 275 400 275 125 50 50 75 20 10 175 125 225 125 50 60 50 20 10 10 50 50 40 25 25 10 10 10 The Amount of Alum Necessary to Produce a Definite Clarification in 13 Hours in the Presence of Silicic Acid. * Parts per million. ** Time expressed in hours, t Milligram equivalents. Ah (864)3 added to reduce turbidity to SiO2 added. 100 ppm. 50 ppm. 10 ppm. .07f .084* 12.4 ppm. .102 .111 24.8 ppm. .112 .129 37.2 ppm. .126 .153 49.7 ppm. .129 15.3 37 TABLE 22 EETARDING EFFECT or SILICIC ACID ON THE COAGULATION OF CLAY SUSPENSION BY A1 2 (S0 4 ) 3 . Tenn. No. 3 Ball Clay Temperature ca 24C. Silicic Acid (SiO 2 ) added. Milligram equiva- lents. ppm.* 12.4 ppm. 24.8 ppm. 37.2 ppm. 49.6 ppm. Turbidity. 1** 24 1 24 1 24 1 24 24 .03 .06 .09 .12 .15 .18 .21 .24 .27 .30 .33 .39 .45 .51 400 400 175 30 200 150 5 5 350 325 200 75 40 40 166 50 25 5 5 5 350 350 350 350 212 100 50 40 40 350 350 350 125 50 50 50 50 50 50 25 10 5 5 100 100 100 40 30 20 5 5 200 75 100 50 50 25 20 20 20 15 12 10 The Amount of Ah (864)3 Necessary to Produce a Definite Clarification in One Hour in the Presence of Silicic Acid. * Parts per million. ** Time expressed in hours. 4- Milligram equivalents. A1 2 (SO 4 )3 added to reduce turbidity to SiO2 added. 100 ppm. 50 ppm. .105+ .115+ 12.4 ppm. .174 .200. 24.8 ppm. .210 .240 37.2 ppm. 49.6 ppm. .240 .270 .270 .330 38 TABLE 23 EETAEDING EFFECT OF SILICIC ACID ON THE COAGULATION OF CLAY SUSPENSIONS BY A1 2 (S0 2 ) 3 IN THE PRESENCE OF ELECTRO- LYTES. Term. No. 1 Ball Clay Turbidity 400 parts per million Temperature ca 21 C Time 3 hours. Milli- equivalent of Al. Milligram equivalents of salts. . 18 NaOH. 1.5 NaHCOs. 2 NaCl. 2 Na 2 SO 4 2 Mg(HC0 3 ) 2 . Dialyzed silicic acid, parts per million as SiO2. 62 62 62 62 62 .03 .06 .09 .12 .15 .18 .21 .27 .30 .36 .39 .45 .60 350 350 325 225 400 400 400 400 400 350 25 400 350 75 350 400 25 400 25* 25 400 400 35 125 300 350 25 10 10 85 400 35 15 125 10 65 350 65 300 ' Milli- equivalent of Al. Milligram equivalents of saits. 2 MgSO 4 . 0.7 Ca(OH 2 ).* 1.0 Ca(HC0 3 ) 2 . 0.7 CaOH 2 f 1.0 ca (HCOa) 2 . Dialyzed silicic acid, parts per million as SiO2. 62 62 62 62 .03 .06 .09 .12 .15 .18 .21 .27 .30 .36 .39 .45 .60 400 400 85 85 85 125 100 180 180 170 170 160 350 100 65 65 400 400 400 400 175 400 50 250 350 50 150 175 325 25 65 150 250 * 13 hours. t Readings expressed as turbidities. 39 O 03 .06 W ,IZ .15 /&. .03 .08 JW .12. ./ff /S .& .24 _4 pn* 5en1 zoo //yi SiO, -j. *4fi LLILL rj. 5 N x \ fvr' MN ( | " ] IJx, f N \ i t .09 M ,15 .16 .21 .24 .Zl .30 .01 -IZ .& JO 1 .01- .Of JZ .15 .16 .21 ^4 .? .30 Milliequivalents of Aluminium. Figure 17. Effect of silicic acid on the coagulation of clay by A1 2 (SO 4 ), 40 .03 .06 .09 OJZ .15 J8 1 4 .27 .30 .33 Milliequivalents of Aluminium. Figure 18. Al necessary to produce a definite clarification in the presence of silicic acid. Plotted from Figure 14. Legend r Zwtilliequhatefrh of 300 .2 J05 .00 .09 ./S /5 JO 1 -34- .27 30 .35 Mill/equivalents of Aluminum Legend. X 0.18 mi Hi equivalents NaOfi O O.ldmillieyuivalenf5 NaOH +6f>.pm. 5iO x 400 3OO SOO ICO t ~\ T l.5millie^ ,v. 'Gf//J /K7//1 yaHC0 3 -h &n \m PL! ^ r~-\ I \; \ T \ x > \ \ \ ^ 7 .C v .0 6 .6 9 ./ f .19 .18 .fl .24 .& 30 .33 .J6 .39 Mi 1 1 /equivalents of Aluminum Figure 19. Effect of sodium salts on the coagulation of clay suspension by A1,(SO 4 ) 3 in the presents of silicic acid. 300+ 300 200 . ,00 ft ft .s N \ i^ ^ 2J. \ \ v s Vr s > 1 ^ > ^ 7 1 H , I ? .03 .06 .09 .f2 .Iff J6 .03 .06 .09 72 ,/S -16 2 Flillieauivs. WgCHCOj) Milliequivs. Hg50 4 400 3OO zoo 100 5/C f 6 < Vv ' 77. ( V \ 1 s -^ fc 4 i .09 ./ ./5" JO .09 .06 .03 Jt J5 JO .21 I Mil/icqu/v. Ca(0ff) z . Ca(HCO,) .O.J 1.0 Milliequivalents of Aluminium. Figure 20. Effect of silicic acid on the coagulation of clay suspensions in the presence of Electrolytes. m 5 eo II i 735 73 o c la s s 01 o j -3 Electrolytes Mgr(HCOs) 2 NaHCOs Absent. Present. Colloidal Silicic Acid. Figure 22. The effect of silicic acid on the coagulation of clay suspensions by A1 2 (SO 4 ) 3 . APPLICATION. Removal of Silicic Acid From Water to be Used for Boiler Purpose, The experiment on a natural water already referred to indicates that silicic acid could be most economically removed by aluminium hy- droxide formed in the reaction of aluminium sulfate with calcium hy- droxide in a solution whose Ph value is 8.0 9.0, and that the precipi- tation is more or less directly related to the ratio of (Ca + Mg) : Na. The higher this ratio the more complete is the removal. By the proper treatment with aluminium sulfate and line it is possible to reduce the silicic acid content from 82.6 to 30 parts per million. The Coagulation of Waters Containing Colloidal Clay. The stability of a suspension of clay seems to be intimately con- nected with the amount of monovalent cations and bivalent anions pres- ent. Thus the alum needed to coagulate will be greater the larger the concentration of sodium ions except in the case when the anion is mainly chlorine. Less alum will be needed as the ratio of the Ca + Mg ions to the sodium ion increases. As the silicic acid content increases more 44 alum will be required to coagulate. In concentrations up to 20 parts per million, from 0.015 to 0.03 milligram equivalents of aluminium (Al) per 10 parts of Si0 2 is needed to combat the influence of the silicic acid. Rate of Reaction. Water containing bivalent ions when treated with alum gives a sharp, abrupt reaction, an increase of 0.3 milligram equivalents of aluminium, Al, coagulates, but when silicic acid or alka- lies are present, other. factors being constant, a much larger amount of alum is necessary to produce the same clarification and the abruptness of the reaction becomes less as the amount of the silicic acid and alkalies approaches a certain maximum where the magnitude of the change pro- duced per unit amount of alum is much smaller than in the former case. This is well shown in Figures 8, 9, 12, 14, 15, and 21. This phe- nomenon is exactly similar to that which occurs when "colloidal waters" are coagulated by alum and lime. These experiments justify the addition of an excess of calcium hy- droxide, Ca(OH) 2 and allowing it to react with the water for some time (8 to 12 hours) before the addition of the alum or ferrous sulfate. This procedure has been effective in purification of water from the Arkansas River at Little Rock, Arkansas when the suspended material is in a colloidal state. SUMMARY. 1. Colloidal silicic acid in dilute solution can be precipitated by aluminium hydroxide. 2. Dilute solutions of dialyzed and undialyzed silicic acid behave towards electrolytes in the same manner as concentrated solutions with the exception that proportionately more reagent is needed. 3. The optimum hydrogen ion concentration for the precipitation of the aluminium hydroxide and the removal of silicic acid by aluminium hydroxide is a concentration of 1 x 10~ 8 . 4. The limiting values of the hydrogen ion concentration, between which the solid aluminium hydroxide phase is present are 1 x lO^ 4 and 11 x 10- 11 . 5. The presence of silicic acid prevents the formation of a preci- pitate of aluminium hydroxide, when the sodium hydroxide, equivalent to the aluminium chloride present, is added all at once. The silicic acid probably as a protective colloid prevents precipitation of the alumi- nium hydroxide. The presence of bivalent cations destroys this pro- tective power. 6. The action of electrolytes on clay suspensoids is the same in dilute as in concentrated suspensions. (Slips). 45 7. Sodium hydroxide, acid carbonate, and sulfate stabilize or dis- perse clay suspensions at one concentration and coagulate at another. 8. The ratio of the coagulating power of calcium and barium hy- droxide to aluminium hydroxide is about 1 to 5. 9. Coagulation of clay suspensions is aided by the bivalent and hindered by monovalent cations in the presence of acid carbonate, car- bonate, sulfate and hydroxyl anions. 10. Silicic acid retards coagulation of clay suspensoids. REFERENCES. 1. Ashley, H. E., The colloid matter of clay and its measurements, U. S. Geol. Surv. Bull. 388 (1909). 2. Ashley, H. E., The technical control of the colloid matter in clays, Trans. Am. Ceram. Soc. 12, 768 (1910). 3. Audley, J. A., Some casting slip troubles, Trans. Eng. Ceram. Soc. 14, 152 (1914-15). 4. Billitzer, Jean, Theorie der Kolloide, Z. Physik. Chem. 51, 150 (1905). 5. Billitzer, Jean, Theorie der Kolloide, Ber. Wien Akad. Wiss. 113. (II) 1159 (1904). 6. Back, Robert, Effect of some electrolytes on clay, Trans. Am. Ceram. Soc. 16, 515 (1914). 7. Black and Veatch, Whipping chemicals into a colloidal water increases efficiency, Eng. Record, 72, 292 (1915). 8. Bleininger, A. V., The effect of electrolytes upon clay in the plastic state, Orig. Com. 8th. Intern. Cong. Appl. Chem. 5, 17. 9. Blitz, Wilhelm, Ueber die gegenseitige Beeinflussung colloidal Stoffe. Ber. cl. d. chem. Gesell. 37, 1095 (1904). 10. Burton, E. F., The action of electrolytes on colloidal solutions, Phil. M. 228, 12, 472 (1906). 11. Burton, E. F., The physical properties of colloidal solutions, p. 1, New York, 1916. 12. Catlett, George F., Colloidal theories, applied to colored water, reduce cost of chemicals. Eng. Record, 73, 741 (1916). 13. Crook, J. K., Mineral waters of the United States, p. 347, New York, 1899. 14. Clark and Lubs, The colorimetric determination of hydrogen ion concen- tration and its application in bacteriology, J. Bact. 2, Nos. 1, 2, 3. 15. Crum, W., Ueber Essigsaure und andere Verbindungen der Thonerde, Ann. d. chem, u. pharm. 89, 156 (1854). 16. Dole, R. B., The quality of surface waters in the United States, U. S. Geol. Surv., Water Supply Paper, 236 (1909-10). 17. Ellms. The coagulation and precipitation of impurities in water purifica- tion, Eng. Rec. 51, 552 (1905). 18. Foerster, F., Ueber das Giessen des Tons, Chem. Ind. 28, 733 (1905). 19. Fleming, W., Gerinnungsgeschwindigkeit kolloidaler Kieselsaure, Z. physlk. chem. 41, 443 (1902). 20. Fuller. Water purification at Louisville, Ky., and Cincinnati, Ohio. 21. Goldberg, A., Die Kieselsaure im naturlichen Wasser, in Alkalisch aus- gereinigten Kesselspeise Wassern und in konzentrierten Kesselwasser, Z. Nahr. Genussm. 27, 265 (1914). 22. Hardy, W. B., On the mechanism of gelatin in reversible colloidal systems, Proc. Royal Society (London), 66, 95 (1900). 23. Hardy, W. B., Colloidal solution, the globulins, J. Physiol. 33, 258 (1905). 24. Hardy, W. B., Structure of cell protoplasm, J. Physiol. 24, 180 (1899). 25. Hardy, W. B., Eine vorlilufige Untersuchung der Bedingungen, welche die Stabilitat von nicht umkehrbaren Hydrosolen bestimmen, Z. physlk. Chem. 33, 391 (1900). 26. Hillebrand, W. F., The analysis of silicate and carbonate rocks, U. S. Geol. Surv. Bull. 422, 13. 46 27. Japp, F. R., and Murray, T. S., Synthesis of pentaCarbon rings, J. Chem. Soc. 71, 148 (1897). 28. Lottermoser, A., Beitrage zur Kenntnis des Hydrosol-und Hydrogelbild- ungsvorganges, Z. physik. Chem. 60, 451 (1907). Ueberflihrung einiger Metalle in den Kolloi'dalen Zustand und Eigen- schaften derselben. Phys. Zeitung, /, 148 (1899). 29. Linder and Picton, Solution, and pseudo solution (sulphides), J. Chem. Soc. (London), 61, 148; 67, 63; 71, 568; 87 (pt. 4), (1906). 30. Mayer, A., Schaffer, G., and Terroine, E., Influence de la reaction du milieu sur la grandeur des granules colloi'daux, Compt. Rend. 145, 918 (1907). 31. Mellor, Green and Baugh, Studies on Clay slips, Trans. Eng. Ceram. Soc. 6, 161 (1906). , 32. Kahlenberg, L., and Lincoln, A. T., Solutions of silicates of the alkalines, J. Phys. Chem. 2, 77 (1898). \ 33. Kerr and Fulton, The effects of some electrolytes on typical clays, Trans, Am. Ceram. Soc. 15, 184 (1913). 34. Kuspert, F., Ein Demonstrationsversuch ueber colloi'dales Silber, Ber. d. d. chem. Gesell, 35, 2815 (1902). 35. Pappada, N., Sulla Coagulazione Dell' acido silicies colloidale, Gazz. chim. Hal., 33, II, 272 (1903), 35, I, 78 (1905). 36. Parmelee and Moore, Some notes on the mechanical analysis of Clays, Trans, Am. Ceram. Soc. II, 467 (1909). 37. Peale, S. C., Lists and analyses of United States mineral waters, U. S. Geol. Surv. Bull. 32, 35 (1886). 38. Pence, F. K., Trans. Am. Ceram. Soc. 12, 43 (1913). 39. Quincke, G., Ueber die Fortfiihrung materieller Theilchen durch strft- mende Elektricitat, Ann. physik. u. Chem. II, 115, 513 (1861) ; Tereschin,' S. Ann. Physik. Ill, 32, 333 (1887). 40. Rohland, Paul, Ueber die Einwirkung von Hydroxlionen auf Kaolinsus- pensionen, Z. Chem. ind. Colloide II, 193. 41. Rohland, Paul, Eine Kolorimetrische Methode zur Bestimmung der Kol- loidstoffe in Abwassern, Z. Anal. Chem. 52, 657. 42. Rohland, Paul, Das Verhalten der Tone und Kaoline gegen Hydroxylione, Biochem, Z. 58, 202, (1913-14). 43. Schultz, H., Schwefelarsen in wassriger Losung, J. F. prakt. Chem. (2) 25, 431 (1882) ; Antimontrisulfid wasserigen Losung 27, 320 (1883) ; ueber das Verhalten von seleniger zu schwefliger Saure. 32, 390 (1884). 44. Schwerin, Ein neues elektrishes Tonreinigungsverfahren (von Dr. M. Stoermer), Tonind. Zeitung, 36, 1283. 45. Thomas, Settling and filtering of fine clays, Trans. Am. Ceram. Soc. 14, 399 (1912). 46. Tirelli, L., Effect of free silica in water used industrially, Rass. min. 26, 134-36. 47. Tyndall, John, On the blue colour of the sky, and on the polarization of light, Phil. M. (4), 37, 384 (1869). 48. Van Bemmelen, J. J., Die Verbindungen einiger fester Dioxhydrate mit Sauren, Salzen, und Alkalien, J. prakt. Chem. 2d. Ser., 23, 324, 379 (1881). 49. Weber, E., The liquefaction of clay by alkalies, and the use of fluid clay casting in the ceramic industry, Trans. Eng. Ceram. Soc. 8, 11. 50. Whitney and Ober, The precipitation of colloids by electrolytes, J. Am. Chem. Soc. 23, 842 (1901). 51. Wiederrian, G., Ueber die Bewegung von Fliissigkeiten im Kreise der geschlossenen galvanischen Saule, Ann. d. Physik, u. Chem. II, 187, 321 (1852). 52. Zimmer, W. H., Hydrated silicic acid in kaolin and its effects in pottery bodies, Trans. Am. Ceram. Soc. 3, 25 (1901). 53. Zsigmondy and Siedentopf, Vortrage und Diskussionen von der 77 Natur- forscherversammlungen zu Meran, Phys. Z. 6, 855 (1905) ; ueber aml- krospische Goldkeime, Z. Wiss. Mikros, 24 (1907) ; Physik. Z. 8, 850 (1907). VITA. The author was born in Yates Center, Kansas, May 12, 1884, and received his early education in the public schools of Green County, Missouri, graduating from the Springfield High School in 1903. In the fall of that year he entered Drury College, Springfield, Missouri, and in 1907 received the degree of Bachelor of Science. For the term of 1907-08 he taught Chemistry and Physics in the Springfield High School. The following two years, 1908-1910, were spent as a graduate student in chemistry at the University of Pennsylvania. In Septem- ber, 1910, he entered the employment of the Iowa State College as assistant chemist for the Engineering Experiment Station and the Mining Department of that institution. In 1912 he was placed in charge of the chemical laboratory of the Engineering Experiment Sta- tion. He resigned that position in 1913 to accept another with the American Water Works and Electric Company of Xew York. In the fall of 1916 he left this company to enter the University of Illinois as a graduate student and assistant in chemistry, receiving' a degree of Master of Science in 1918. For nearly two years he has been chemist for the Illinois State Water Survey. THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $I.OO ON THE SEVENTH DAY OVERDUE. .'*N 26 1935 LD 21-100m-8,'34 UNIVERSITY OF CALIFORNIA LIBRARY