KARTUNG STUDIIS WITH THE MICBOBAL4XCE. PART II. 2691 CCCLXVIII.-&udies with the Microbalance. Part I I . The Photochemical Dewmposition of Silver Chloride. By ERNEST JOHANNES HARTUNG. THE investigation of the action of light on silver chloride by gravi-metric methods has not hitherto given very definite r d b . It is well known thak the halide loses chlorine and diminishes in weight when insolated but the changes are usually insignificant and afford little evidence as to the nature of the products of the action. Baker (J. 1892,61 728) tried to analyse the dark photochloride produced when 51 g. of silver chloride were insolated with repeated shaking, but only 14 mg. of chlorine were evolved after considerable exposure to light. Richardson (J. 1891 59 536) found 8% of the total halogen to be freed when 26 g.of silver chloride under water were insolated but Koch and Schrader (8. Pliysik 1921 6 127) using extremely small amounts of the dry substance could not detect greater changes in weight than 1 or 2%. A preliminary test carried out by the author with the microbalance (J. 1922,121,688) showed that a very thin film of the halide weighing 0-0880 mg. lost 81% of its total chlorine when insolated for 12 days in a vacuum. This line of attack wm therefore promising and the d h of an extended investigation tare now given. The Steele-Grant microbalanc 2692 H A R ~ G STUDIES w - r r ~ THE ~CROBALBNCE. PART n. employed carried a load of 105 mg. and was sensitive to 2 x mg. ; it was used in ideal conditions in a roomy cellar and had been thoroughly tested for 2 years.The weighings with this instrument a m recorded to the nearest 104 mg. and they are thoroughly trust-WO*~ to this degree of precision for small discrepancies in every C- where they were noted could be traced to the inevitable errors of manipulation and were never due to inaccuracy of the balance. The general plan of the work was similar to that used for the case of silver bromide (J. 1924,125,2198) ; very thin iilms of the chloride on vitreous silica sheets were sealed up in glass vessels containing suitable chlorine absorbents and insolated for defhite periods. The films were then weighed again and finally tested in various ways. Some experiments have also been made on the rate of chlbrination of thin films of silver and of the product of insolation of silver chloride.Preparation of F i l m of Pure Silver Chloride.-Films of pure silver on vitreous silica were made by chemical deposition and subsequent ignition at 4-00' as already described (h. cit.). These films were then chlorinated by exposure to dilute chlorine which was made in the usual way from manganese dioxide and hydrochloric acid, followed by washing with water and d r p g with concentrated sulphuric acid. The rate of chlorination was comparatively slow unless the gas was dilute and the weight of the resulting silver chloride usually agreed closely with that calculated from the weight of silver taken. The maximum divergence was 1 part in 24-0 parts and the average was 1 part in 1400 parts. These results are dis-tinctly leas consistent than those obtained previously with silver bromide where the maximum divergence was 1 part in 10oO parts and the average 1 part in 2400 parts.Indeed the whole process of chlorination of silver exhibited peculiarities which have not been observed in bromination. For example a certain optimum concen-tration of chlorine in mixtures with air was found at which addition of the halogen to silver occurred most easily and the speed of chlorina-tion rapidly diminished as the chlorine concentration became greater than this optimum. Also chlorine which was made from bleaching powder and dilute sulphuric acid gave chlorinated films in which the added halogen was from 3 to 4% in excess of that required to form silver chloride. The excess could be removed by cautious heating at 200" and could be prevented by igniting the chlorine before use, This pointed to some oxygen compound it8 the cause which supposi-tion was strengthened by hding that mixtures of chlorine and moist air which had stood for some days had a tendency to give slightly high results in chlorinating silver and that intermittent chlorination of silver films gave high results also.These effects were not notice m PB[OTOCHEMICAL DECOMPOSITION OF s.nm CHLORIDE. 2693 in the bromination of silver ; in any case the discrepancies are small, but their reality was established by repeated tests. Photocliemical Decomposition of Silver CMde.-The glass apparatus in which the films were sealed up and insolated was essentially similar to that used for silver bromide and has already been described.The technique adopted was similar also but great care was required to prevent contamination of the film by flame gases during the sealing for silver chloride seemed to be more sensitive in this respect than the bromide. With proper precautions, however the operation could be performed without appreciable change in weight of the h. Before exhaustion the glass vessel was filled with air nitrogen or hydrogen which had been carefully purified dried and flbred before admkaion. The final pressure in the apparatus was usually either 10 mm. or 0.001 mm. but as heating of the glass during exhaustion was inadrmss ible the latter pressure especially was not maintained over the whole period of insolation owing to the liberation of admrbed gas from the glass.For this reason the term " residual " is used in Table I to describe the conditions when the pressure of the experimental gas in the appamtus was 0.001 m. at the time of sealing off from the pump. Copper was used as chlorine absorbent but if insolation were per-formed with hydrogen present in other than residual amount solid sodium hydroxide was substituted. Insolation was performed on the roof of the laboratory and as the colour opacity and reflecting power of the h s varied steadily no estimate of the amount of radiant energy absorbed was possible. Nevertheless the times of exposure in days afford an approximate measure of the relative amounts of energy received in each case. The pearly white films always darkened rapidly to dull purple or slate which passed slowly to purple-brown; after several days this had faded to very pale greyish-yellow and no further change was visible.Meanwhile the copper in the side tube which was shielded by an opaque cover, became heavily tarnished at the nearer end and subsequent analysis showed always the presence of chlorine on it. After a definite period of exposure filtered air nitrogen or hydrogen was slowly admitted, and the film removed and weighed. The subsequent treatment varied; some of the exposed i3.m~ were rechlorinaw either in one or in several progressive operations whilsf others were used in attempts to discover the nature of the products of the phofochemical decomposition. A summary of the experimental results is shown in Table I, weights being given in milligrams.Comparison of the results in the presence of air nitrogen and hydrogen discloses a striking resemblance between them and it i 2694 €L4RTUNG STUDIES WITH THE MICROBALANCE. PART II. TABLE I. 0.2543 0.4180 0.2354 0.2428 0-3206 0-2775 0.4117 0- 1750 0.1360 0.2113 0-2634 0.1816 0-2204 0.2 108 0.2305 0.1641 0.3365 0.5556 0.3126 0.3227 0.4247 0.3690 0-5467 0.2326 0.1812 0-2809 0.3496 0.2417 0.2937 0.2808 0-3068 0.2181 0.3379 0.5554 0.3128 0.3226 0.4261 0.3688 0.547 1 0.2325 0.1807 0-2808 0-3500 0.2413 0.2929 0.2801 0.3063 0.2181 0.2899 0.4468 0.2472 0.2449 0-3432 0.2862 0-4632 0-1863 0.1406 0.2219 0.2828 0.1852 0.2375 0.2172 0.2466 0.1669 0.3362 0.3146* 0-3250* 0-3706* 0-5486 0.1829 ---- - -0-2820* 0-2 185 -56.0 2 79.1 9 84.7 11 91-1 27 78.3 10 90.5 87 61.9 87 80.4 15 89.9 88 77.5 10 94.2 62 76.7 10 90.9 91 78-9 10 94.8 111 84.4 12 * Chlorination in steps.evident that oxygen is not necessary for the photochemical decom-position of silver chloride. The experiments with air in the apparatus exhibit plainly the diminution in the speed of decomposition as the gas pressure rises due to the adverse effect of adsorbed gas h s . For nitrogen the data are insufEicient to warrant conclusions being drawn and for hydrogen the gas pressure appears to exert little influence on the change. This behaviour is not surprising for hydrogen acts as a sensitiser in virtue of its power of combining with chlorine under the influence of light.As in the case of the bromide the figures indicate that the products of the photochemical decomposition of silver chloride are silver and halogen although the final few units yo of the latter are held tenaciously in the h. There is no evidence of the formation of any sub-chloride. On rechlorinating the product of insolation colour changes were observed similar to those seen on insolation but in the reverse order. Eventually the films regained their original pearly white-ness and were then usually rather heavier than in the former un-exposed condition. This behaviour was always most marked when the chlorination had taken place in steps with interruptions for the purpose of weighing and it is believed to be due to slight oxidation during the addition of the halogen but not at other times.It was also observed to some extent when pure silver was chlorinated pro TKE PHOTOCHEMICAL DECO-MPOSITION OF SILVEB CIELOBXDE. 2695 gressively. Hence it is probable that the figures given in Table I for the percentage of chlorine lost on insolation are not appreciably affected by oxidation due to expasure to air before weighing. Chemical analysis of the insolated material would be decisive but, with less than 0-5 mg. available in each case no accurate procedure has yet been devised. Two simple methods were tried and neither proved to be quite satisfactory. In the first method the film was treated with aqueous sodium thiosulphate to dissolve any unchanged silver chloride but in spite of the greatest care insoluble particles became detached when the solution touched the film and a quanti-tative estimation was impossible.Much of the residue remained adherent to the silica however and chlorination testa proved it to be practically pure silver. In the second method the insolated film was exposed to the vapour of purified iodine at room temperature (partial pressure 0.3 mm.) in the hope that the free silver presumably present would be attacked by the iodine to form silver iodide the amount of which could be measured by the increase in weight. In principle this is obviously a more rigorous means of testing the nature of the insolated film than by chlorinating it. Unfortunately, however silver chloride itself was shown to be attacked by iodine vapour although the complete conversion into silver iodide of even very thin films of the chloride took several days.When insolated silver chloride was exposed to the iodine vapour the film rapidly darkened and then steadily changed t o the pale yellow colour of silver iodide in about 15 minutes. The first stage of the action was now taken to be complete and the film was weighed. Table I1 summarises the results weights being given in milligrams. TABLE 11. Wt. of original Wt. of insolated Wt. dtar Wt. calculated for film. an. iodination. mixed AgI and AgCI. 0-2326 0.1863 0.3646 0.3520 0.2417 0.1852 0-3898 0.3874 0.2809 0.2219 0.4338 0.4331 0-2937 0.2375 0.4489 0.4386 The weight of the iodinated film in each case is grater than that calculated on the assumption that the insolated material is a mixture of silver and unchanged silver chloride of which only the former is attacked by the iodine.In two instances the discrepancy is less than 1% and on the whole the results are favourable to the hypothesis mentioned for further iodination caused steady increase in weight in each case until after some days all the chlorine had been displaced by iodine. The method is therefore not thoroughly sound for the initial stages in the displacement of the chlorine proceed appreciably in a very few minutes and may of course b 2696 ~ " a STUDIES WITH THE MICBOBAT,ANCE. PKBT II. more rapid with previously insolated films than with pure silver chloride. Rate of CWruation of Silver Films.-Further evidence as to the nature of insolated silver chloride was obtained by comparing the rates of addition of chlorine to pure silver and to the insolated films under the same conditions.The apparatus and methods of ex-periment were essentially analogous to those used in the bromination of silver (J.y 1924,125,22O4) and have already been described. The tests were carried out in a thermostat at 25" (regulated to 0.1") with mixtures of chlorine and air in various proportions in the chlorinat-ing vessel. The concentration of the halogen was estimated by absorption in sodium hydroxide and titration in the usual way with FIG. 1. T i m of expornre in minutes. sodium thiosulphate. In every case the total area of the film was 364 sq. mm. and the average thickness 040015 mm.These results are shown graphically in Fig. 1 where the chlorine absorbed by the filmy as percentage of that necessary to form pure silver chloride in each case is plotted against total time of chlorjna,-tion in minutes. The broken lines indicate rechlorination of previ-ously insolated silver chloride films and the numbers denote chlorine concentrations in the gas phaae in mg.-atoms per litre. The curves are continuous and show no irregularities which might point to the transient formation of sub-chlorides in the film. Also the remarkable fact is disclosed that the speed of chlorination of pure silver rapidly dmumshes as the concentration of the halogen in the surrounding medium increases. This effect was not noticed in the bromination .. of silver (Zuc. cit.) but much smaller halogen concentrations were then employed owing to the very rapid attack of the silver fihw by bromine vapour. It is evident that an optimum concentrcttion must exisf at which chlorine attacks silver most readily. The precise value of this concentration has not been measured and it is hoped to investigate the whole action thoroughly in the near future. For the present purpose it was sdicient to show that isolated dver chloride h s exhibited the same behaviour a,s silver itself when exposed to chlorine. That this is so will be evident from an in-spection of the broken curves in Fig. 1 ; the existence of the optimum in this case also has beenestablished byrepeated tests and furnishes strong evidence that silver is present in the insolated material.It cannot be there in the same form aa ordinary metallic silver, however for the optimum chlorine concentration for the product of insolation is much greater than that for the metal itself whilst the whole rate of chlorination is comparatively very much slower for small concentrations of the halogen. In this respect chlorine stands in marked contrast to bromine and iodine both of which even in small concentration attack the product of insolation of silver chloride very rapidly. The work described in this paper does not contradict the conclusions of Baker (J. 1892 61 728) that perfectly dry silver chloride is unaffected by light and that perfectly dry chlorine will not attack silver. Owing to the danger of injuring the photosensitive films, it waa not possible to bake out the glass exposing vessels and the presence of phosphorus pentoxide has been found to be most objec-tionable because of the fine dust which arises from it with changes of pressure.Also in the chlorinafion experiments the halogen was '' dried " with concentrated sulphuric acid only and little alteration in the results was noted when moist chlorine was substituted for it. Summary. 1. The photochemical decomposition of silver chloride in air, nitrogen and hydrogen has been investigated by means of the microbalance. 2. The maximum percentage loss of the total chlorine in thin films of silver chloride when insolated was found to be in air 91-1% in nitrogen 89-9% and in hydrogen 94.8%. 3. Evidence is adduced to prove that the photochemical decom-position products of silver chloride are silver and chlorine and that oxygen is not necessa,ry for the action. 4. The rates of chlorination of silver and of previously insolated silver chloride have been studied and it has been shown that optimum concenbations of the halogen exist in each cam at which th 2698 MORTON AND BOQERS: chlorination is most rapid and above which the speed of reaction rapidly dimini8hes with increasing chlorine concentration. 5. No evidence of the formation of silver sub-chlorides has been found. UNlvEaSrry OF &LEOwNE. [Received June W h 1925.