2178 RHEAD AND WHEELER: THE EFFECT OFCCXXVL-The Eflect of Temperature on theEquilibrium 2CO CO,+C.By THOMAS FRED ERIC RHEAD and RICHARD VERNON WHEELER.THE fact that carbon monoxide dissociates under the influence ofheat, yielding carbon dioxide and carbon, or, in other words, thefact that the reaction:is reversible, was discovered by H. Sainte-Claire Deville in 1864(Gompt. rend., 1864, 59, 873; 1865, 60, 317) by means of hisI' hot-cold " tube. He was able to observe only a small degree ofdissociation a t a temperature a little lower than the melting pointof silver, whilst at temperatures above 1000° none at all couldbe detected.Since i t was assumed that the degree of dissociation of carbonmonoxide, like that of carbon dioxide and steam, should increasewith increased temperature, doubt was cast on Deville's first experi-ments, and it was suggested that the formation of carbon dioxideand the deposition of carbon were due to the chemical action ofthe glaze of the porcelain tubes he employed.I n 1869 Sir Lothian Bell (Journ.Chem. SOC., 22, 203), whilestudying the reactions taking place in the blast-furnace, found thatsuch portions of the iron ore as had been subjected to the actionof carlon monoxide at comparatively low temperatures in the upperpart of the furnace were impregnated with carbon, presumablyarising from the dissociation of the gas. He thereupon institutedlaboratory experiments to determine the action of carbon monoxideon different oxides at different temperatures.As a result of these experiments, Bell was able to state that thereduced metal was as effective as the oxide ip determining thedecomposition of carbon monoxide, and he gave the equation of thereaction as being:GO, + c = 2c"o2co = co, + c.He also showed that the oxides of nickel and cobalt and thereduced metals acted in a similar manner to iron oxide and reducediron.The influence of temperature on the amount of decomposition ina given time was also studied by Bell, comparative figures beingobtained from the quantities of carbon deposited.This is wellshown in the following series of experiments, in which carboTEMPERATURE ON THE EQUIL~BR~UM 2co ;~r CO,+C. 2179monoxide was passed slowly over reduced iron at different tem-peratures during six hours :Tem1)erature ,.. , . , ,. . , , . , , , . . . . . . ... . ,. . , , .. . . . . . . . . . , , . .Carbon deposited. Grams per 100 grams of iron 4.7 181 95 6 0.3It is thus apparent that a low temperature favours the decom-position, a result which explains the failure of Deville and otherst o obtain evidence of dissocia.tion a t 1000°, and is in accordance withthe fact that the reaction 2C?O=CO,+C is exothermic. The heatof reaction is shown by the following eqmtions:250" 400" 500" 600" 800"(1). CO + 0 = CO, + 6s-0 Kg.C. units.(2). C + 0 = CO + 29.0 ,, 3 9whence 2CO = GO, + C + 39.0 ,, 9 )_____Since the reaction is reversible, an equilibrium must be establishedbetween the quantities of carbon dioxide and monoxide that canexist together in the presence of carbon; and, in accordance withvan't Hoff's principle of mobile equilibrium, the quantity of carbonmonoxide will be increased by lowering the temperature.The equilibrium a t different temperatures has been studied by0. Boudouard (Ann.C'hirn. Phys., 1901, [vii], 24, 5 ) , who has givenfigures for 650°, 800°, and 925O.I n studying the reaction 2CO-+CO, + C, Boudouard made useof iron, nickel, and cobalt as catalysts. The finely divided metalswere obtained by impregnating broken pumice with the nitrates andigniting, the oxides thus formed being afterwards heated in astream of carbon monoxide until reduction was considered to becomplete.For the experiments a t temperatures below 700°, glass tubes,6 to 7 cm. long and 1-5 cm.in diameter, were used, the total volumeof gas being from 12 t o 15 C.C. Above 700°, a porcelain tube,40 cm. long and of 2.4 cm. internal diameter, was employed, thepumice containing the catalyst occupying the middle 10 cm. of thetube, and the remainder being packed with broken porcelain.The main results were as follow:Carbon dioxide,Temperature. Catalyst. Duration of heating. per cent.44 5" iron 6 horns 100445 nickel 1 hour 100445 cobalt 1 7 9 100650 cobalt 7 Y , 61800 nickel 4 ,Y 6.7800 cobalt 4 9 ) 6 -5I n one experiment at 4 4 5 O , using a very small quantity of ironoxide (reduced by carbon monoxide) as catalyst, Boudouardobtained 52.3 per cent. of carbon dioxide and 47-7 per cent. ofcarbon monoxide remaining after six hours' heating2180 RHEAD AND WHEELER: THE EFFECT OFThe reverse reaction, CO2+C-+2C0, was studied in a similarmanner, but without the use of catalysts, purified wood charcoalbeing employed.The carbon was heated in an atmosphere ofcarbon dioxide in tubes sealed at one end, with the other end openand bent so as to dip under mercury, the object being to avoidbursting of the tubes due to increased pressure as the reactionproceeded.The results were as follow :Duration of heating, Carbon dioxide,Temperature. hours. per cent.650" 9 62'4650 12 61 *5800 6 6 -7800 6 6-3I n addition to the above, two experiments were made at 925O, inwhich a measured volume of carbon dioxide was passed in a, slowstream through the heated charcoal, and the resulting gases bubbledthrough baryta water, the barium carbonate that was precipitatedbeing afterwards weighed. Assuming that a single passage of thecarbon dioxide over the heated charcoal was sufficient to establishequilibrium, Boudouard calculated from one experiment 3.3 percent., and from the other 4.5 per cent., of carbon dioxide remainingin equilibrium with carbon monoxide over carbon at 925O.Boudouard's experiments prove that the equilibrium ratioCO/C'O, in contact with carbon is a function of temperature, andthe results are in general agreement with the laws respectingequilibria in gaseous systems.R.Schenck and F. Zimmermann (Ber., 1903,36, l), while studyingmore particularly the order of the reactions taking place, have beenable to prove that the oxides of iron, nickel, and cobalt are quiteineffective in determining the dissociation of carbon monoxide, andthat it is only the reduced metals that act catalytically.This is indirect opposition to the views advanced by Boudouard. A t thesame time Schenck and Zimmermann give results for the equilibriumat low temperatures (445O and SOSO) that are entirely at variancewith those of Boudouard. At 445O Boudouard regards the dis-sociation of carbon monoxide as complete ; while Schenck andZimmermann, using reduced iron as catalyst, obtained 52.8 percent. of carbon monoxide as the quantity remaining in equilibriumat that temperature. It is interesting to note that this figureagrees fairly well with that obtained by Boudouard when usingonly a small quantity of iron as catalyst (a result which he discards),and it seems probable that in those experiments, otherwiseinexplicable, in which he obtained complete decomposition of carbonmonoxide, the oxides of the metals used as catalysts were incomTEMPERATURE ON THE EQUILIBRIUM 2co co,+c.2181pletely reduced before the tubes were sealed, and that oxidation ofthe carbon monoxide took place.We considered it desirable to determine the equilibrium ratiomore accurately and for a greater number of temperatures, avoidingthe use of catalysts, for Boudouard's method of experiment did notappear to us to be calculated to give very accurate results, andhis figures were not in agreement with those obtained by us duringthe course of an investigation on the mcde of burning of carbonon which we are still engaged.The method we have adopted has been to circulate carbondioxide continuously over purified wood charcoal packed in aporcelain tube, and heated in an electric resistance furnace.We have obtained in this manner the following figures for thepercentages of carbon dioxide and monoxide that are in equilibriumin the presence of excess of carbon at different temperatures:Temperature.850"9009501000105011001200C:trbon dioxide. Carbon monoxide.Per cent.by volnme.r6-232.221 *320.590 '370'150.0693.779 i . 7 898.6899.4199.6399.8599 9 1The percentages are calculated as those of the nitrogen-free gases.The gases usually contained from 1 to 2 per cent.of nitrogen.I n Le Chatelier's general formula for equilibrium in gaseoussystems :L = the total heat of the reaction at absolute temperature Y'.P = the prcssure in atmospheres.N and N'= the Lumber of molecules on the left- and on the right-hand side of the equation.ml, ~ 2 ' ~ , , .. and n2, nf2, ... = the number of molecules of the differentsubstances taking part in the reaction, index 1 meaning the initial andindex 2 the final system.cl, c ' ~ , ... and c2, c'~, ... 5 the concentrations of the different sub-stances, icdices as above.I n the system:2co = CO, i- cnl&= 2 ; n2= 1 ; n',=O ; c',=O.and the expressionIf the system is in equilibrium at atmospheric pressure, Y = l ,( N - N)lo.g.,P=O2182 KHEAD AND WHEELER: THE EFFECT OFAssuming with Le Chatelier that the heat of reaction is constant,and introducing its value (39.0 Kg.C.units), the equation thenbecomes :i- log, !i! = k. 19,500I' c2The values for Ic calculated from our resulk are as follow:T.11 23"117312231273132313731473c1-0.93770.97780.98680'99410,99630*99850.9994c,.0.06230.02220.01320.00590-00370.00150.0006k.20.0120'3920 '2420'4420'3220.7020'65EXPERIMENTAL,The Epuilib rium Furnace.-In designing the equilibrium furnace,the two chief considerations were (a) the obtaining of a uniformtemperature, and ( b ) the attainment of rapid cooling of the gasesafter they had left the zone of reaction, in order to "fix" theequilibrium. We had, moreover, to recognise the fact that a ttemperatures above 1000° both porcelain and fused silica or quartztubes, such as we intended to employ for the reaction vessel, becomeslightly porous to gases.The construction of the furnace, which was made for us byMessrs.C. W. Cook and Co., at the University Engineering Works,Manchester, is shown in Fig. 1. It consists essentially of a glazedBerlin porcelain tube, 51 cm. long and of 28 mm. external and20 mm. internal diameter, wound with platinum wire, through whichan electric current can be passed. The winding is arranged so asto give a uniform temperature throughout the central portion of12 cm., and is carried on either side close up to the gunmetal water-jackets, J, being insulated from them by thin disks of porcelain.By winding the coils closer near each end than along the rest ofthe tube, we are able, when a fairly rapid stream of water is passingthrough the jackets, to obtain a sudden reduction in the temperatureof the tube from 1000° in the central uniform portion to 400° orless within a distance of 1.5 cm., while the temperature falls to below150° within a distance of 5 cm.This result is not attained solely as the effect of water-coolingand increasing the length of resistance wire a t the ends, but is inpart due to the double-jacketing arrangement, A , which is intendedprimarily to avoid any error due to porosity of the porcelain tubeat high temperatures. A nickel tube, 22 cm.long and of 7.1 cm.external and 5'7 cm.internal diameter, is fixed coaxially with thTEMPERATURE ON TEE EQUILIBRIUM 2 0 ) C0,t-C. 2183'2.f ar"Rcf-mat- - 2184 RHEAD AND WHEELER: THE EFFECT OFporcelain tube, and through the annular space a slow stream ofdry nitrogen is passed. The nitrogen, prepared by Harcourt’smethod, enters under a slight pressure through the central tube B,and issues at C, C through wash-bottles containing concentratedsulphuric acid. The passing of this stream of dry nitrogen, inaddition to ensuring that no oxygen or water vapour enters theporcelain tube if it becomes porous a t high temperatures, causes amore uniform distribution of heat throughout the length of thefurnace, an effect which is enhanced by the position of entrance ofthe gas.The furnace tubes are surrounded by a thick layer of kieselguhrto prevent loss of heat by radiation, and the whole is encased ina jacket of sheet iron.The carbon used throughout this research has been wood charcoalpurified by first digesting with concentrated hydrochloric acid (ina bolt-head flask fitted with a reflex condenser) to remove the ash;washing with distilled water ; and subsequently heating at 1000°in a stream of dry chlorine, washing, heating in a stream ofhydrogen, and finally in a vacuum at 1000° for forty-eight hours.It is crushed and sieved so as to pass through a 10-mesh sieve andremain on a 10-mesh, and about 6 grams are then loosely packedinto a thin tube of quartz, 12 cm.long and open at both ends, whichslides easily into the porcelain tube.After the insertion of the quartz sheath containing the carbon,a plug of silica, 16.5 cm.long and 1.9 cm. in diameter with a hole3 mm. in diameter drilled through the centre, is introduced a t eachend. These plugs serve to keep the carbon surface in position inthe zone of constant temperature, but they are intended moreespecially to cause the stream of gases, after passing over the heatedcarbon, to pass rapidly out of the tube, and thus ensure that theequilibrium determined shall be that of the experimental tem-perature recorded.The Measurement of Temperature.-The temperatures aremeasured by means of a platinum and platinum-rhodium thermcFcouple, and recorded by one of the Cambridge Scientific InstrumentCompany’s ‘‘ Thread-Recorders.” The couple is embedded in themiddle of the carbon, the leads being insulated by thin quill tubingof quartz, and the whole enclosed in a sheath of thin quartz, whichpasses easily through the bore of the plug P.Some little difficulty was at first experienced in maintaining aconstant t,emperature, owing to fluctuations in the voltage of theelectric current supplied to the furnace.Since the experimentsextended continuously over several days, or, in some cases, severalweeks, personal attentioL was found to be impossible, and a meanTEMPERATURE ON THE EQUlLlRRlUM 2CO C0,fC. 2185had to be devised of autoniaiically regulating the voltage. Themethod finally employed, for the suggestion of which we are indebtedto Mr. E.Muller, of the Cambridge Scientific Instrument Company,is as follows.The voltage of the main current is first cut down by the largeresistance, R (Pig. Z), t o within a small margin of that required toobtain the experimental temperature in the furnace. It then passesthrough the Nernst lamp steadying resist~ances, N , of which asufficient number are arranged in parallel to allow the requisitequantity of current to pass round the circuit. These steadyingresistances take as their normal current 1 ampere at 15 volts, whilstthe furnace, when hot, takes about 3 amperes. The exact numberthat are required to ensure perfect regulation and automatic adjust-ment of the voltage across the furnace terminals depends on theexperimental temperature employed ; there must be a sufficientFIG.2.It A 1number to ensure that the spirals of fine iron wire within theexhausted globes of each shall glow a dull red without becomingoverheated; for they depend for their action on the change inelectrical resistance that occurs in iron wire at a temperature ofabout 775O.After passing through the Nernst lamp resistances, the currentis divided, part going through the furnace, and part through theshunt, S, containing a rheostat. About equal quantities of currentp a s through the furnace and the shunt. The final adjustment ofthe voltage across the furnace to that required to obtain a giventemperature is made by means of the rheostat in the shunt, thealteration of which does not interfere with the main current, sincethe whole of it passes through the Nernst lamp resistances.This method has proved eminently satisfactory, the experimentaltemperature being maintained without any serious fluctuationscontinuously auring several weeks,VOL.XCVII. 7 2186 RHEAD AND WHEELER: THE EFFPCT OFGeneral Arrangement of A ppuratus.-The reacting gases arecirculated without interruption over the heated carbon untilequilibrium is attained. The general form of the circulationapparatus is that designed by one of us in conjunction withW. A. Bone for the investigation of the slow combustion of hydro-carbons (Trans., 1903, 83, 1074).The porcelain tube containing the carbon carries a ground glassjoint at each end held firmly in position by strong springs. Thesejoints make connexion on either side, through the mercury-cuptaps T, T' (Fig.l), with the cylindrical vessels C, C', each of200 C.C. capacity. These vessels mainly determine the capacity ofthe apparatus, and are cylindrical in form in order to allow ofbeing heated to drive off any traces of gas that may have a tendencyto stick to the glass.On the rightrhand side, connexion is made, through the calciumchloride drying-tube D, to the head of the Sprengel pump S. Onthe left is fused a long tube of wide bore, which passes horizontallyacross the front of the furnace and is then bent downwards at rightangles, forming a manometric tube, which stands over the delivery-tube of the Sprengel pump in the mercury trough M . A shortT-piece near the left-hand cylinder, closed by a mercury-cup tap,serves f o r the introduction of the gas.With the exception of the ground joint connexions to the porcelaintube, the apparatus is of fused glass throughout:The gases are drawn, by means of the automatic Sprengel pump,through the furnace, and delivered under mercury in the trough Minto the manometric tube B, whence they return along the horizontaltube H to the cylinder C, and are again drawn forward through thefurnace.The automatic Sprengel pump, the general constructionof which is described in the paper referred to above (Zoc. cit.,p. 1079), is actuated by suction produced by a doubleacting Gerykpump driven by an electric motor, and is so arranged that the headof mercury in the reservoir R allows of the gases being circulated atatmospheric pressure.The total volume of the apparatus, measured at Oo, is 570 c.c.;that of the packed porcelain tube 96 C.C.The Gus Analyses.-The carbon dioxide was prepared by droppingboiled concentrated sulphuric acid from a separating funnel into aboiled solution of sodium carbonate contained in an Erlenmeyerflask.The gas evolved waa passed through two sulphuric acidworms, and collected in a glass gas-holder containing concentratedsulphuric acid, over which it wa.s stored for two weeks before beingused for an experiment.The carbon monoxide was prepared from sodium formate, madTEMPERATURE ON THE EQUILIBRIUM 2CO CO,+C. 2187into a stiff paste with distilled water, by the action of concentratedsulphuric acid. It was washed through two worms containingpotassium hydroxide solution, and stored over sulphuric acid inthe same manner as the carbon dioxide.The gases remaining after an experiment were analysed volu-metrically in a, Bone and Wheeler gas analysis apparatus overmercury, from 200 to 300 measure of gas being taken for ananalysis.Carbon dioxide was estimated by absorption with as smalla quantity of aqueous potassium hydroxide as possible, or whenonly small quantities were present, by absorption with a concentratedsolution of barium hydroxide. Carbon monoxide was absorbed byam ammoniacal solution of cuprous chloride, prepared by passingammonia gas into distilled water containing the freshly precipitatedcuprous chloride in suspension until the latter was dissolved.Alittle ammonium chloride was added to the solution as thus preparedto reduce the tension of ammonia vapour. The gas was treatedtwice with small quantities of this solution, and afterwards washedwith dilute sulphuric acid.Any residue (which never amounted to more than 2 per cent.of the total gas) was afterwards exploded with a measured volumeof air and oxygen, to which a few C.C. of pure electrolytic gas wereadded. Any contraction in volume after explosion, or afterabsorption with potassium hydroxide, wits then determined. Atrace of hydrogen due to insufficient drying of the gases, or tomoisture in the circulation apparatus, was detected in severalexperiments the results of which were discarded.dlethod of Conducting am Expem’ment.-The apparatus havingbeen thoroughly exhausted, the glass being heated in a large blow-pipe flame to drive off the last traces of air, the carbon dioxide(or monoxide) is introduced in such quantity that when the reactionis complete the gases shall be as nearly its possible at atmosphericpressure. As a preliminary to a series of experiments, a, certainquantity of carbon dioxide is introduced to the carbon heated to1000° and allowed to circulate for several days.The resulting gasesare then pumped out, and the furnace brought to the experimentaltemperature. This preliminary treatment serves to remove anytraces of water-vapour that still remain in the apparatus; it wasin the gases resulting from such experiments that we were able todetect traces of hydrogen.In the experiments proper the gases are allowed to circulate fortwenty-four hours after the volume, ag indicated by a scale fixedbehind the manometric tube, has ceased to change.The reactiontube is then shut off from the rest of the apparatus, and samplesare withdrawn for analysis.7 D 2188 RHEAD AND WHEELER: THE EFFECT OFResults of Experiments.The results of our experiments can bestform aa follow :Experimentnumber.E 14E 9E 16E 4E 18E 5E 6Duration of heating,Temperature. hours.850" 240900 180950 1441000 481050 481100 481200 48be expressed in tabularCoiiiposition of resultinggases (calculated asnitrogen-free mixture). - co,. co.6.23 93.772 '22 97.781'32 98-680.59 99.410 -37 93.630.15 99.850.06 99 94In addition, w e have made two attempts to attain equilibrium a ta temperature of 800° with the circulation apparatus, but haveabandoned the experiments after they had continued for six weekswithout showing signs of coming to a conclusion, 20 per cent.ofcarbon dioxide still remaining after that time.Boudouard, in the reduction of carbon dioxide by wood charcoalwithout the presence of a catalyst, states that equilibrium wasattained in his experiments after six hours' heating at SOOO andafter twelve hours' heating at 650°, the percentages of carbondioxide remaining in equilibrium being 6.3 and 61.5 per cent.respectively at the two temperatures. The volume of his apparatuswas only from 12 to 15 C.C.as against our 570 c.c., but we do notthink that this fact is sufficient to explain the marked discrepancybetween our results, since we used a correspondingly larger quantityof carbon surface. I n another series of experiments that we aremaking on the rate of reduction of carbon dioxide by wood charcoalat different temperatures, we have been unable to obtain a dis-appearance of carbon dioxide of more than 0.7 per cent. after122 hours' heating a t 700O.All our experiments recorded above have been made startingwith an initial concentration of 100 per cent. carbon dioxide; forthe rate at which the reverse reaction proceeds was too slow toenable us to attain equilibrium in a reasonable time without thepresence of a catalyst, the use of which we wished to avoid.The relative rates of the two reactions during the initial stagesare well shown in the experiments recorded below.A temperatureof S50° was chosen as being that at which the reduction of carbondioxide by carbon was fairly rapid, and the dissociation oE carbonmonoxide readily appreciable.The rates of the reactions are calculated, by means of the relationZ C - log -2 = k,t CTEMPERATURE ON THE EQUILIBRIUM 2CO Z= CO,+C. 2189from the partial pressures (concentrations) of the carbon dioxide atdifferent time intervals in experiment R 13, and from the partialpressures of the carbon monoxide in experiment R 15 :Experiment R 13.co, -I- c = 2co.Time.( U n i t = l hour). P. at 0".0 258'61 292.32 317-84 356.36 389-08 415.812 439.5Temperature 8 5 0 O .pcoz.257'6224.9199*4160.9128'2101.477.7kCop -0.05900.05550'05110-05050'05060'0434Experiment R 15.2co = co, 3. c.Time.(Unit=I hour). P. a t 0".0 463'024 459.248 453.972 452'196 448'0120 447-2Temperature 850O.Pco. kco.453.7 --446'1 0*00030435-5 0.00037431.9 0.00030423'7 0-00031422.1 0'00026It will thus be seen that the reduction of carbon dioxide bycarbon takes place at 850° at a speed 166 times as great as thedissociation of carbon monoxide at the same temperature.We may incidentally draw attention to the fact that the goodt ctfor a unimolecular reaction, points to both reactions beingessentially surface phenomena; the rate of reduction of carbondioxide and the rate of dissociation of carbon monoxide both varyingdirectly with the partial pressure of the gas. It is our intentionto discuss this question more fully when we have concluded aresearch, on which we are at present engaged, on the relative ratesof reaction between carbon dioxide, carbon monoxide, oxygen, andcarbon at different temperatures.agreement of the constant k, calculated from the expression 1 -1og- CfJThis work has been undertaken in connexion with the experimentsnow being carried out by the Mining Association of Great Britainon coal-dust explosions. We are extending it to the investigationof the influence of pressure on the equilibrium ratio,ALTOFTS