COAGULATION OF METAL SULPHIDE HYDROSOLS. PART I. 461 Y XX.V.-Coag?clation qf Metal Sulphide Hydrosols. Payt I. Injhence of Distance b e t d e n the Particles of’ Q Sol on its Stability. Anomalous Pwtectiw Action of Dissolved Hydrogen Sulphide. By J%ANENDRA NATH MUKHERJEE and NAGENDRA NATH SEN. THE coagulation of the sulphide sols has been studied by a fairly large number of investigators. There is however fundamental disagreement between the results obtained by different authors. It was suggested by one of us in a previous paper (17. Amer. Chem. Soc. 1915 37 2024) that the discrepancy is due to the difference in the methods of observation some of which are undoubtedly defective. The method used in that paper is a comparative one, and is based on visual observation of the changes in the sol with time.It will be conceded that no objection can be taken against it, although it has one disadvantage in so. far as it is not instru-mental. This does not in any way interfere with the trend of the results. The method is found to be the most suitable one and gives concordant numbers. Contrary to the observations of Freundlich (Zeitsch. physikd. C’hern. 1903 44 129) it’ was shown that dilution with pure water increases the stability of arsenious sulphide hydrosols to coagulation by electrolytes. The electrolytes studied were all salts of uni-valent cations. The difficulty in explaining the observed facts on the basis of the adsorption theory as developed by Freundlich (Zoc. c i t . ; ibid. 1910 73 385; 1913 83 97; 85 398 641) was men-tioned.This point has also received attention from Kruyt and Spek. (Prnc. I<. Alcacl. T?7etensch. Amsterdam 1915 17 1158) who examined three electrolytes namely the chlorides of potassium 462 MUKHERJEE AND SEN COAGULATION OF barium and aluminium but could not find the stabilising influence of dilution of the sols for ions other than potassion. They do not seem to recognise that the adsorption theory as it stands is in-sufficient to explain all the facts observed and do not consider the influence that the distance between the particles of the sol may have on the stability of the sol. Further Yonng and Neal ( J . Physical Chem. 1917 21 14) in a thorough study of cupric sulphide hydrosols remark “the amount of electrolyte required is independent of the dilution of the sol within wide limits.This latter was found to be true within rather close limits by Freundlich for arsenic sulphide sols.” The method used by Young and Neal consists in mixing equal volumes (2 C.C. each) of electrolyte and sol and noting the respective con-centration of electrolyte that just produces complete separation of the colloid in twenty-four hours and that which just fails to do so. These two limiting concentrations give a measure of the stability of the sol o r the coagulative power of the electrolyte. This method is one due to Freundlich and generally used by other workers. Ti. will be noted that the concentrations of electrolytes employed are, necessarily such as would require a fairly long time for the com-plete separation of the colloid.I n the earlier paper (Toc. f i t . ) the process of co’agulation was discussed in detail and i t was statetl that “the time for complete settling is not vharacteristic of t . 1 1 ~ rate of coagulation.” The justification of ally rriethod lies jn so far as it indicates the progress of coalescence. The increased mass of particles wit-h progress of coalescence introduces a new factor, namely their gravitatioiial effect which masks the true behaviour of the sol as will be clear from the following observations on mercuric sulphide sols. These sols are opaque unless very di1ut.e. On the addition of electrolytes there is a quiescent period followed by a sharp clearing of the whole liquid. At this stage the liquid loses its homogeneous appearance and visible clots are found suspended throughout the liquid.As the change is sharp the times noted by different observers agree satisfactorily. I n this way it is found that a mercuric sulphide sol on saturation with hydrogen sulphide, requires a longer time for the observed change than when it is not so treated-the electrolyte concentration of coarse being identical. These experiments leave no doubt that dissolved hydrogeii sulphide iiicreases the stability of the sol. The subsequent settling of these clots however requires a very long time which is about the same for both samples. So long as the respective times required for the clearing of the sols are very small compared with the time required f o r the subsequent settling of the clots i t is found tha METAL SULPHlDE 1IYI)ROSOLS.PART I. 463 the times that are necessary for the complete separation of the colloid do not differ much in the two cases. However with electro-lyte concentrations where the clearing requires intervals comparable with that required for the subsequent settling of the clots regular differences in stability are observed even if the times necessary for complete separation of the colloid are noted. ITlufEuence of DiLutiom om the Stability of a Sol. The method employed is to mix 5 C.C. each of sol and electrolyte in carefully cleansed test-tubes. For reasons discussed in the previous paper the electrolyte is always added to the sol. Thorough mixing is secured by pouring the mixed liquids from one test-tube to another and repeating the process. The mixed liquid is shaken regularly a t short intervals and the changes with time are observed.As before the times required for perceptible change in the sol for the attainment of maximum opacity and for the first appearance of visible particles are noted. The time for com-plete separation of the colloid is also noted when thought desirable. The comparison is always carried out side by side. For sols of mercuric and cupric sulphides the time for the appearance of visible particles is noted. With dilute sols it is necessary to note the time for complete settling of the particles. Experiments were made on arsenious sulphide sols with solutions of hydrogen ammonium potassium lithium barium and aluminium chlorides aluminium sulphate and thorium nitrate. With sols of mercuric and cupric sulphides solutions of potassium, ammonium and barium chlorides aluminium sulphate and thorium nitrate only were studied.All the glass vessels were cleansed by dipping them for twenty-four hours in chromic acid d u t i o n after they had been washed with dilute alkali hydroxide. The need for scrupulous care i n the washing of the vessels and avoiding dust or other impurities cannot be too strongly emphasised. This holds especially for the extremely dilute solutions used in some cases. (a) Arseniozrs Sztlphide Sols. In the presence of salts having univalent cations dilution of the sol increases its stabilit'y in each case. The magnitude of the stabilising effect of dilution will be evident from t,able I. The electrolyte concentrations given are end concentrations that is, what results after mixing.Sols mentioned in the several tables are diff erent unless stated otherwise. Coagulation " means the T 464 MUKHERJEI!! ANb SEN COAatJLATION OF breaking up of the colloid into flakes 8 0 that the liquid is clear or very slightly coloured. T-LE I. Sol A contained 17.58 millimoles of arsenious sulphide per litre. Sol B was prepared by diluting sol A five times and sol C by diluting sol L4 ten times with pure water. Electrolyte lithium chloride. Dilution (after mixing). Sol. A. Sol. B. 5Nf16 ... - -5Nf32 ... c Coagulation after 30 seconds. Coagulation after N / 8 . . . . . . Instantaneous coagulation. half an hour. N/16 ... Change perceptible Change just after on mixing. mixingnot per-Coagulation not ceptible.observed after 1 & hours. Sol. 0. Coagulation af tor 20 seconds. Coagulation after 50 minutes. Perceptible change after 8 minutes. Coagulation after 2$ hours. Change perceptible after 45 minutes. On the other hand in the presence of the salts of aluminium and thorium the stability decreases on dilution as will be seen from table 11. The data refer to the same three sols. The observations in tables I and I1 were completed within two days, and neglecting the slight “ ageing” during t.his interval the data may be taken as comparable. TABLE 11. Electrolyte thorium nitrate. Dilution. Sol. A. N / 10,000 Instantaneous coagulation. N/20,000 Perceptible turbidity just after mixing, Sol cha.nges N/30,000 Perceptible turbidity after half an hour.slowly. N/40,000 -Sol. B. Sol. c. Instantaneous Instantaneous coagulation. coagulation. Coagulation i n 2 Coagulation within minutes. half a minute. Perceptible turbidity Coagulation in 4 - Coagulation in 53 after 5 minutes. minutes. IlliRUttXl. Solutions of salts of bivalent cations display an interesting aspect With dilute of the effect of dilution of t.he sol on its stability METAL SULPHIDE HYDROSOLS. PART I. 465 sols the stability increases on dilut,ion whereas with sols very rich in sulphide content the stability diminishes on dilution and for a rich sol it is possible to reach a limit where on further dilution t~he sol becomes more stable. Moreover from table I11 it will be seen that the stability relations on dilution vary with the concen-tration of the electrolyte itself.TABLE 111. Electrolyte barium chloride. Arsenious sulphide sol containing 19-45 millimoles per litre. Dilution of N/800 ... Complete coagula-tion in 1 minute. Electrolyte. Original Sol. N/1000 . . . Change perceptible after half a minute. Clots appear through-out after 17 minutes. Nil200 . . . Change perceptible in 2 minutes. Clots appear after 1 hour 7 minutes. Sol diluted 4 times. Change perceptible in 20 seconds. Coagulation after 4 minutes. Change perceptible in 1 minute. Clots appear throughout after 18 minutes. Change perceptible in 2 minutes. Clots appear after 52 minutes. Sol diluted 16 times. Perceptible changv after 1 minute.Coagulation afte 12 minub. Change perceptible after 2 minutes. Clots a p p e a r throughout after 26 minutes. Change perceptible a f t e r a b o u t 3 minutes. Clots appear a f t e r 1 hour 1 minute. (b) Mercuric and Cupric Sulphidc Sols. In the case of mercuric and cupric sulphide sols it is found that dilution increases the stability of the sol irrespective of the nature of the electrolyte. The effect cannot reasonably escape observation. All these apparently anomalous facts can be explained on the assumption that distance between t3he particles of a sol is an important factor in determining its stability. On the adsorption theory of Freundlich coagulation is due to the neutralisation of the charge of the particles of the sol by adsorbed cations.Other things being equal i t follows that an increase in the total surface of the colloid would mean a decrease in the surface concentration of the cation so that a higher concentration of the electrolyte would now be necessary to neutralise the charge on the particles. The amount adsorbed is necessarily small and its effect can only be perceptible when (a) the difference in surface is great compared with the total quantity of the electrolyte present that is ( b ) when the electrolyte concentration is sufficiently low. It is evident that for the concentrations employed in the case of salts of univalent T* 466 MUKHERJBX AND SBN COAGULATION OF cations the difference in stability predicted by the adsorption theory on dilution would be negligible.The adsorption theory thus pre-dicts that dilution of the sol will always diminish its stability and in the limiting case of salts of univalent cations this theoretical diminution may not be perceptible. It has been assumed in the above discusbion that the individual particles in the sol do not change in any way on dilution and hence the total surf ace of the colloidal particles will decrease proportion-ally to the dilution of the sol. The observed increase in stability cannot thus be explained by the adsorption theory as it stands. However if it is considered that dilution also increases the distance between the particles of a sol it explains easily the increase in stability observed. It may be stated here that Freundlich's adsorption theory does not con-template any effect of the distance between colloidal particles on the stability of the sol.The increased distance samehow decreases the facility Gr coalescence and thus increases its stability as will be evident from the sequel. Dilution thus brings into play two factors which have opposite effects on thb stability of the sol. The observed increase or decrease in stability is due to the predominating influence of one over the other. The observations given in table I11 are instructive. I n order that the difference in total surface may have a decisive effect on the stability the quantities withdrawn by adsorption shoald be comparable with the total quantity of electrolyte present that is, appreciable differences in the concentrations of the electrolyte in the bulk of the liquid should result with the different dilutions of the sol employed.As the dilution of the electrolyte increases the differences in h t a I surface become more dominant in determining the stability of the sol. On the other hand if the electrolyte concentration is kept constant then as the dilution of the sol increases the total surface of t'he colloid decreases rapidly and the effect of adsorption becomes counterbalanced by that. of the increase in distance. This is apparent from table IV. TABLE IV. Sol contained 19-45 millimoles of arsenious sulphide per litre. Electrolyte aluminium sulphate. Dilution N / 4000. Sol diluted Dihted Dilubd Original Sol. 4 times. 16 times. 20 times. Coagulation in Coagulates im- Coagulation in Coagulation in 7 minutes.mediately on 40 seconds. 50 seconds. mixing METAL SUL9WD!E HYDROSOLS. PART I. 407 Further dilutions could not be examined as it became increaa-ingly difficult to follow the changes in the sol. With mercuric and cupric sulphide 9019 much higher concestra-tisns of electrolytes are required for coagulation and it is interest-ing to note that in the case of these sols dilution always increases their st'ability. This may be ascribed to the fact that these sols are generally poorer in cdloid content and that the adsorption is much smaller in comparison with the arsenious sulphide sols used. It should be remarked that the total surface varies directly with the dilution whereas the mean distance between the particles varies with the cube root of the dilution.C o m p m t i v e Stability of Sols having the same Colloid Content but dijjfem'ng in the Degree of Dispersion. In the foregoing it has been assumed that on dilution the individual particles da not suffer any change. The observations of Coward (Trans. Faraday Soc. 1913 9 142)show that dilution does not bring about a proportionate decrease in the number of sub-microns and that the migration velocity of the particles in an electric field changes on dilution. Young and Neal (loc. cit.) have also observed an increase in migrabion velocity on dilution with cupric sulphide sols. I n view of these observations a comparison of a pair of sols which have the same colloid content but differ in the degree of dispersion was thought desirable. This is possible with arsenious sulphide sols.Such a comparison has the advantage that in reality two sols are compared one of which has a smaller number of larger particles whilst the 0 t h ~ has a larger number of smaller particles for the same volume. A simple calculation will glhow that the mean distance between the particles and the total surface of the colloid in a given volume differ in the same ratio. The rahio is given by ~ ( T z ) y(n2) where " njl " and " n2 " denote the number of particles present in each case. The relative effects of these factors can thus be compared directly. The finer sol will evidently contain a greater number of partides )khan the coarser one The resulh leave no doubt as to the greater stabilising effect of increased distance. Of course here also with dilute electrolyte solutions and sols differing greatly in the degree of dispersion the surf ace effect is perceptible.By varying the conditions of experiment a seriet3 of sols having the same sulphide content was prepared. For comparison bhe coarsest and the finest sols are selected 468 MIJKHERlEE AND SEN COAQULATION OF TABLE V. Electrolyte strontium chloride. Both sols contained 8.52 millimoles of arsenious sulphide per li tre. Dilution. Sol I (fine sol). N/200 ......... Coagulation after a few N/300 ......... Coagulation after 2 N/400 ......... Coagulation after seconds. minutea, 13 minutes. Sol I1 (coarse sol). Coagulation after a minute. Coagulation after 4 minutes. Turbidity perceptible after 1 minute. Aportion of the colloid had separated after 40 minutes.NlljOO] ......... The greater portion had separated after 40 minutes. It appears that the magnitude of the difference in stability is roughly the same for fhe different electrolytes. It will be seen from the sequel that dissolved hydrogen sulphide has an anomalous effect on the rate of coagulation of arsenious sulphide sol in the case of certain salts. Here also the greater stability of the coarser sol is quite marked. I n table VI are given the respective con-centrations of an electrolyte which corresponds with about the same coagulation time for these two sols. TABLE vr. Comparable concentrations. - Electrolyte. Sol I. Sol 11. SrC1 ............ N/600 N/400 ............ N/400 N/300 KCI ............ N/20 21’116 LiEi ............y;;; NH,Cl ......... $\& Remark. In presence of H,S. Influence of Dissolved Hydrogen Sulphide an the Stability of Metal Sulphide Sols. (a) Arsenious Sulpkide. I n the paper referred to it was stated that dissolved hydrogen sulphide stabilises arsenious sulphide sols against coagulation by electrolytes. The electrolytes used a t that time were salts of uni-valent cations. It has subsequently been observed that arsenious sulphide sols behave in an anomalous manner. When solutions o METAL SULPHIDE HYDROSOLS. PART I. 469 barium and strontium chlorides magnesium sulphate and thorium nitrate are used the sol containing hydrogen sulphide becomes less stable. Tables VII and V I I I show that the diminution in stability is as marked as the increase in stability observed with the other electrolytes.In each set of experiments the sol and the electrolyte were both saturated with hydrogen sulphide freed from impurities. Five C.C. of each were withdrawn by means of a pipette with the help of a rubber hand pump and kept in separate test,-tubes. The liquids were then mixed as usual and kept well corked with india-rubber stoppers. Care should be taken %hat the liquids do not touch the rubber. TABLE VII. Arsenious sulphide sol containing 17.58 millimoles per litre. Electrolyte aluminium sulphate. Dilution. Has absent. H,S present. N/24,000 ... Complete coagulation Partial coaguIation after N/30,000 ... Partial coagulation after Only slight turbidity after after 2 minutes. 34 minutes.6 minutes. Complete 21 minutee. after 11 minutes. TABLE VIII. Comparable Concentrations. Sol I. Sol 11. Sol 111. - - H,S H2S H2S H,S H,S H,S Electrolyte. absent. present. absent. present. absent. present. KCl ............ NIL8 - - N/16 NH,C?l ......... NJ20 $2 N/20 N/12 N/20 BaC1 - ............ N/800 N/1000 N/800 N/1000 - SrC1 - ............ N/300 N/400 NJ300 N/400 - Th(NO,) ...... N/10,000 N/12,000 - - - -The data with thorium nitrate refer to the sol mentioned in Sols I and I1 are the same as those in tables 'v and Sol I11 is a fine sol containing 34.8 millimoles of arsenious The resulk show that the magnitude of the stabilising effect A quantitative table VII. VI. sulphide per litre. varies somewhat with the quality of the sol used. comparison is beyond the scope of the present paper.(b) Mercuric Sulphide. Hydrogen sulphide has a similar influence on mercuric sulphide sols. Increase in stability was observed for ammonium an 470 MUKHERJEE AND SEN COAGULATION OF potassium chlorides and a diminution for barium and strontium chlorides. The sols were prepared as usual from the freshly precipitated hydroxide or sulphide after they had been washed free from electro-lytes. The coagulation of these sols differs in one respect from that of arsenious sulphide sols possibly due to the fact that they are comparatively poorer in sulphide content. Sols unless very rich in sulphide show a minimum coagulation time (as defined here); for example a mercuric sulphide so1 gave a clearance time of about two minutes from Nj20- to Nl3OO-banum chloride.With more dilute solutions the coagulation time increased rapidly as usual. The sols had a blackish-grey appearance. (c) Cupric Sulphide. It is well known that dissolved hydrogen sulphide markedly stabilises the pure sulphide sols both in aqueous and non-aqueous media (Lottermoser J. p r . Chem. 1907 [ii] 75 293) and facili-tates their solution. It is natural to conclude that the same pro-tective effect would be observed with the sulphides of different metals in the presence of electrolytes. This is however not the case here. It has been found that hydrogen sulphide diminishes the stability of cupric sulphide sols. This holds good for all the electrolytes studied namely potassium ammonium strontium, and barium chlorides and aluminium sulphate and the anomaly observed with arsenious and mercuric sulphide sols is absent.Young and Neal (Zoc. cit.) could not find any perceptible effect of hydrogen sulphide on the stability. This is probably due to the method they used. The observed diminution in stability can be understood from an observation of Young and Neal. They find that hydrogen sulphide diminishes the velocity of migration of the particles of a cupric sulphide sol in an electric field. It follows from the well-known Helmholtz-Lamb formula (Rep. Brit. Asso'c. 1887 495) that a diminution of the electric charge of the particles takes place pro-vided that other factors remain constant. The result will be a diminution in stability as a smaller amount of adsorbed cations will now be required to discharge the particles.As it is not clear that simultaneous measurements of viscosity and other properties were made the parallelism loses much of its significance. Solutions of ammonium potassium barium and strontium chlorides and aluminium sulphate were studied. With ammonium and potassium chlorides nearly satmurated solutions have to be used METAT SUIiPHIDE HYDaOSOLS. PART J. 471 TAEXLE IX. Clearance time. .A J 7 Electrolyte. Dilution. H,S absent. H,S present. KCl .................. - 6 minutes. A few seconds. BaCl .................. N/20 9 9 ) 4 minutes. Al,(SO,) N/2000 13 9 9 9 9 9 ............ Protective Action of Alkali Sulphides and Alkali Hydroxides. Solutions of potassium and sodium sulphides have a more marked protective action.This stabilising influence has been found for ammonium potassium barium and strontium chlorides and aluminium sulphate on hydrosols of cupric mercuric and arsenious sulphides. A trace greatly facilitates the preparation of sols rich in sulphide and largely increases their stability. The protective action of alkali sulphides is probably due to the free alkali hydroxide present as a result of hydrolysis. Since these substances dissolve arsenious sulphide with the form-ation of arsenites and thioarsenites i t is not possible in this case to refer the protective action observed to the hydroxidions alone. Indeed the liquid obtained by dissolving in a few C.C. of dilute alkali hydroxide as much arsenious sulphide as possible has an equally marked protective action on the sols of t.hese three sulphides.However as alkali hydroxide does not react with the other sulphides and produces similar protective action i t is prob-ably that' the effect of all these substances is due to the trace of free alkali hydroxide present as the result of hydrolysis. Freundlich (Zeitsch. physikul. Chem. 1903 44 144) has observed t'hat salts of alkali metals with organic anions of large mass have a lower coagulating power than the corresponding salts of inorganic acids. He refers this to the protective action of these anions due to greater adsorbability but in view of the preceding this can as well be due to the trace of alkali hydroxide present on hydrolysis. I n conclusion it may be stated that' the anomalous influence of hydrogen sulphide is not without? parallel. Recently Freundlich has found somewhat similar behaviour with ferric hydroxide hydrosol (Biochem. Zeitsch. 1917 81 87). An actual reversion of stability was notl however observed in this case. So far as can be understood from the abstract of the paper he explains these irregularities as due to selective adsorption (loc. cd.). It remains The behaviour of the sols is thus very regular. This is also the case with alkali hydroxides 472 ASTON A SIMPLE FORM OF APPARATUS FOR to be seen how far these observations can be explained on the basis of tjhe existing theories. Our best thanks are due t o Sir P. C. RBy and to Dr. J. C. UNIVERSITY COLLEUE OF SCIENOE, Ghosh. CALCUTTA. [Received Fehruary 41h 1919.