IRLF 112 QC 721 LIBRARY UNIVERSITY OF CALIFORNIA, RECEIVED BY EXCHANGE Class Ube mnix>ersft2 of FOUNDED BY JOHN D. ROCKEFELLER STUDIES IN RADIO-ACTIVITY A DISSERTATION SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE SCHOOL OF SCIENCE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (DEPARTMENT OF CHEMISTRY) t>7 f,i I UNIVERSITY I BY STEWART JOSEPH LLOYD 1910 TTbe mniv>ersit of Gbicaao FOUNDED BY JOHN D ROCKEFELLER STUDIES IN RADIO-ACTIVITY A DISSERTATION SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE SCHOOL OF SCIENCE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (DEPARTMENT OF CHEMISTRY) BY STEWART JOSEPH LLOYD UNIVERSITY OF 1910 Cte, 1 OF THE ( UNIVERSITY ) .IFOR Nil THE BETA ACTIVITY OF URANINITE BY STEWART J. I^OYD The percentage of the total a activity of uraninite con- tributed by each radioactive constituent of that mineral has been made the subject of investigation by Boltwood, * by McCoy and Ross, 2 and to a lesser extent by Mme. Curie. 3 Also the absolute maximum ionization current due to the activity of i gram uranium in the forni of an infinitely thin film has been determined by McCoy and Ashman 4 in con- nection with their work on uranium oxide as a standard of radioactivity. The present investigation was undertaken, at the suggestion of Prof. McCoy, with a view to ascertaining the percentage of the total ft activity to be assigned to each constituent of Uraninite, and, if possible, to determine by comparison with the a activity of uranium oxide the total ionization currenf produced by the /? activity of each. The latter object was however found impossible of realiza- tion. It was at first thought that possibly the 3 rays, like the w, might possess a definite fixed range beyond which they ceased to produce ionization, and that the apparent absorp- tion following the exponential law might be due to scattering alone; 5 that if the latter effect were eliminated, this range might be determined for each substance, and by the use of an electroscope of suitable form and sufficient size the total ionization current obtained, just as in the case of the a activity. With this in mind, a large cylindrical electroscope was constructed, in which the active material emitting the particles could be placed at the centre, by which arrangement the particles "bent out" of an ordinary p ray electroscope by the scattering due to absorbing material would be re- 1 Am. Jour. Sci., 25, 269 (1908). 2 Jour. Am. Chem. Soc., 29, 1697 (1907). 3 Comptes rendus, 126, 1101. 4 Am. Jour. Sci., 26, 521 (1908). 5 Crowther: Proc. Roy. Soc., 8oA, 186 (1908). 211029 5io Stewart J. Lloyd tained within the ionization chamber, and their ionizing effect included. Upon trial, however, it was discovered, as has recently been shown also by Makower, 1 that actual true absorption as well as scattering of the /? rays does take place, so that a determination of the absolute ionization current in a way comparable to that used for the rays seemed im- practicable. The measurements of p activity in this paper are therefore merely relative, and hold accurately only for the particular electroscope used. The order of magnitude will be the same, however, whatever measuring vessel may be employed, and the variations from the numbers here given will not be great, so that the latter may fairly be taken as representing the relative magnitudes of the /? activities of the different substances. The constituents of uraninite known to emit /? rays capable of producing appreciable ionization after passing through a sheet of aluminium foil 0.044 mm thick are U X, Ra B, Ra C, and Ra E 2 . In order to determine their respective activities with any degree of accuracy it was necessary to devise means of separating them quantitatively not only from the mineral itself, but from any considerable amount of foreign substances introduced during their extraction. Prac- tically no work has been done on the separation of the radio- active substances in a quantitative way, and very little on their general chemical properties, consequently the latter were studied in some detail, especially in the case of U X. In the following pages therefore, after a description of the electroscope used, there will be found measurements of the ft activity of uraninite, U X, Ra B, Ra C and Ra E 2 , referred to the amounts associated with i gram of uranium as a unit; descriptions of the methods found convenient for obtaining these substances in a condition suitable for measurement, and the results of a few other measurements of related in- terest. The electroscope used in practically all the measure- ments is that shown in Fig. i. It was designed especially 1 Phil. Mag., [6] 17, 171 (1909). The Beta Activity of Uraninite 511 for the examination of /? activity, and differs from electro- scopes commonly used for that purpose in allowing the active material to be placed either at the centre or at the bottom of the ionization chamber, instead of outside. It was made of galvanized iron, had a diameter of 40 cm 1 and the same height. As it was feared that the highest potential difference that could con- veniently be used, about 500 volts, would be insufficient, with such an instrument, to give a potential gradient high enough to produce the saturation current, four upright brass rods were attached to the base, each at a distance of 7 cm from the centre, and four similar rods, attached to the central elec- trode as shown in the figure, were placed between and outside them. In this way there was obtained a potential gradient of not less than 75 volts per cm throughout the ionization chamber, except at a very c - Char s in g wire - D. Branches of central electrode. few points. Separate experiments showed that with this arrangement the saturation current was being obtained with /? ray preparations of ordinary strength. The central electrode and the rods attached to it could be rotated at will. A door sliding vertically afforded access to the interior of the electroscope. The central elec- trode was made up of two sections screwing into each other, so that preparations could be measured, if desirable, at the centre. In most of the measurements, however, the active material was placed on the bottom of the ionization chamber. As a standard of activity, a film of uranium oxide pre- pared accord ng to the method of McCoy and Ashman 2 was employed. It may be of interest to note that the /? activity Fig. i E. Rods attached to base. A. Ionization chamber. F. Gold leaf electroscope. B. Central electrode. 1 The electroscope was cylindrical instead of square as shown in Fig. i. 2 Am Jour. Sci., 26, 521 (1908). 512 Stewart J. Lloyd of such a film (weighing 0.753 gram) measured in an ordinary a ray electroscope such as that used in this laboratory by McCoy and Ashman constitutes hardly 4 percent of the whole; in the electroscope just described it amounts to 12 percent. The ratio of the volumes of the two electroscopes is 1-9.5. Uraninite In measuring the /? activity of uraninite it is necessary of course to take account of the absorption due to the mineral itself, just as in the measurements of the a activity, though the effect is of less magnitude here. Hence three films of each of two samples of uraninite free from thorium, one con- taining 58.1 percent U, the other 45.1 percent were measured; the ratio of weight to activity plotted against the weight, and this ratio for an infinitely thin film, where no absorption takes place, determined graphically. 1 The activities are given in terms of the standard referred to above, divided by 100 to avoid decimals. As was first pointed out by Boltwood, 2 a small part of the emanation is spontaneously evolved when uraninite is powdered, and hence Ra B and Ra C are present in thin films in less than equilibrium amount. It will be shown later that Ra B and Ra C together contribute 52 percent of the total /? activity of uraninite. The amount of emana- tion and hence the /? activity lost was determined essential!) 7 as described by McCoy and Ross, 3 namely by boiling off in a mixture of nitric and sulphuric acids the emanation from a weighed quantity of the powdered mineral, sealing up the mixture to allow the emanation to accumulate again, boiling off and measuring once more, and calculating the maximum amount of emanation, using as period 3.75 days. Since the /? activity of the immediate emanation products Ra B and Ra C constitutes 52 percent of the whole, evidently the activity found requires to be increased by the activity of 52 percent of 1 McCoy: Jour. Am. Chem. Soc., 27, 402 (1905). 2 Phil. Mag., [6] 9, 603 (1905). 3 Jour. Am. Chem. Soc., 29, 1698 (1907). The Beta Activity of Uraninite the emanation lost in this way. This correction is applied in the following table. The a activity was excluded by covering the films with sheet of aluminium foil. The amount of /? activity cut off by the foil was determined graphically by adding successively i, 2, 3, and 4 layers of foil, and producing the curve, activity- foils, backwards. It was found that one sheet of aluminium foil 0.044 mm thick cut off 13.65 of the total /? activity of uraninite. Uraninite, 58.1 Percent U Weight Activity Em. lost Corr. act. w\a 1 a per wja g. uran- inite a per g. u. 0.2175 0.4100 0-6934 6.841 II .42 12.59 4.2 % of total emanation 7.01 11.68 12.87 0.0312 0.0352 0-0539 0.0307 32.6 56.1 Uraninite, 45.1 Percent U Weight Activity Em. lost Corr. act. w\a w\a Q g "-!n er inite g ' l 0.3398 0.4422 0.6650 7.713 9-365 10.23 3-8 % of total emanation 7.87 9-55 10.44 0.0431 o . 0463 0.0637 0.04 25-0 55-4 In the preceding tables w refers to the weight of the film of mineral, a to the activity thereof, and a to the activity of an infinitely thin film, determined graphically as described above. The average of the two values for a per gram U is 55.75. That is, the 1 activity of an infinitely thin film of uraninite containing i gram uranium and all the successive products of the latter in equilibrium amounts, is, in the electroscope described, 0.5575 times as great as the total activity of the standard film of uranium oxide employed. Uranium X As it has repeatedly been shown that uranium X is produced by uranium, and since uranium itself is readily 514 Stewart J. Lloyd determined quantitatively in uraninite, it seemed much simpler, instead of separating U X from uraninite, to separate it quantitatively from some pure compound of uranium which contained it in equilibrium amount. This separation may be effected to a greater or less extent in several ways; a quantitative separation, however, in which all of the U X is obtained without any uranium, and in addition sensibly free from other impurities, is distinctly difficult. The compound of uranium most readily obtained is the nitrate. Uranium X has been obtained from it in several different ways : 1. By dissolving uranium nitrate in ether, and separa- ting the aqueous layer, which contains most of the U X and a little U. This separation is not quantitative. 1 2. By precipitating uranium nitrate with ammonium carbonate and dissolving in excess, whereby the U X remains undissolved. This also is incomplete, and requires the presence of a considerable amount of impurity, such as iron, to make it of any use whatever. 1 3. By precipitating barium sulphate in the aqueous uranium nitrate solution, whereby U X is carried down mechanically. Three successive operations of this kind re- move all the U X. It was found possible also to remove the U X from the sulphate precipitate by boiling with hot concen- trated hydrochloric acid, but to obtain a complete extraction acid of such concentration had to be used that appreciable quantities of barium sulphate were also dissolved. 2 4. By stirring into an acetone solution of uranium nitrate some freshly prepared ferric hydroxide. Results obtained by this method proved very erratic, from 40-90 percent of the U X being removed, never more, even after several operations. 3 5. By boiling aqueous uranium nitrate with animal charcoal. 4 This method also was not quantitative. A 1 Crookes: Proc. Roy. Soc., 66, 409 (1900). 2 Becquerel: Comptes rendus, 133, 977 (1901). 3 Moore and Schlundt: Phil. Mag., [6] 12, 393 (1906). 4 Levin: Phys. Zeit., 8, 585. The Beta Activity 0} Uraninite 515 modification of it, in which soot was substituted for animal charcoal, and acetone for water proved to be quantitative, however, and entirely satisfactory. It was found, also, more convenient to stir the soot into the acetone than to boil it with the latter. The complete process employed in the separation was as follows: Five or 10 grams of uranium nitrate, containing about 50 percent U were dissolved in 150 cc of acetone, and stirred for thirty mimites with i gram of soot prepared by burning naphthalene. The soot need not be purified before using, as the colored hydrocarbons and other substances in it are removed later, and the fresher the soot the more efficient it appeared to be. The mixture of soot and acetone was then filtered, another gram of soot added to the uranium nitrate- acetone filtrate, and the stirring repeated. Three treatments of this kind suffice. Separate experiments showed that 95 percent of the U X was removed in the first treatment, and practically all in the first and second together. Upon rapid evaporation of the filtrate and measurement of the uranium oxide produced by ignition, scarcely a trace of /? activity could be detected. The three portions of soot were then united, and without washing boiled twice with dilute hydro- chloric acid for fifteen minutes at a time, the first boiling re- moving about 95 percent of the U X. This dissolves out from the soot the U X and the small amount of uranium nitrate which had adhered to the soot. The latter can not be re- moved even by prolonged washing of the soot with ether, alcohol or water. This solution was then evaporated to a volume of 100 cc, 0.05 gram iron wire added, and when the latter had dissolved, ammonium carbonate was added care- fully until the uranium had just redissolved. Separate ex- periments showed that this amount of iron was sufficient to carry down all the U X, and that no uranium remained in the precipitate. Without the iron the separation is quite in- complete. The ferric hydroxide-U X precipitate was then thoroughly washed with water, and dissolved in 10 cc of hydrochloric acid. The resulting solution was shaken three 516 Stewart J. Lloyd times with ether which had been saturated with gaseous hydrochloric acid, whereby the ferric chloride dissolved in the ethereal layer, and the U X with an inappreciable amount of the iron salt, remained in the aqueous, acid layer. If concentrated acid is used, and the ether is freshly distilled, this separation is quite sharp. The uranium X solution was then evaporated, finally on a gold plate, and its activity measured. Two samples, each containing 10 grams uranium nitrate, which upon analysis yielded 47.36 percent U, were treated as indicated above, and the resulting activity measured in the electroscope. The residue on the plate weighed in one case 0.0015 gram in the other 0.0027 gram, and had therefore no appreciable absorbing power for /? rays. As is well known, U X emits, besides its j rays, radiation of two kinds, "hard" penetrating rays, and "soft" rays which are unable to pass through 0.044 mm f aluminium foil. Since in the measure- ment of uranium itself these rays were excluded by the cover- ing of foil, it became necessary to determine here what per- centage of the total activity of the U X, as measured in this electroscope, was due to them. The preparation was there- fore covered successively by i, 2, and 3 layers of aluminium foil, and by extrapolation the required value found. It was found that 26.7 percent of the total ionization due to U X was contributed by these soft rays. Schmidt 1 and others have recently shown that these soft rays are not a rays of short range, as had once been supposed, but ft rays of slight penetrating power. The above measure- ment affords additional evidence of a distinctly different kind in favor of this view, as is shown by the following con- siderations. The ratio of the activities due to the two kinds of rays in an ordinary a ray electroscope is approximately as 7-3. Had the weaker radiation been a in character and of short range, its effect would not have increased in magnitude when measured in the larger electroscope, while the activity due to 1 Jahrb. fur. Rad., 5, 451 (1908). The Beta Activity of Uraninite 517 the penetrating /? rays increases about threefold (page 512). Hence in the present electroscope the ratio of the two activities would have been, had the weaker been a in character, about 7-1 instead of 3-1 as it is. The following table gives the data obtained for two samples of uranium nitrate. Allowance was made of course for the time elapsing between the separation of the U X from the uranium nitrate and the measurements. Weight of uran. nitrate grams Weight of uran. Activity | Activity of hard rays Activity of hard rays per g. U. 10 IO tnl 114.6 116.9 83-83 85.72 17.7 18.1 Mean of two values 17.9. Percentage of total ft activity of uraninite due to hard rays of U X, No. 32.1. The activity of one of these preparations was measured over several months with the object of redetermining the period of U X. Both hard and soft radiations decayed at the same rate, indicating a period of 22.4 days. The activity of U X was determined in still another way. Three films of uranium oxide were prepared, allowed to stand until they had grown, their maximum amounts of U X, and then measured, the absorption due to the oxide itself, and the a radiation, being allowed for as in the case of uraninite. The results were found to- agree essentially with those obtained from the measurements on pure U X extracted from pure uranium nitrate. Weight of U 3 8 Activity w\a w/a Activity per g.u. 0.625 0.694 0.782 8.941 9-734 9-731 0.0692 0.0713 0.0761 0.0675 17-45 518 Stewart J. Lloyd It was observed also that one layer of aluminium foil 0.044 mm thick cut off 8.7 percent of the /3 activity of U X (hard rays). Radio-Uranium In a recent article Danne 1 has published some measure- ments which may indicate the existence between uranium and U X of a ray less substance which produces U X, and to which he has given the name radio-uranium. This substance was obtained by him along with U X by precipitating barium sulphate in an aqueous solution of uranium nitrate. As some irregularities and peculiarities had been observed in the course of the present work on U X, it was thought worth while to attempt the detection of this new substance, in an indirect way at least. If such a substance exists, and has a period at all com- parable in length with that of U X, and if the uranium nitrate be freed at once from it and from U X, the rate of growth of the uranium nitrate in /? activity should be much different from what it would be if the uranium directly produced U X. Even if the uranium nitrate be only partially freed from this hy- pothetical substance, the recovery curve of /? activity should still be somewhat different from what has been regarded as the normal, the difference depending upon the amount of radio-uranium removed. Danne obtained his radio-uranium by precipitation of aqueous uranium nitrate with barium sulphate, and subsequent elaborate treatment of the precipitate. In the present experiment therefore, 10 grams of uranium nitrate were dissolved in water, and barium sulphate pre- cipitated in the solution twenty-five times successively. The uranium nitrate, from which, presumably at least, part of the radio-uranium had been removed, was then precipitated with ammonia, ignited to the oxide and made into a film. Measurements of the growth of /? activity extending over seventy days gave a period of 22.7 days, practically the same as that so frequently observed for U X. Hence, if such a 1 Le Radium, 6, 42 (1909). The Beta Activity of Uraninite 519 product does exist, and is the parent of U X, it is obviously not removed by barium sulphate. Indeed the phenomena observed by Danne could all be accounted for by the assump- tion that his uranium nitrate had not been freed entirely from radium. Chemical Behavior of Uranium X In the attempt to devise a suitable method for extracting UX from uranium nitrate some new and interesting facts with regard to the chemical behavior of the latter were obtained. As the chemical properties of the less abundant radioactive substances will probably assume considerable importance in the future, especially in the light of the possible position of the latter in the periodic table, these facts are given below. Uranium X, as has been observed by other investigators, resembles ferric iron very closely in all its reactions. It is precipitated quantitatively from its hot hydrochloric acid solution by the addition of a small amount of ferric chloride and ammonia. This reaction affords a means of determining the amount of U X in a solution which is much more accurate and rapid than the evaporation of the solution and subse- quent measurement. The amount of iron added need not be so great as to cause any appreciable absorption of the activity. When to a hydrochloric acid solution of U X, ferric chloride and aluminium chloride are added, the two metals precipitated by potassium hydroxide and excess of the latter added, the uranium X remains undissolved, with the iron, not a trace following the aluminium. Ammonium carbonate, when added to a precipitate of ferric hydroxide containing U X, dissolves a small amount of the iron when cold, and quite a considerable amount when hot. The U X is much less soluble in the carbonate than is the iron, being scarcely dissolved at all. The only way known by which UX can be completely separated from iron is by treatment with ether and hydro- chloric acid (page 516). This reaction is of importance, as the most convenient method of obtaining U X from a solution is by precipitation with ferric hydroxide. 520 Stewart J . Lloyd Uranium X when precipitated from a solution of uranium nitrate by barium sulphate is not dissolved out by boiling sodium carbonate, but remains with the resulting barium carbonate. Lead sulphate is not so efficient in carrying down U X as is barium sulphate, the ratio being 7-10. When lead sulphate containing U X is treated with sodium thiosulphate, the U X does not follow the lead, provided there is sufficient impurity present to form a nucleus of undissolved matter. Ammonium acetate removes from a lead sulphate U X precipitate about 35 percent of the U X in two extractions. The precipitation of basic ferric acetate in a solution of U X carried down 70 percent of the activity. Boiling acetic acid does not remove 'any pf the activity from a barium sulphate-U X precipitate. Radium B and C Rutherford's 1 examination of the immediate active de- posit from radium led him to the conclusion that radium C alone emitted /? rays. Certain irregularities in the decay curves however brought about subsequent investigations by Schmidt, 2 and by Bronson, 3 both of whom demonstrated the existence of /? activity in radium B, the latter showing that the /? activity of B probably exceeded that of C. The results of their work, so far as the existence of /? activity in radium B is concerned were substantiated by Duane, 4 using a very different experimental method. He measured the amount of negative electricity emitted by the active deposit instead of the ionization current due to it. In the present investigation the total p activity due to B and C together was obtained, and by an indirect method the ratio of the two activities, from which data the individual activities themselves were calculated. To determine the total /? activity of the active deposit of radium, that is, the /? activity of B and C together, portions 1 Phil. Trans., 198 (1904). 2 Phys. Zeit., 6, 897; 7, 764. 3 Phil. Mag., [6] 12, 73(1906). 4 Le Radium, 5, 65. The Beta Activity of Uraninite 521 of a radium solution were evaporated on platinum plates, allowed to stand for forty days, at the end of which time the activity had reached a maximum, and then measured in the electroscope, interference by the emanation and by the a activity in general being prevented by a closely adhering sheet of aluminium foil. The amount of ft activity absorbed by the foil was determined essentially as in previous similar cases. The radium films themselves were so thin that no ap- preciable absorption took place in them. In order to deter- mine the amount of radium of which the activity was being measured, portions of the same radium solution were com- pared in a gas electroscope with a weighed amount of uraninite whose uranium and hence radium content was known. Boltwood, 1 however, had previously shown that radium preparations in the form of thin films lose spontaneously quite considerable amount of emanation, and that therefore the maximum activity of a radium preparation was less than it should be, were all the emanation retained. Since a correction for the loss of emanation was made in the case of uraninite, and since in any event the losses for the two materials are by no means proportional, it was necessary to make a similar correction here. To do this, a portion of the same radium solution was evaporated on a small thin copper plate, the activity allowed to reach a maximum, the plate placed bodily in a small flask, and the emanation boiled off and measured in the usual way. The flask was then sealed up, the emanation allowed to grow for a few days, measured again, and the maximum amount calculated. Three determinations of this kind gave 10.7, 1 1. 6, and 11.4 percent of emanation lost, an average of 11.2. The correction is applied in the accompanying tables. | No. Act. Corr. act. Act. per g. U Average 1 i I 10. 61 12 .2 28.7 2 11.13 12.8 29.9 29.0 3 10. i 11.4 28.5 Phil. Mag., [6] 9, 603 (1905). 522 Stewart J. Lloyd Percentage of the total /? activity of uraninite contributed by B and C 51.8. The total /? activity due to the active deposit of radium was obtained in still another way, after the method used by McCoy and Ross 1 for obtaining the a activity due to emana- tion + A + B + C in uraninite. A portion of the uraninite sample containing 58.1 percent uranium was finely ground, treated with nitric acid, evaporated to dryness, and the process repeated three times at intervals of two hours, to free the material from emanation and to allow A, B and C to decay. After the last evaporation the residue was heated strongly enough to decompose all the nitrates present. The resulting residue, mostly oxides, was made into films as quickly as possible, and its activity determined in the usual way. The films were then allowed to stand until the maxi- mum ft activity had been attained, about 35 days. The percentage of the total /? activity contributed by the active deposit could thus be directly calculated. The results are contained in the following table. Initial ft activity (due to constituents other than BandC) Final activity (due to all con stituents ) Act. corr. for loss of eman. Act. of B + C 4.24 5.98 8.2 7 11.72 8.8 12 .46 4-56 6.48 Percentage of /? activity of uraninite contributed by B and C therefore 52.0. The /? rays from uranium X, radium B, radium C and radium E 2 are unequally absorbed" by the uranium oxide, and as no allowance for absorption was made in these measure- ments, the close agreement between them and those im- mediately preceding is fortuitous only. To determine the individual /? activities of Ra B and Ra C advantage was taken of the difference in the tempera- tures at which they volatilize. A copper plate kept negatively Jour. Am. Chem. Soc., 29, 1702 (1907). The Beta Activity of Uraninite 523 charged was placed in a vessel containing emanation, and allowed to remain until the deposit on it had reached a state of radioactive equilibrium. It was then removed, allowed to stand for twenty minutes to ensure the complete decay of Ra A, and the ratio of its a to its p activity measured. The P activity was due to B and C together, the a activity to C alone. Another plate which had been exposed in a similar way was heated for five minutes in an electric furnace to a temperature of 700 C. According to Makower, 1 Ra B volatilizes entirely at this temperature, while Ra C is un- affected. The ratio of the a to the p activity was then measured, both activities in this case being due to Ra C. From these two measurements the individual activities were calculated Let x == p activity of Ra C. y == P activity of Ra B. z = a activity of Ra C when all are in radioactive equilibrium. Then z/x + y == ratio of a to p activity after exposure to emanation and decay of RaA = m r z/x ratio of a to p activity after heating to volatilize Ra B = w 2 , m, m v The results of several measurements did not agree very closely among themselves, as might be expected from the fact that at the time of measurement, B and C are not present in equilibrium amount. In general, however, it appeared that of the total p activity due to B and C, radium C con- tributed about 68 percent, Ra B about 32 percent. Percentage of total activity of uraninite due to B is 15.9 ' C " 35-9 Radium E, The determination of the p radiation of E 2 was some- what more difficult than that of any one of the preceding Proc. Manch. Phil. Soc., 53, II, 1-8. 524 Stewart ] . Lloyd substances. Radium B and C depend immediately upon the easily estimated radium emanation, while uranium X may be grown readily from weighable quantities of uranium. Radium E 2 may, indeed, be obtained without trouble from Ra D, but the quantitative extraction of Ra D from uraninite had not been worked out. Possibly the most satisfactory though rather tedious way to obtain the desired result would be to obtain a strong preparation of Ra D, await the growth in it of E t , E 2 and F, and when equilibrium had been reached, measure the activity of Ra F, the only a ray product present. Since the activity of F in terms of uraninite has been deter- mined, 1 we should know the weight of mineral corresponding to the amount of Ra D in our sample and could then deter- mine E 2 with accuracy. No supply of Ra D with its products was available, however, so that it became necessary to study the quantitative extraction of the latter from uraninite, and to obtain Ra E 2 from it in such a manner as to allow of its activity being measured. A preparation of Ra D was made however, set aside to permit of the growth of Ra F, and will be measured later. In order to determine the conditions under which Ra D could be extracted quantitatively from uraninite, a study of its chemical behavior was made. As its name radio-lead would indicate, it clings very closely to lead, quite as closely as U X does to ferric iron, following the lead quantitatively everywhere, provided a sufficient amount of the latter is present. In the form of sulphate it dissolves quantitatively in ammonium acetate and in sodium thiosulphate, is pre- cipitated quantitatively by hydrogen sulphide, sulphuric acid, and by sodium carbonate. If the quantity of lead associated with it be small, however, some of the Ra D is lost. The one condition to be fulfilled in order that Ra D may follow quantitatively the reactions of lead appears to be the presence of considerable quantities of that metal. In obtaining Ra D from uraninite therefore, the follow- ing method was used. 1 Boltwood: Am. Jour. Sci., 25, 269 (1908). The Beta Activity of Uraninite 525 10 grams uraninite containing 58.1 per cent U were dis- solved in dilute nitric acid. The uraninite was practically free from sulphides, so that oxidation of the latter to sulphates was not feared. The residue, mostly silica, showed activity, which disappeared in a few hours. The solution was evaporated to dryness, 100 mg of lead nitrate added, and solution effected by the addition of water and a few drops of nitric acid. Dilute sulphuric acid was then added to precipitate the sulphates of barium and lead. The mixed sulphates were then ex- tracted repeatedly with ammonium acetate, to remove the lead and radio-lead, and the resulting solution precipitated with hydrogen sulphide. The filtrate from the barium-lead sulphates was also treated with hydrogen sulphide, and the resulting precipitate containing lead, bismuth, Ra F, etc., added to the former. The combined precipitate was con- verted into nitrates, then into carbonates, and then into nitrates again. The nitrates were precipitated by sodium hydroxide, enough of the reagent added to redissolve the lead, and the solution filtered. The lead, all of which was in the nitrate, was converted into a carbonate, then into a nitrate, and the solution evaporated to dryness. At this time it possessed no appreciable activity, and was set aside until Ra E 2 should have grown to maximum amount. Before measuring the activity of Ra E 2 it was necessary to separate it from the very considerable amount of solid matter, over 125 mg, associated with it, as the absorption of the soft rays of radium E 2 takes place readily. The separa- tion was made in the following manner. When to a lead nitrate solution containing radium D, E t , E 2 , and F, sodium hydroxide is added, complete pre- cipitation takes place, but upon addition of excess of the re- agent only the lead and Ra D redissolve, provided there is sufficient solid matter in the residue to afford a nucleus. The separation becomes quantitative when a drop of ferric chloride solution is added. The solid residue was filtered off, dried, and measured at once, so that no correction needed to be made for decay. 526 Stewart J. Lloyd Weight of uraninite Activity Act. perg. mineral Act. per g. Uran. IO 35-4 I 3-54 6.1 Percentage of total /? activity of uraninite due to Ra E 2 , 10.9. The ft activity of uraninite per gram uranium was found to be 55.75 in arbitrary units. The activities of the several constituents in the same units, and in percentages are Uranium X 17.9 3 2 -i% Radium B 9.0 16. i Radium C 20.1 36 . i Radium E 2 6.1 10.9 53-i 95-2% Whether the missing 4.8 percent is contributed by some constituent as yet unknown, or is due to experimental errors, is hard to say. It was expected that the total activities as given above would show some intimate relation to the absorbability of the various rays. Such is hardly the case however, as the follow- ing table 1 shows. The first column contains the thicknesses, in cm of aluminium foil, required to reduce the activity to half value in each case, the second column the total activities. Uran X 0.048 32.1 6.014 RaB 0.053 J 5-9 o . 0078 RaC 0.0534 35-9 0.0131 Ra E 2 0.016 10.9 Indeed the rays from any one substance do not appear to be homogeneous, so that to obtain comparable numbers for their absorbabilities it would be necessary to know much more accurately than we do at present the relative amounts of the hard and soft rays in each. It should be noted that account is taken only of those rays which produce appreciable ionization after passing through 0.044 mm of aluminium foil. It is quite possible, indeed probable, that other rays, like the soft rays of U X, 1 Le Radium 6, i (1909). The Beta Activity of Uraninite 527 exist, and in establishing the numbers of /? particles given off by each substance such rays would have to be taken into account. Hahn 1 has recently advanced the hypothesis that a single radioactive substance is capable of emitting rays of one kind only, either homogeneous a or homogeneous /? rays. If such is the case, the number of accepted radioactive substances will, obviously, have to be greatly increased, especially the number of those emitting ft particles. Although, as was stated at the beginning of this article, the absolute ionization current due to the /? activities of the constituents of uraninite could not be determined with ex- actness, it is possible, however, from the measurements made, to assign a lower limit to it in each case. According to McCoy and Ross 2 the total activity of i gram of uranium is equal to that of 796 square cm of a thick film of U 3 O 8 . In the electroscope used throughout the present work, the U X associated with i gram uranium in radioactive equilibrium was found to possess an activity equal to that of 7.53 sq. cm (a only) of a thick film of U 3 O 8 . Hence, the hard rays of U X in radioactive equilibrium with i gram uranium have an activity equal to that of 0.00946 gram uranium, and hence an ionization current of 4.36 io~ 12 amp. Since U X furnishes 32.1 per cent of the total ionization due to the /? activity of uraninite, the /? activity of the latter is equal to the a activity of 0.0295 gram of uranium, and gives an ionization current of 1.358 io~" amp. If the soft rays of U X be included, the figures for U X become 0.0129 and 5.95 X io~ 12 ; for uraninite 0.0329 and 1.516 X io~~ u . If U X is genetically connected with radium, and if the periods of the two be taken as 22 days, and 1760 years respectively, the weight of U X in equilibrium with i gram uranium is 1.116 X io~ 6 gram. Weight for weight therefore U X (hard rays) is at least 8.48 X io 7 as active as uranium. Including the soft rays it is 11.02 X io 7 as active as uranium. Kent Chem. Lab., Uni-v. of Chicago 1 Phys. Zeit., 9, 697. 2 Jour. Am. Chem. Soc., 29, 1698 (1907) THE ESTIMATION OF RADIUM BY STEWART J. LLOYD In the course of an investigation which involved frequent determinations, by the emanation method, of the amount of radium present in minerals and in solutions, some irregulari- ties were noticed, which led to the examination of the con- ditions necessary for the accurate determination of radium in this way. The facility with which radium may be esti- mated by means of its emanation, w T herein it is unique among radioactive substances, renders the method of considerable importance. Indeed so simple and convenient is the process that it bids fair to be used as an indirect means of determining substances other than radium, just as the iodine -thiosulphate titration is employed in the estimation of substances ranging from copper to chlorine. At present it undoubtedly furnishes the simplest method of determining uranium in minerals. Unlike most other methods of analysis, too, it has the ad- vantage that the sample is not destroyed in the course of the work, but may be re-examined any number of times. It is quite within the range of possibility, therefore, that the gas-electroscope will take its place in the analytical laboratory along with the polariscope and the refractometer. The method in question is well known, in outline at least, to all those concerned with measurements in radioactivity. The solution containing radium is boiled to expel all emanation present, sealed up for a definite time, usually several days, to permit the emanation to accumulate, and the emanation then drawn off into a gas-electroscope where its activity is measured. Since in a radium solution which has been freed from emanation the latter grows to half value in 3.75 days, it is possible to calculate the total maximum value of the activity from the formula i==i (i-*0 where "I" is the activity observed (the reciprocal of the time of discharge), "I " the maximum activity, "A" the The Estimation of Radium 477 decay constant, and "/" the time; and by standardizing the electroscope from time to time with a solution of known radium content, it is possible to determine the actual amount of radium present in any case. Rutherford appears to have been the first to use this method. After him, Boltwood, McCoy, Joly, and numerous others have practiced it with various modifications. The present article is not concerned with the mode of transferring the emanation from the vessel containing it to the electro- scope, which has been the point of difference between them, but with the effect which the state of the solution exercises upon the giving up of the emanation. It had been noticed that the presence of foreign sub- stances exerted a marked effect on the accuracy of the esti- mation. To test the extent of this effect and to determine the conditions under which accurate results might be obtained, the following experiments were made. A dilute radium solution prepared from uraninite residues, and therefore containing barium, was made up, and 10 cc of it placed in each of several 150 cc flasks, 100 cc of water added, the solution boiled vigorously for fifteen minutes, and then small quantities of the following reagents added, one to each flask. The flasks were then tightly stoppered, and allowed to stand for several days. The emanation pro- duced in that time was then measured and the maximum amount calculated. The results are found in the following table. No. Reagent Time of discharge for maximum Sec. I H 2 S0 4 35-6 2 HC1 18.4 3 HNO 3 18.2 4 Na 2 CO 3 27.7 5 K 2 CrA 19.0 6 KOH 19.6 7 MoO 3 18.7 8 HP 19.6 9 Hg 23-3 478 Stewart J. Lloyd From this table it is evident that only in the presence of hydrochloric or of nitric acid is the emanation completely evolved. Sulphuric acid and sodium carbonate, which pre- cipitate the sulphates and carbonates of barium and radium, hinder very noticeably the evolution of emanation. The other reagents have but slight effect, mercury more than the rest. To determine whether other reagents which effected precipitation in the solution had the same effect as sulphuric acid and sodium carbonate, some of the solutions which had been previously used, after boiling to remove emanation, were treated, respectively with the following reagents, sealed up for some days and the emanation measured as before. Discharge To No. i was added BaCl 2 producing BaSO 4 27.35 AgNO 3 " AgCl 18.2 CaCO 3 CaCO 3 24 . 9 BaCl 2 BaCO 3 50.2 Pb(NO 3 ) 2 " PbCr6 4 19.1 AgN0 3 . Ag,0 18.5 It will be seen from the table that the production of a precipitate in a solution does not necessarily mean an impair- ment of the accuracy of the determination. Only in the case of sulphates and carbonates is the interference with the evolu- tion marked. Further experiments were therefore made to determine if possible the cause of this difference. Four similar radium solutions were made up, sealed in the usual way, and the emanation measured after an interval. Discharge Sec. No. i contained no H 2 SO 4 "2 " 5 cc H 2 SO 4 (dilute) 3 " " " " plus BaCl 2 4 Na 2 CO : 54 63 149 74 The radium solutions of course contained originally a small amount of BaCl 2 . It is apparent from the table that The Estimation of Radium 479 the greater the amount of BaSO 4 , formed the less emanation is given off. These experiments did not however throw any light upon the mechanism of the retention of emanation, so another series was made. To each of four radium-barium chloride solutions were added equal amounts of BaCl 2 , and then equal amounts of H 2 SO 4 (excess) were run in, with stirring, from a burette, under identical conditions except that No. 2 was precipitated cold, the others hot. 1. The pptd. BaSO 4 was filtered immediately, and both the precipitate and the filtrate sealed up at once. 2. As No. i except that precipitation was made in cold. 3. Stirred continuously for thirty hours, after precipita- tion, then treated as i. 4. Heated continuously for thirty hours after precipita- tion, then treated as i. After a suitable interval the radium contents of filtrates and precipitates were determined. No. Residue Sec. Filtrate Sec. I 500 I2OOO 2 420 11400 3 144.2 10800 4 133-6 I 1 100 This table indicates that in every case practically the same amount of radium remains unprecipitated by the sul- phate, but that the readiness with which the emanation is given off from the part that is precipitated depends very largely upon the treatment to which it is subjected, stirring and heating, especially the latter, facilitating the evolution very markedly. A similar set of experiments involving the precipitation of BaCO 3 instead of BaSO 4 was made. Sodium carbonate in excess was added to each of three radium-barium chloride solutions. No. i was filtered at once. Stewart J. Lloyd No. 2 was heated for fifty hours and then filtered. No. 3 was stirred for fifty hours and then filtered Upon examination for emanation they gave: No. Residue Sec. Filtrate Sec. I 348 2970 2 144 I2OO 3 339 2724 The results here are quite analogous to those obtained Avith BaSO 4 except that stirring does not seem to be so effi- cacious, and the precipitation of radium by the carbonate is not so complete. A further experiment was made to determine what effect the physical condition of BaSO 4 has upon the retention of emanation. A hydrochloric acid solution of radium which gave an emanation content corresponding to a time of dis- charge of 1 60 seconds was precipitated with sulphuric acid and barium chloride, and its emanation content measured at intervals of four days. The times of discharge correspond- ing to the maximum amounts were: Days Sec. 4 8 12 16 20 370 J75 165 157 163 BaSO 4 previously made and heated for some time was then added to the solution and thoroughly shaken with it. The time of discharge was not however affected, remaining 163 seconds. To ascertain whether or not the total amount of emana- tion was finally obtained from solutions in which sulphates had been precipitated, two solutions containing the same amount of radium, the first acid with hydrochloric acid, the The Estimation of Radium 481 second having a heavy precipitate of BaSO 4 produced in it, were measured from time to time. Days HC1 solution Sec. BaSO 4 solution Sec. 4 57 167 9 58 87 13 56-5 62 17 58 59 22 58.5 58 Conclusion. For the accurate determination of radium by the emana- tion method the presence of HC1 or HNO 3 is necessary. If the sulphate or carbonate of barium is present in the solution, prolonged boiling or repeated determinations will be necessary to ensure the extraction of the total emanation. The retention of emanation by the freshly precipitated BaSO 4 is doubtless mechanical only, and is due to the en- tanglement of the radium chloride or sulphate within the fine almost amorphous precipitate. Recrystallization under the influence of heat releases the radium, and permits the removal of the emanation. Kent Chem. Lab., Univ. of Chicago OF UNIVERSITY Of The preceding work was done at the suggestion and under the direction of Prof. H. N. McCoy, whose encouragement and advice are gratefully acknowledged by the author. 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 5O CENTS ON THE FOURTH DAY AND TO $1.OO ON THE SEVENTH DAY OVERDUE. APR 1 g 184? LH 21-100w-7,'39(402s;