THE R6LE O F PROTECTIVE COLLOIDS IN CATALYSIS. 109 XHL-I’he R81e of Protective Colloids in Catalysis. Part I. By THOMAS IREDALE. TEE effect of protective colloids in inhibiting the catalytic decom-position of hydrogen peroxide by colloidal platinum has already been noted by Gr6h (Zeitsch. physikal. Chem. 1914 88 414) who found that they appeared to follow the order of Zsigmondy’s “gold number ” series (Zeitsch. anal. Chem. 1901 40 697; ‘( Colloids and the Ultramicroscope,” 1914 79) in their activity in this respect. It appeared desirable however to extend investigations in this direction as the experimental results obtained by Gr6h were too few to permit of a general statement as to the relative effects of different protective colloids. A number of these colloids have now been examined in this regard and it has been found that in general the stronger a substance is as a protective colloid the greater will be its inhibition of catalytic activity and that a sub-stance like sucrose which is without protective effect is likewise without inhibitive effect.Now in view of the circumstances in which the protective colloids are examined in the two cases-Zsigmondy’s coagulation method and Gr6h’s cat,alytic method-there is nothing altogether surprising in these results. From superficial considerations they might almost be anticipated and the author was inclined t o believe that they might throw considerable light not only on the mechanism of pro-tective action but also on the processes involved in the hydrogen peroxide decomposition Unfortunately owing to war conditions, Gr6h’s paper itself was not available and the author had to be content with abstracts (A.1915 ii 239; Chem. Abstracts 1915 7). From these it does not appear that Gr6h advanced any thorough argument to account for his results but be seems to have en-deavoured to obtain a relation between the gold number of protec-tive colloids and the extent of their inhibitive effect the significance of which seems a t present a little obscure. It is necessary to examine the analogy more thoroughly. In this coiinexion mme recent remarks made by workers in this field are of interest. Bailcroft ( J . Physical Chem. 1917 21 775) considers that a substance like gelatin may increase the degree of dispersity of the catalyst thus exposing a larger surface with increased cata-lytic activity but that this effect may be more than counter-balanced by the presence of the gelatin itself which hinders th 110 IREDALE THE R ~ L E OF adsorption of the hydrogcn j)eroxicie.There is no evidence that protective colloids increase the degree of dispersity of a metal sol already formed and in view of Rusznyak’s work (Zeitsch. physikal. C‘hem. 1913 85 681) on the decreased catalytic activity with imreased dispersity Bancrcft’s argument seems scarcely reason-able. Rideal ( J . Amer. Chem. Soc. 1930 42 749) considers that diffu-sicn is the chief factor concerned in the rate of decomposition of hydrogen peroxide and argues against the idea of a colloidcom-plex formation. If this is the case why should a strong protective colloid inhibit.to a greater extent than a weak one? What part, can diffusion play in the ordinary method of measuring the value of protective colloids as announced by Zsigmondy? The change from red t o blue in the colour of gold sols is assumed to be due to a union of the gold particles after their charges have been neutralised by the adsorption of certain ions. Protective colloids may hinder this change for one or other of two reasons. It may be that after the neutralisatim of their charges the gold particles are prevented from uniting owing to the presence of the protective colloid. On this theory it is difficult to see where the analogy exists in the case of the catalytic process. I f we assume however that the protective cdloid hinders the adsorptioii of the ions that would bring about the coagulation then the analogy is quite complete.The rate of decomposition of hydrogen peroxide is probably determined by a number of factors of which adsorption is undoubtedly one of the chief. Anything which hinders the adsorption of the hydrogen per-oxide by the catalyst will retard t,he velocity of reaction as measured in the usual way and a strong protective colloid which hinders the adsorption of ions better than a weak one may also hinder the adsorption of the hydrogen peroxide more efficiently. It remains to be seen however if the gold number is really expressive of these relations. Zsigmondy (“ Colloids and the Ultramicroscope,” 1914 150) has endeavoured to follow the mechanism of protective action under the ultramicroscope.The assumed union of the gold and gelatin ultra-microns is followed by decreased mobility in the former except a t a certain concentration of gelatin below which there does not seem to be any retardation in the movements of the gold particles. It will be seen later that the inhibitive effect of gelatin is noticeable a t much lower concentrations than the critical one mentioned by Zsigmondy and this inhibition cannot be due therefore to any decreased motility in the particles of the catalyst. As far as the Brownian movement is concerned however the part played by it4 in the catalysis is still somewhat obscure PROTECTIVE COLLOIDS M CATALYSIS. PART I. 111 Bredig (Zeitsch. physikab. Clhem. 1901 37 14) has shown that the adsorption of poisons by the catalyst follows the logarithmic law and it was anticipated that the adsorption of protective col-loids might also obey the same law.From results obtained with gelatin a t very low concentrations it appears that the process is more complicated than a simple calculation can possibly account for owing t o the continual subdivision of the gelatin ultramicrons over a certain range of dilution. T’he results of experiments on the poisoning of protected metals ~ i l i be made available in a later communication. E x P E R I M E N T A L . The hydrogen peroxide used in all these experiments was care-fully purified by distillation under diminished pressure. The col-loidal platinum solutions were prepared by Bredig’s method using a current of 110 volts and 10-12 amperes the temperature of the water being kept below 25O.Solutions made by this method may bc diluted to the extent desired and after allowing the larger par-ticles to settle may be used directly without filtering. They appear, however to be much more sensitive than filtered ones and cannot be used for very exact work where it is desired to follow the course of a reaction with the maximum of accuracy. The velocity constant falls slightly during the reaction instead of rising as is usually the case. The solutions of protective colloids were prepared by simple dis-solution of the materials in water adopting the usual procedure for gelatin and starch. I n the case of gum tragacanth and egg-albu-min which gives extremely turbid solutions a known weight of material was dissolved as much as possible in water and the amount of undissolved matter ascertained after filtration.With a know-ledge of the weight of substance in the filtrate it could then be diluted to the concentration required. T’he concentration of protec-ive colloids when first prepared was 0.04 per cent. and lower concentrations were obtained merely by dilution from this strength. The initial concentration of the hydrogen peroxide in all the experiments was 21.1 / 40. The concentration of the platinum solutions was the same throughout any one series. All the reactions were carried out a t 25O and in every instance the plat,inum solutions on admixture with the protective colloids were allowed to remain for fifteen minutes a t the temperature of the experiment before the addition of the hydrogen peroxide.At different intervals 10 C.C. of the reaction mixture were titrated, after addition to dilute sulphuric acid with standard permanganat 112 IREDALE TI-IE R ~ L E OF (about J / 4 0 ) . It was not i'ound necessary t o apply a correctic to the titrations for organic matter present as the concentration ( thc latter was apparently too small to affect the results. The velocity constant was calculated from the usual formuIa : k = 0.4343 k = - 1 loglo- a t a-x ( t in minutes and a-x in terms of C.C. of potassium permar ganate). The values of k given in the tables are the progressive one obtained durizig any reaction and the mean of these in each cas gives the same result on comparison as the time for 50 per cenl decomposition.The ratio value3 were calculated by taking the velocity constan wit,h unprotected metal as unity. TABLE I. Protective Colloid Preparations Six Days Old. Series. Protective colloid. k. Mean. Rat'io. - I. none ... .. ............... 0.026 0.023 0.023 0-024 1 0.01 yo gelatin ... .. .... 0.0043 0.0043 0.0044 0.0043 0.17 , glue ............ 0.0046 0.0043 0.0041 0.0043 0.17 , egg-albumin . 0,0055 0.0052 0-0050 0-0052 0.23 , gum arabic ... 0.014 0.012 0.012 0.013 0.54 , sucrose ...... 0-027 0.025 0.025 0-025 1 11. none ..................... 0.039 0.038 0.037 0.038 1 0.001% gelatin ... ... 0-0074 0.0082 0.0078 0-0078 0-20 , egg-albumin 0.013 0.012 0.012 0.012 0.32 , gum arabic. 0.032 0.031 0.031 0.031 0.82 , sucrose . ..... 0.038 0-036 0.036 0.037 1 , glue ......... 0.0080 0.0078 0.0077 0.0078 0-20 TABLE 11. Preparations One Day Old. Series. Protective colloid. k. Mean. Ratio. - I. none ..................... 0.057 0.052 0.056 0.055 1 O.Olyo gelatin ......... 0-0058 0-0061 0.0060 0-0059 0 . 2 2 , glue ............ 0-0071 0.0073 0-0071 0.0072 0.13 , egg-albumin. 0.0093 0-0095 0.0094 0-0094 0-17 , gum arabic ... 0-037 0.035 0.034 0.035 0.64 11. none .......... .. ....._. ... 0.025 0.024 0-025 0.025 1 0.001% gelatin ... ... 0.0044 0.0043 0.0045 0.0044 0.18 , glue ............ 0-0055 0.0059 0-0053 0.0056 0-22 , egg-albumin . 0-0068 0.0072 0.0071 0.0070 0.28 , gum arsbic ... 0.020 0.021 0.020 0.020 0.8 PROTECTIVE COLLOIDS IN CATALYSIS. PART I. 113 From these results it is evident that the inhibitive effect is in the order gelatin and glue>egg-a1bumin)gum arabio>sucrose, which does not appear to affect the reaction a t all.Tbis order is also followed in Zsigmondy’s coagulation experi-ments but the author has n o t been able to discover any exact rela-tionship between the gold numbers of these colloids and their inhibitive activity as indicated in the ratio table. Subsequent determinations of the gold number by the usual method (Zsigmondy loc. c i t . ) gave values of 0.006 and 0.008 for the samples of gelatin used and 0.2 for gum arabic so that the author was not working with materials showing any great anomalies in this respect. The gold numbers seem therefore to be only a useful guide to enable one to predict the probable order of inhibitive activity.The protective colloids are themselves without appreciable action on hydrogen peroxide. Eredig (Zeitsch. physilcal. Chem. 1899 31, 342) showed this in the case of gelatin and it has been found that the stability of a hydrogen peroxide solution is not appreciably affected by tha addition of protective colloids of the concentrations indicated irr any of these tables. The extent of the inhibition produced by some of the weaker protective colloids is shown in the following table: TABLE 111. Preparations Two Days Old, Series. Protective colloid. 1. none ......................... 0.01 % gum tragacanth , dextrin ............ , starch ............. , gun arabic ...... 11. none .......................... O * O l ~ o egg-albumin ...... , tragacanth ........111. none ........................ 0.001 % egg-albumin ... , tragacanth ...... IV. none ........................ 0.01 yo sodium oleate ... o*Oo5~0 , Y ? ... o.Ooly0 , 9 7 * * a 0.0025% , ? ? k. 7-0.080 0.083 0-086 0.030 0.029 0.027 0.032 0.031 0.030 0.042 0.042 0.040 0.044 0-044 0.040 0.017 0.018 0-018 0.0036 0.0038 0.0039 0.0078 0.0074 0.0073 0.017 0.018 0.019 0.0057 0.0058 0.0056 0.012 0-013 0-013 0.017 0.018 0.019 0.021 0.023 0.023 0-0130 0.0130 0.0132 0.0127 0.0129 0-0127 0.0150 0.0152 0.0151 Mean. Puatio. 0.083 1 0.025 0.34 0.031 0.37 0.041 0.50 0-043 0.52 0.018 1 0-0038 0.22 0.0075 0.42 0.018 1 0.0057 0.32 0.013 0.72 0.018 -0.023 -0.0130 -0.0128 -0.0151 -From these results it is evident that tragacanth inhibits to a less est,ent’ thau egg-albuniin but is somewhat more effective than dextrin which is inore effective than starch and gum arabic 114 THE RbLE OF PROTECTIVE COLLOIDS IN CATALYSIS.Sodium oleat,e behaves abnormally as it must be completely hydrolysed at these low coiicentrations and the velocity constant will rise owing to the presence of hydroxyl ions. It is interesting t o observe however that on dilution from 0.01 per cent. concentra-tion the protective effect of the soap begins to dominate the situa-tion and the velocity constant therefore falls but rises again on further dilution of the protective colloid. This observation is of great importance as it shows that the protective action of soaps on gold sols is not due to the stabilising effect of the hydroxyl ions alone-the concentration of the latter in a 0.01 per cent.sodium oleate solution could not be greater than N / 3000-but more prob-ably in greater part to the acid-soap residue which is more complex, perhaps than is generally realised. Gelatin appears to be active as an inhibitor a t extremely low concentrations. With a platinum solution of medium concentra-tion (about 1 /30,000 gram-atoms per litre) the following results were obtained : TABLE IV. Series. Protective colloid. k. Mean. I. none ..................... 0.005% gelatin ...... 0.001 % , ...... 0*0001 yo , ...... 0.00005y0 , ...... 0~00001 yo , ...... 0*000005~0 , ...... 0 ~ 0 0 0 0 0 1 ~ , ...... 11. none ..................... 0.001% gelatin ......0.000 1 % , ...... 0.00005% , ...... 0~0000 1 % , ...... 0.000005% , ...... 0*000001% , ...... 0:0148 0-0152 0.0151 0.0154 0.0028 0.0026 0.0026 0.0027 0.0032 0.0032 0.0031 0.0029 0.0045 0.0044 0.0041 0.0040 0.0050 0.0049 0-0050 0.0051 0.0105 0.0108 0.0108 0.0108 0.0137 0.0142 0.0141 0.0139 0.0149 0.014-8 0.0152 0.0154 0.0158 0.0161 0.0167 0.0162 0.0030 0.0032 0-0030 0.0029 0.0043 0.0039 0.0038 0-0038 0.0060 0.0047 0.0045 0.0047 0.0108 0.0110 0.0114 0.0117 0.0123 0-0123 0-0129 0.0133 0.0155 0.0156 0.0159 0.0160 0-0151 0.0027 0.003 I 0-0050 0.0107 0.0140 0.0151 0.0162 0.0030 0.0039 0.0048 0.0112 0.0127 0.0158 0.004.3 Series I and I1 were carried out with different samples of g cla tin.) The most striking fact about these results is the gradual rise of the velocity constant with diminishing gelatin concentrations down to 0*00005 per cent.and the rapid rise on further dilution of the protective colloid. Now Menz (Zeitsch. physikat. Ghem. 1909 66 129) found that the protective action of gelatin increased on dilution but the results were usually dependent on the mode of preparation of the original solution. It seems not improbable however that on diluting a gelatin solution of low concentration the larger gelatin ultra-microns split into smaller ones and these being more strongl YEAST CROPS AND THE FACTORS WHICH DETERMINE THEM. 116 adsorbed by the gold or platinum particles will partly make UP for the decreased concentration of the protective colloid.Hence the velocity constant will only rise very slowly until this subdivision process ceases when further dilution of the protective colloid will now bring about its more rapid elevation. Summary. (i) The inhibitive effect of protective colloids on the catalytic decpmposition of hydrogen peroxide by colloidal platinum has been examined in a number of instances. (ii) It has been found that the stronger a substance is as a pro-tective colloid the greater will be its inhibition of catalytic activity. (iii) I n the case of a strong protective colloid like gelatin the inhibitive effect is noticeable a t very great dilution for example, 0.000005 per cent. or one part in twenty million parts of water. (iv) The inhibition is explained on the ground of selective adsorption resulting in a decreased concentration of hydrogen peroxide a t the platinum surface and a consequent fall in the value of the velocity constant. (v) There is no precise relation between the gold numbers of protective colloids and the extent of their inhibition. (vi) The reaction may be used not only to detect adsorption effects but probably also changes in state of the protective colloid owing to the subdivision of its ultramicrons. I am indebted to the Committee on award of Science Research Scholarships in this State for a scholarship which has enabled me t o carry out this investigation. UNIVERBITY OB SYDNEY. CHEMICAL LABORATORY, [Received September 14th 1920.