T H E COXI,EECET\'CE OF AT\' UNFILTER.4BLE PRECIPITATE OF BARIF31 SULFATE* BY H. M. TRIJIBLE'"
Introduction It has long been known that a precipitate which has just been thrown down from solution, and which is so fine as t o pass through an ordinary filter, may be made filterable by digesting it for some time at an elevated temperature in contact with the mother liquor from which it was separated. Ostwald' s a s apparently the first to advance an explanation for this phenomenon. Cnder these conditions, he states, the smaller particles dissolve and the larger mes grow by deposition, so that finally a precipitate results which is sufficilntly coarse to filter well. This explanation was an application of the prin:iple enunciated by Sir William Thomson.? later Lord Kelvin, according to svhich the vapor pressure or solution pressure of a substance in the form of very small particles varies inversely as the size of its particles. Some work (soon to Le published), carried out by the author with Professor S. Lawrence Bigelow indicated that, while it is undoubtedly true that the vapor pressure of a small particle is greater than that of a larger one, yet this difference is vanishingly small for particles of measurakle size. I t was found, too, that any transfer of material from smaller t o larger particles due to such difference is exceedingly slow. By analogy it seemed that the rate at which larger crystals of a nearly insoluble substance ill grow at the expense of smaller ones in contact with its solution must be very low; probably too low t o account for the coalescence of precipitates as it occurs in ordinary analytical practice. Evidence is not wanting to show that this is, indeed, the case. G.A. Hulett,a in the course of a study of this very matter, added three grams of very finely pulverized gypsum to z j o cc. of a solution which had been saturated over large plates of this material and agitated the preparation gently a t constant temperature. The concentration of the dissolved electrolyte was determined from time to time by measuring the electrical Conductivity of the solution. The small particles were, on the average: about 0.3 microns in apparent diameter a t the start of the experiment. The conductivity rose t o ' a maximum almost immediately, and then fell slowly. At the end of 48 hours it had fallen nearly to the conductivity of a normally saturated solution; but it was only after nine days that this value was actually reached. Parallel * Contribution from the Chemistry Laboratory of the Cniversity of Michigan. * * T h e work presented in this paper is taken from a dissertation presented in partia fulfillment of the requirements for the degree of Doctor of Philosophy in the University of hIichigan. T h e investigation was carried o u t under the direction of Professor 9.Lawrence Bigelow. ' 11. Ostwald: ".lnslytische Chemie" pp. 14, z z (1894). Thomson: Proc. Roy. Soc. Edinburgh. 7, 63 (1870). 3 G .A. Hulett: 2. physik. Chem., 37, 385-405 (1901);47, 357-357 (1904).
602
H. M. TRIMBLE
with this decrease of the conductivity, which indicated a decrease in the concentration of the solution, went the disappearance of the smallest particles. At the end of nine days they had disappeared completely, and the smallest particles t’hen present were approximately two microns in diameter. Hulett concluded that above this size, the solubility of gypsum is no longer a function of particle size. It is noteworthy that equilibrium was approached very slowly during the latter part of this experiment. If we accept the explanation of coalescence which Ostwald advanced, it follows that the rate of elimination of the’last of the smaller part,icles is very low. Hulett found in similar experiments that barium sulfate behaved in the same manner as did gypsum. In particular he found that, at z5”C., its solubility is no longer a function of the size of its particles if these latter are larger than about’ I .8 microns in diameter. Von Weimam’ found that colloidal barium sulfate is converted to regular crystals in the course of time. Thousandth-normal solutions of barium thiocyanate and manganous sulfate were mixed in the cold, and the precipitate was allowed to stand quietly in contact with its mother liquor without stirring or shaking. His description follows: “During the first few hours after the appearance of the opalescence the greater part of the precipitate consists of particles which seem to be spherical under the highest magnifications of the microscope. After twelve hours the part of the precipitate n-hich has settled upon the bottom of the vessel consists of tiny crystals in the form of parallelopipeds of different, sizes, and of aggregates of spherical particles (as seen under a magnification of I joo times). After a month it is readily seen that the crystals have grown; but the aggregates of spherical particles still constitute an important part of the precipitate. It requires a half year for the aggregates to disappear almost completely; and at that lime the precipitate consists of characteristic barium sulfate crystals.’’ The individual units in the aggregates of spherical particles cannot be seen in the photographs as given. I t seems certain that they were very small.
If the colloidal material of this experiment was amorphous barium sulfate, it constituted an unstable phase in the precipitate. The difference in solution pressure between a n unstable and a stable form of this substance is certainly greater than any such difference between particles of comparable sizes in either phase. The transfer of material in a system containing these tv:o phases should be more rapid than in one which contains only crystals of different sizes. But von Weimarn holds that all mattei, even in the colloidal state, is crystalline. If this is true, the system with which he worked was the same in nature as that which we are considering. The question cannot be decided by microscopic examination, since very small particles of any kind show up as tiny discs of light, quite without distinguishing features. No matter which view we take, however, the experiment furnishes direct 1
P. P. yon Weimarn: Kolloid-Z., 3, 289-90 (1908)
COALESCENCE O F BARIUM SULFATE
603
evidence that, the growth of larger particles of barium sulfate a t the expense of smaller ones, at, room temperature and without stirring, is a very slow process. But for lack of space, various other observations of this kind might be cited. In every case, so far as t’he author is aware, the changes studied took place at or near room temperature. But digestions for the purpose of rendering an unfilterable precipitate filterable are invariably carried out at elevated temperatures. I t seemed worth while, then, to determine whether the same results would be found at these higher temperatures.
Experimental Part Obviously a precipitate can be made filterable by digestion only if all particles small enough to pass through the pores of the filter can be eliminated in this way. The average size of the pores in ordinary thick filter paper has been found to be’ about 3.3 microns. This value was calculated from the permeability of this paper to water, and may or may not indicate the magnitude of the actual maximum pore size. A number of esperiments were run, accordingly, to determine the size of the largest particles which will pass through such filter papers as are commonly used in quantitative analysis. Findy pulverized baIium sulfate was suspended in water and this suspension was poured into filters made up in the usual way using three well-known kinds of filter paper. The part of the precipitate which passed through was collected and its particles were examined under the microscope. In every case particles as large as three microns in diamet,er were not at all uncommon, and some which were as large as four microns were found. The maximum pore sizes, then, must be very much as calculated from the permeability to water. I t seems safe to set somewhat more than four microns as the size of the smallest particles which will certainly be retained by ordinary quantitative filter paper. But a filler is by no means a sieve. Filter paper is a colloidal material which is usually negatively charged. As such it tends to hold barium sulfate particles of any size. since they are positively charged when prepared, as in quantitative analysis, by precipitation from dilute acid solutions of sulfates by the addition of barium chloride. Due to this and other causes its pores soon become clogged and its efficiency is thereby increased. Filters of other materials might be used in separating particles of different sizes: but at the best the retaining power of any given filter is only very imperfectly reproducible. While, then, filterability serves as a convenient practical criterion as to particle size, it is not a very certain or useful one. The direct measurement of particles with a microscope, using a filar micrometer, is much better. The unavoidable error which is involved may be reduced to a minimum by measuring a large number of particles from each precipitate which is examined. This method was used in the present study, supplemented by numerous tests as t o filterability using filters of ordinary quantitative filter paper. “Third Report o n Colloid Chemistry.” B r i t . Ass. Adv. Sci., L. p . 68 (1920)
604
H. M. TRIYBLE
Barium sulfate is perhaps more frequently digested to render it filterable than any other precipitate which is handled in quantitative analysis. It,s low solubility should make the growth of larger particles at the expense of smaller ones very slow. But it is found in practice that a few hours digestion at a high temperature in the presence of a small amount of hydrochloric acid, without stirring, will render a precipitate of this substance readily filterable, provided it has been prepared in the proper manner. For this reason it seemed the most suitable subPtance for study. It was used in h o s t of t,he experiments which are described below. The results have been confirmed, however, by experiments in which barium chromate, lead chromate, lead sulfate and calcium oxalate mere used. The experiments had for their purpose to determine whether particles of these substances smaller than about four microns in average diameter can be eliminated from precipitates cont,aining them by digestions such as are carried out in ordinary quantitative analysis. Rhombohedral crystals of barium sulfate about ten to twenty five microns in length were prepared by slow addition of barium chloride solution to a hot dilute solution of sulfuric acid containing a sinal1 amount of hydrochloric acid. Small particles were prepared for the first experiments by crushing this material in an agate mortar, suspending the powder in a large quantity of water and pouring off the suspension which had not settled out after standing quietly for one or two minutes. These particles were recovered by allowing the suspension to settle for about an hour and then pouring off the supernatant, liquid. This left a heavy suspension which was used at once in an experiment. I t was impossible to filter these precipitates clear. As measured under the microscope t,he individual particles were found t'o be between one and four microns in diameter, the smallest ones being vastly in the majority. Particles smaller than one micron were few. About one-tenth gram of larger rhombohedral crystals was added to a port,ion of this suspension which contained one fourth to one half gram of material. The volume was then nihde up to about joo cc., hydrochloric acid was added, and the preparation lyas' dividcd into approximately equal portions after stirring it up. One portion was then allowed to digest a t about, 9 j oto IOO'C.without stirring. The other was heated under the same conditions, except that it was stirred during the progress of the experiment by means of a mechanical stirrer. K a t e r and a little hydrochloric acid were added from time to time as needed to replace the liquid lost by evaporation. S i n e experiments were run in this manner. The precipitates which were digested wit,hout stirring fi1t)eredclear after two to four hours in every case. llicroscoplc examination showed that the small fragments which were introducfid at the start had, in these experiments, formed aggregates. Their average size remained almost unchanged. The other preparation which had been digested with stirring would not filter clear even after three days or more. Microscopic examination showed. that the smallest particles had disappeared during this digestion. KOparticles smaller than about' two microns in diameter could be found in the precipitates. The size of the larger rhombohedral crystals lyas not appreciably changed in any case.
COALESCENCE O F BARIUM SULFATE
605
In ten similar experimenh suspensions of barium sulfate which had been prepared by mixing normal solutions of barium chloride and sulfuric acid in the cold were used. Such volumes of the solutions were taken that about one gram of barium sulfate was thrown down. The particles which fornied were all very small; none of them being larger than one micron in apparent tliameter. The larger rhombohedral crystals which were present (luring t'he digestions did not change appreciably in size. The portions iyhich were digested without stirring coalesced more slowly than before: tint fire t o ten hours sufficed to render the precipitates filterable in every case. At this stage the particles were again aggregated together. The apgrepates \yere all made up of unit particles two microns or less in diameter. In the portion5 which were digested with stirring all the smallest particles disappeared. so that :it the end of three days the smallest ones present were two microns or more in diameter. The precipitates could not be filtered clear in any case. thoupli four digestions were carried on for two weeks in the hope of attaining this end. It was, of course, impossible tmodetermine whether or not the total amount of fine precipitate had diminished. Even a t the end of two weeks digestion the smallest, particles present were still about two microns in diameter Xumerous experiments were carried out in the atttnipt to grow crystals a t the expense of finely divided and apparently amorphous barium sulfate at' elevated temperatures. The precipitates were prepared by mixing in the cold equivalent volumes of barium chloride and sulphuric acid. The concentrations of these solutions varied between one hundredth normal and twice normal. The amount of the precipitate at the start, the concentration of hydrochloric acid a t the start and the time of digestion Tyere all varied lietween rather wide limits. The digestions were carried out both with anti without stirring, but large crystals were not introduced in any case. The precipitates were examined from time to time with a microscope, using a mnpification of 1000 diameters, but no crystals which were recognizable as such were ever found. The largest particles before digestion were less than one micron in diameter. After digestion for some days they were between two and t,hree microns in diameter, on the average. Those precipitates which were digested without stirring became filterable after a day or two in sonie cases. I n others t,hey could not be made filterable. S o n e of those which were digested with stirring were rendered filterable. Ammonium acetate and sulfuric acid mere introduced in a number of experiments instead of hydrochloric acid. Digestions were carried out both in the presence and in the absence of large rhombohedral crystals of barium sulfate. It was hoped that these solvents might aid in the growth of larger particles a t the expense of smaller ones. This hope was realized, in that all particles smaller than about two microns disappeared after a few days digestion. Particles of this size, however, could not be eliminated by further digestion. I t seemed that any loss of matter by the smaller particles to larger ones in these experiments might be revealed if t,he total quantity of the former was very small. I n some further experiments, therefore, just enough barium
606
H. hl. TRIMBLE
chloride and sulfuric acid were added to distilled water to give a faint opalescence. The particles which were present at this stage were not measured, Large rhombohedral crystals were then introduced and the preparations were digested hot, with stirring. Hydrochloric acid, ammonium acetate and sulfuric acid were introduced in different experiments as solvents. .\fter two or three days digestion the precipitates were collected and studied. The very small particles which at first formed the opalescence had disappeared and in their place were many particles about two microns in diameter. The small particles were not removed entirely, as might have been expected. The findings of t'he earlier experiments were thus confirmed.
Discussion of Results This work shows that particles of barium sulfate larger t'han two microns in apparent diameter cannot be made to go over to yet larger ones, even by protracted digestion a t a high temperature. Measurements of large crystals before and after digestion, too, showed that their growth is inappreciable. It follows, then, that the coalescence of an unfilterable precipitate which renders it filterable cannot be explained in t,erms of the growth of 131yer particles at the expense of smaller ones. But we must still account for it. for the change undoubtedly takes place. The fact t,hat, on digestion without stirring, aggregates form; while they do not form when the digestion is accompanied b y stirring shows the way to the answer. In view of all the facts the best explanation seems to be as follows: The precipitated particles when allowed to settle come into intimate contact with one another and form aggregates which adhere rather firmly. As the digestion proceeds barium sulfate, either from the still supersat'urat'ed solution or from the dissolving of very small particles, deposits upon these aggregates and cements them together. This effect is enhanced when the solution is allowed t o cool. In th;s way large bodies which will not pass through the porm of a filter are formed. On the other hand, when the digestion is arcompanied by stirring all the crystals present. grow alike at the expense of any material which may tie separating from solution. There is no opportunity for cemented aggregates of crystals to form. Thus, while larger crystals do grow at the expense of smaller ones, if the latter are less than about two microns in diameter, the amount of such growth is very small; so small, indeed, that it could not render filterable a precipitat,e which is otherwise unfilterable. Summary The resulk of this study show that the solution pressure of bariutn sulfate a t about I O O O C . ceases to be a function of particle size for particles larger than about two microns in diameter. The coalescence of barium sulfate precipitates on digestion which renders them filterable cannot be adequately explained in terms of the growth of large particles at the expense of smaller ones. It is brought about, rather, by the collection of particles into relatively large aggregates, followed by the cementing together of the unit particles in the aggregates.
