1280 MAXTED THE INFLUENCE OF HYDROGEN SULPHIDE ON CXXXVII1.- The I n j h e n c e of Hydrogen Sulphide o n the Occli~~ion of Hydq-ogen by Palladium. Part 11. By EDWARD BRADFORD MAXTED. IN a previofus papelr (T. 1919 115 1050) it has been shown that whilst colmparatively small quantities of occluded hydrogen sulphide are sufficient totally t o inhibit the1 subsequent occlusion of hydrogen by palladium this occluded hydrogen sulphidel on being heated to1 looo in a vacuum evolves an equal volumel of hydrogen and passes into a sulphur complelx which is f a r less poisonous than hydrogein sulphids itself. The inhibitive effect of sulphur in the l a t h r form could be prediclted hy assuming the folrmation of a, oolmplelx Pd,S which is incapable of occlusion and ascribing to the re6dual palladium its normal occlusivei poiwer for hydrogen.In the preivious paper the1 minimum volumei of hydrogen sulphidei as such required f o r the1 complete1 inhibition of the1 occlusive poiwer s f palladium f o r hydrogen was not1 determined, it being noltled holwever that the volume of gas absorbed by palladium on exposure' for a short time to hydropn sulphide-this amounting in an experiment to1 13.5 C.C. per gram of metal T€I% OCCLTJSION OF RYDROGICN BY PALLADIUM. PART I J . 1281 was sufficient to1 prevent the ordinary occlusion of hydrogen on subsequently admitting the1 gas. . The1 subjeict has now been investigated further with the object of following quantitatively the inhibitive effect' of successive incre-melnt,s in the! hydrogsn sulphide content of a given welighti of palladium this hydrogen sulphide being present in the form in which itl is olriginally occluded and being added t o the palladium in small known quantities which for the majority of the1 measure-inelnts were less than the volume1 required for saturation.The poisoning curve1 for undecomposeid occluded hydrogen sulphide was FIG. 1. found t o be) linear in form. Evidence was also1 obtained of the gradual and spontaneoas dissociatioln of the1 occluded hydrogen sulphide men a t %he ordinary temperature( with the formation od hydrogen and of the Pd,S coinplex. This reaction is aocompanieid by a colrresponding change in the ocdusive power. E X P E R I M E N TAL. The apparatus elmployeld is shown diagrammatically in Fig. 1. It consists of a small absorptioln tube -4 containing a known weight oif palladium which was weighed out as chloride and redumd in situ as before.I n order t o intlroduce into the system hydroge 1282 MAXTED THE INFLUENCE OF HYDROGEN SULPHIDE ON or hydrogen sulphidei as requireid two gas burettes U a i d C' wcrc provided. These! had a capacit?y of 25 C.C. and 5 C.C. reapectively, the larger burethe being employed for hydrogen. The1 systelni zoluld be exhausteld by means olf a Sprengel pump IT or where more convenient by lowering and raising the mercury in B the three-way cock being turned as required. All parts of the apparatus weIra joined by fusion and all stopcocks weirel mercury-sealeld. I n view of tha inhibitive effect of mercury on the1 occlusive polwer of palladium obseirved by Paal and Hartmann (Be?-.1918, 51 711) a short plug of gold leaf 11 was inserted as a precau-tion against the1 possible! penetration 04 traoes of melrcury vapour into the absorption tube. The hydrogen useld for reducing the palladium chloride in A and for filling B was obtained frolm a cylindelr of the electrolytlic gas and was freed from traces of oxygeln by passagel through a " half-watt" lamp K follolwed by a soda-lime tube G and phosphoricl oxide tube F . The hydrogen sulphide was prelpared by the1 action of helat on magnesium hydroc sulphidel and was dried by passage elver calcium chloride. In order to relduce the volume! of the1 absorbing system t o as small a value as possible capillary connexiolns were1 employed throughoutl. The volume olf this system was redetesmiiied for each a,bsorptisn tubel the1 value being relquired in order tlo obtain the tlrue olaclusive powelr of the preparation froim the vollume of hydrogen absorbed from B .Each measurement was begun by sealing on a fresh absorption tubs containing a known weightl of palhdium ahloride. This was retduced t o metal by means of hydrogen a t looo the1 current of gas passing out of the system by way of the open end of A which was subsequently sealed off. The vodumel of hydrogeln which was capable of beling evolved a t 100' or occluded a t the ordinary temperature having been determinefd as before A was exhausted a t looo and after cololing to the ordinary texnpelrature E wa5 clmeld and a known small volume1 of hydrogen sulphidel allofwed to pass from C into the! absorption tube!.This gas was found to be quickly and practically colmpletely absorbeld by the palladium and, on subselquentlly admitting hydrogeln from B occlusion of this gas tofok place ta an extent delpendent? on the1 degree 04 poisoning which had been induced by the preliminary treatment with hydrogen sulphide. The reeults of the various measurements are summarised in table I in which the elxperimelnts are arranged in order of inareasing hydrogen sulphide content. The vollumes of gas are in every aase reducetd to Oo and 760 mm. the experiments being, however carried outl a t 20' No. of expt. 1 2 3 4 5 6 7 8 9 Weight of palladium. Gram. 0.1033 0.2380 0.0976 0.1714 0*2000 0.2795 0.1890 0.1422 0.0993 Original oc c h i ve power.7.25 16.45 6.86 C.C. H2. 12.1 14.1 19.0 13.1 9.9 6-25 Equivalent occlusive power in C.C. H per gram Pd. 70.0 69.2 70.1 70.5 70.5 68-0 69.3 69.6 68.0 TABLE I. C.C. H,S. absorbed. 0.23 0.65 0.38 0.78 1.35 2.00 1-61 1.68 1.75 C.C. H,S per gram Yd. 2.2 2.74 3.9 4.56 6.75 7.16 8.52 11.8 17.6 Subsequent primary occlucion. C 1284 MAXTED 1 THE INFLUENCE OF HYDROGEN SULPIITDli OW It will be s e a that the1 occlusive power of palladium for hydrogen is diminished progressively by the presence of increasing concentra-tions of hydrogen sulphidel. Provided that the ratio of hydrogen sulphide to palladium is less than a certain critical quantity a volume of hydrogen varying with the concentration of the inhibit<ant is occluded at once in a normal manner as is the case with unpoisoned palladium.This primary prolcess of practically inst,antaneous absorption is followed by a secondary process during which a further quantity od hydrogen is gradually and slowly ocduded. Evidence will be adduced later in the preaelnt< paper that secondary absorption is due tol dissooiation of the1 occluded hydrogen sulphide. On passing through a critical va,lue of aboutl 8.5 C.O. of hydrogen SO 70 60 60 4 0 30 20 10 0 2 4 - 6 C.C. of H,S pcr gram of Pd. sulphide per gram of palladium primary absorption celases and seloondary olcdusion only is observed ; further on plotting occlusive power against hydrolgen sulphide content itl is seen that the influence of this inhibitant both on primary occlusion and on total oloclusion is linear in nature.This result which is analogous t o that obtained for decomposed hydrogen sulphidel is illustrated in Curvels I and I1 in Fig. 2. In oases whelre the concentra4tlion of hydrogen sulphide is con-siderably in excess of that necessary totally to inhibit primary occlusioln the firstl stages of secondary absorption may take place sufficiently slolwly to mask the process unless the) system is kept under olbselrvation folr a oonsiderabls time after admimion of hydrogen. The prolcress of wcolndary absorption after beginnin THE OCCLUSION OE” HYDROGEN BY PALLADIUM. PART 11. 1285 slowly gradually increases in velocity and finally the velocity diminishes to zero as the limit of total occlusion is approached.With concentrations of hydrogen sulphidej below the1 clritical rattio, secondary absorption may om the okher hand take place a t a velocitly sufficieliit to reInder necessary a very careful following of the moveineint of the( mercury in the hydrogen burette in order to map accurately the progress oi the) first stages of the prolcess and to differentiate i t from primary absorption. Measurements in which readings of the1 volume occluded were takea in such cases, every few seconds during the first stagea of secondary occlusion, FIG. 3. 0 10 SO 30 40 50 60 70 80 90 100 Time in minutes. have however shown that the general form of this absorption curve although more rapid.is similar t o that4 observed with higher concentrations of hydrogen sulphide in which secondary occlusion is slower and more easily followed. The form of the various absorption curves is well shown in Fig. 3 in which the progress of absorption of certain of the) measurements is plotteld against the ti me. In order to obtain information reapelcting the nature of the process involveld during secondary occlusion careful analyses were iriade of the gas remaining in the absorptio$n pipette after secondary absorption had taken place for the purpose of ascertaining whether VOL. OXVII. 3 1286 MAXTED THE INFLUENCE OF HYDROGEN SULPHIDE ON or not hydrogen sulphide is evolved during the! change, simultaneously with the slow absolrption of hydrogen.I n no case was hydroagen sulphide in appreciable quantity oibselrved and eixhaustion of the] gas follolwed by thhe introduotion of fresh hydrogen a150 faileld to) cause! the' frecsh increase in occlusive power whioh wonld be expected if selcolndary olcclusion were! duel either to the! gradual displacement of hydroigein sulphide by hydrogen or to the relduction of a sulphide complex with elvolution of the sulphur as hydrogen sulphide. The! process appears on the1 other hand to be due tot it change FIQ. 4. Time in minutes. in the condition of the occluded hydrogen sulphide itself in that the change in ocollusive power is found to take1 place1 spontaneously on standing. In order to investigate this point' three nieasurements were carried out with three specinlens of palladium which were in each case allotwed to! absorb about 11.8 C.C.olf hydrogen sulphidel per gram of palladium the1 subsequent8 admission of hydrolgen f o r the determinatioii o l the ocalusive power being delaye'd for six minutes, four hours and twenty-three hours respectively. The absorption ourves olbt'ainecl are reproduced in Fig. 4 from which it will be seen that whilst' the1 final voli~n~e of hydroigen occluded by pre-parations containing an approsiinately q u a 1 volume otf hydrolgen sulphide is the sariiel in each caw. the form of thel absorption curv THE OCCLUSTON OF HYDROG~N BY PALLAD~UM. PART IT. 1287 has been influeiiced by the time which has been allowed t o ellapse beltween treatment with hydrogen sulphidei and subsequelnt measure-ment of the1 occlusivs power f o r hydrogen almost complet0 suppression of the slow process of secondary occlusion having bem induced by alloiwing the system to stand.The1 nature of the change1 undergone by occluded hydrogen sulphidel during the1 comparativeily slow passage of th0 system a t 20° from a state of lower t o one of highelr occlusive power would seem to be! similar to) thatl obtained instantaneously at looo in whioh the hydrogen of the1 hydrogen sulphide has belea shown to be wolved as such the1 sulphur being retained by the palladium as a Pd,S ooimplelx. This view is suppolrteld by considering the dissociation relaction H,S -C 2H+S in connexion with the1 volume od hydrogen occluded during primary and secondary absorption. During this dissociation a volume of hydrogein equal whilst in the! moleoular condition to that of the hydrogen sulphidei wcluded is setl free and must be added to the volume of hydrogen subsequently absorbed in order to delrive the true ocolusive power of the preparation.The total volume of hydrogen thus contained in the palladium shoald if the relaction postulated is the correct onel fall on the1 known poisoining curve, caloulateld oln the1 basis of the1 presence of Pd,S complex in amount equal to the1 sulphur coiitained in the hydrogeln sulphide omluded. The dose agreelmeliit beltween the observed olcclusive powers and the values calculateid from the Pd,S curve is shown in tablet 11 the reeults being also plotted graphically in curve 3 of Fig. 2. TABLE 11. No. of expt.1 2 3 4 5 6 7 8 9 C.C. H,S per gram palladium. 2.2 2.74 3.9 4.56 6.75 7.16 8.52 11.8 17-6 Hydrogen occluded. C.C. 64.0 62.5 59.5 59.8 53.7 51-9 47.8 43.0 27-7 Total hydrogen. C.C. 66.2 65.2 634. 64-4 60-5 59.1 56.3 54.8 45.3 Total hydrogen calculated from Pd,S curve. 66.0 65.4 63.8 62- 9 60.0 59.5 57.7 53-3 45.6 For the purpose of the calculatioa required fotr the above table, the occlusive power for hydrogen of 1 gram of unpoisoaed palladium under the1 giveln conditions has been taken as 69 o.c., this value being in agreement both with the) measurements 3 B 1288 R~DIIAL AND HAWKINS CATALYSIS 1" THE: colntained in the present papelr and with t h m previously publisheid .The palladium in addition to fundioning as an occluding medium would thus appear to induce the catlalytia disslmiation of the ooduding hydrogen sulphide. This gradual dismcliation affords Blsol EL ready explanation of the non-elxistence of a definite1 solu-bility of hydrogen sulphide in palladium. It has previously been sho1w-n that after preliminary saturatioa a slow but continuous absolrption of hydrogen sulphide takes place occlusion of as much as 42 0.0. per gram being obselrved by allowing palladium to remain folr three wmks in excess of hydrogetn sulphide. It will be seeln tlhaii the gradual disappearance od occluded hydrogen suIphide as such due t o the passage of the sulphur into1 the Pd,S form should aausa the system to cease to be satarated for hydrogeln sulphide. Fresh hydrogen sulphide is thus olacluded and the process continues until the olocluding palladium has been eliminated as Pd,S complelx. It would be expeoted that the catalytic activity of the palla,dim, necessary for causing dissolciation woluld also decrease progressivelly as poisoning proceeds so that the velocity s f this dissooiatioln and consequently of the absorption of fresh hydrogen sulphide should rapidly diminish with inmaasa in the sulphur colntent of the palladium. This deoreaw in the velocity of the dissociation of hydrogen sulphide with increasing sulphur content is demonstrated in it striking mannelr by the curvea of Fig. 3. It should be emphasiseld that the coimplex Pd,S cannotl without furtshelr investigation be assumed to be a true sulphide of palladium the charaoter of this mmplex having yet to be detelrmineld . CHARLES STREET, WALSALL STAFFS. [Receiued September 30th 1920.