604
TED€3. FLANAOAN AND B. S. RABINOVITCH
light, on one of tho photo-cells, PC2 or PC3. If tha light falls on PC2, a potential will be imprefisad on t,lw grid of t'he t i i h V1 (RCA t8ypa 1171;7/ M7-GT), in phase with the plate voltage so that crirrcnt passes which operates the relay R y l (Nlicd model BK-6A), with the result that the reversing niotor turns to produce an increase in the rmistnnce of the electrolytic rheostat, ER2. This diminishes slightslythe current in the coulometer circuit and causes the galvanometer mirror to shift the reflected line of light back again from photo-ccll PC2 to cell PC3. I n this condition the potentials impressed on the grid will be out of phase with the plate voltage of the tube, VI. This in turn releases the relay, Ryl, so that the motor, M, operates in the opposite direction. In this way the electrolytic rheostat,, ER2, is continilally recciving slight adjustments to maintain the current constant. A difference of potential at the galvanometer terminals of less than one-tenth microvolt will cause the motor, M, to reverse its dircction. Since the standard cell, SC, has a potential of about one volt, the regulation, as now used, is thus of the order of one part in ten million, under the best conditions. Momentum of the anode of the electrolytic rheostat, its motor and the coil of the galvanometer, causcs tho image of the filament to move somewhat more than the minimum noeded for reversal: one microvolt at the galvanometer terminals deflects the image by two centimeters. A conscrvntive estimate of the regulation attainable with this apparatus is thus one part per million. As shown, the photo-cells PC2 and PC3 are each in series with 15K protective resistances R8 and R11, and form a bridge with two 10K resistances, R9 and RlO. The operation of the apparatus is thus uninfluenced by the illumination of the room since changes in the background lighting produce substantially equal variations in the resistances of the two photo-cells. To protect the galvanometer and the standard cell in case of any failure in the apparatus, provision is made to release the contacts of the relay, Ry3, in case the light reflected from the galvanometer mirror reaches the photo-cells, P C t or PC4. Normally, current passes through the tube, V2, holding the contacts of the relay, Ry3, in closed
Vol. 61
position. However, illuminntion of PCt or PC4 will change the grid bins of thc tuheV2, thuscffccting the rclease of t,hese contacts with the resrilt that the galvanomctcr circuit is opened and the relay Ry2 breaks the circuit of the motor M. To make this system also independent of the background lighting the photo-cells PCt and PC4 with their series rcsistances form a bridge with the 10R resistances R9 and RlO. Normally the mid-point between R7 and R12 will be at zero potential with respect to the center of the bridge arms R9 Rnd RlO. If, however, the light from the gnlvanometer mirror reaches PC1 or PC4, this point will be subject to a fluctuating potential. This potential is converted into a pulsating negative bias on tube V2 with the germanium diode rectifier Re (lNGS), the 10M resistance R13, and a 0.0Fj microfarad condenser C2. This bias causes V2 to stop conducting, thus releasing the relay, Ry3, and o ening the galvanometcr and motor circuits. T e microfarad condensers C1 and C3 eliminate the chatker of the relays RyJ and Ry3. The batteries Ba2 and Be3 furnish adjustable biases on the grids of tubes VI and V2. As already stated, it is necessary for the purpose of this research that current be passing through the principal circuit, with the value determined by the quotient of the values of the potential of the standard cell and the standard resistance, R1, at the moment at which the current is started through the coulometer. To meet this requirement the switch S1 is turned to the position opposite to that shown in Fig. 3, eo that the current is passing through a variable resistance, R3, which must have a resistance equal to that of the coulometer, Co. I n order t o make the appropriate adjustment, a bridge system is provided by opening switch S1 and closing switch 84. Audiofrequency current from the generator Q is impressed on the bridge aa shown. Since the resistance arms R4 and R5 are both 100 ohms, equality of the resistances of Co and of R3 is secured by adjusting R3 and is indicated hy a minimum of sound in a telephone connected a t jack J1. Switch 54 is then opened, and 51 is put in the position opposite to that shown in the figure until it is desired t o pass current through the coulometer, when it is reversed.
b
E
EXCHANGE AND ISOMERIZATION OF ~ T U ~ ~ - E ' P " Y L E ON N E -NICKEL ~~ IN THE PRESENCE OF DEUTERIUM' BYTEDB. F L A N A Q A N AND 13. S. RABINOVITCH Conlribuwn from the Department of Chemialry of the Universily of IYashinglon, Seallle, Waahinglon Received Dsoambet I Y , 1866
The products of the exchange and isomerization reactions of trans-ethylene-& and deuterium on nickel wire can be aatisfactorily cxplained b assuming that the rcsctions proceed through an halE-hydrogenated state. The activation energics for isomerization andYexchsnge from 0-56" wnre found to be 6.0 and 7.3 kcal., re~pectively. In this range of temperature and extent of reaction investigated, hydrogen:ition and exceea exchnngc (ethylene-& minus ethylene-&) were unimportant. ( 1 ) Abstraoted in part from a thcals Rubmittcd b y Ted 8. Flanagan to the Graduate School in pmrtial fulfillment of the raqulrementa for &hedogrce or Doctor of I'hilosophy at the Uuivarsity of washington.
1
Introduction The exchange reaction, CZHd
+ Da
3
CzITaD
n
EXCHANGE OF IranS-ETHYLRNE-d2 ON NICKEL IN
May, 1957
+
HD, was discovered by Farkas, Farkas and Itideala in their study of the nickel-catalyzed hydrogenation of ethylene with deuterium. The rnechanisrn of this exchange has been controversial since its discovery. Much support has been given a reaction proceeding from associatively absorbed ethylene through a half-hydrogenated intermediate,3 e.g. HK!--CHz * * + D* +CHID-CHz + HzC-CDH
* *
+ H*
(1)
although there have been a number of different proposals regarding the mechanism of the over-all process. Jenkins and Ridea14 have recently presented evidence in support of an alternative mechanism suggested by their adsorption studies of ethylene on nickel. C&(g)
+ Q +DCHz-C+Hz
+CaHaD
+
Materials and Apparatus.-The ethylenes were the same as used in earlier work.618 Deuterium (99.50/0), obtained from the Stuart Oxygen Co., was passed over hot platinized asbestos and through liquid nitrogen traps before use; masss ectral analysis showed it to be free from contaminants. #he vacuum apparatus and catalyst (nickel wire) were the same as employed earlier. The catalyst was activated in tohesame manner and was protected at all times from stopcock grease and mercury with a cold trap. Procedure.-truns-Ethylene-& was frozen into the catalyst chamber, the required deuterium was added and the Iruns-ethylenedz was allowed to evaporate. To stop the reaction] the gases were removed by evacuation through a liquid nitrogen trap. The deuterium lost in this way was replaced with fresh material after each determination; eince the reaction was only carried to a small percentage reaction where excess exchange (defined below) was negligible, replacement of deuterium made no significant difference in the isotopic content of the deuterium. (2) A. Farkaa, L. Farkaa and E. K. Rideal. Proc. Roy. SOC.(London), A164, 630 (1934). (3) J. Horiuti and M. Polanyi, Tvans. Faraday SOC.,80, 1164 (1934). (4) C. I. Jenkins and E. Rideal. J . Cham. Soc., 2490, 2496 (1955). (6) T.B. Flanagan and B. 8. Rabinovitch. THISJOURNAL, 60, 724 (1956). (6) T. B. Flanagan and B. 8. Rabinovitch, ibid., 60, 730 (1956). (7) Various mechanisms were eliminated 86 possibilities in the earlier
work b y virtue of their inability to explain the observed ratios of isomerisation to exahange; upon further examination, however, it is found that a mechanism involving an assooiatively adaorbed vinyl radical. i.6.. CaHd(g) + CH-CH. H, cannot be eliminated for these *
+
DEUTERIUM 6AFi
The catalyst waa found to maintain its activity in the presence of deuterium] &B contrasted to the situation in the absence of deuterium, EO that no difficulty waa encountered in working with a catalyst of constant activity. Analysis.-The isomerization and exchange reactions were followed as before by infrared and maas-spectral analysis. Greater weight is attached to the experimental percentage of ethylene4 than those of ethylene-dl, eince the mass spectral analysis for ethylene4 was more accurate.
Results and Discussion Time Reaction Order.-Table I shows a typical run in which percentages of isomerization and exchange of trans-ethylenedz in the presence of added deuterium are shown at various times (0",5.8 cm. of each reactant) ; hydrogenation was negligible. The rate constants were calculated from the equations
(2)
In recent studies of nickel-catalyzed isomerization and exchange of trans-ethylene-d2 in the absence of hydrogen, Flanagan and Rabinovitch6J found strong experimental evidence favoring the half-hydrogenated state as an intermediate.' Pliskin and Eischenss have presented spectral evidence for the existence of the half-hydrogenated state when ethylene was added to a hydrogen covered nickel catalyst. Thus, there is strong but not unequivocal evidence regarding the half-hydrogenated state as an intermediate in the exchange reactions of ethylene in the presence of hydrogen. To obtain further information, a study of isomerization and exchange of trans-ethylene-da in the presence of deuterium was undertaken; circumstances have prevented as detailed a study as would have been desirable. Experimental
I
T H E P R E S E N C E OF
L
reasons but ia unlikely on other grounds (see ref. 11). (8) W. A, Pliskin and R. P. Eischena, J . CAsm. Phya., 24, 482 (1966).
(4)
where C n refers to an ethylene with n deuterium atoms, C, refers to cis-ethylene42 and Cao refers to the initial trans-ethylene-& These expresaions are for first-order reactions (reversible for isomerization), Equation 3 is solved by approximate method0 and (4) may be integrated directly. The rate constants for exchange refer to rates of inter-exchange; excess exchange, defined below, was unimportant where quantitative significa?ce was attached to the rate constants, e.g., activrttlon energy determination, The justification and adequacy of these approximate equations was discussed earlier.6 Relative Rates of Isomerization and Exchange.From Table I it may be seen that isomerization proceeded faster than exchange (ethylene-dl plus ethylene-da); this was also true in all other runs. The rates of isomerization and exchange of transethylene-dz were many times faster in the presence of added deuterium than in its absence, and isomerization and exchange were accelerated .in a quantitatively similar manner by the addition of deuterium. Both reactions were also affected equally by variation of catalyst activity. TABLE I Sample run (O', 6.8 cm. of deuterium and ethylene-&) co, kl x IO,, k x 108, CI, Ca,
Time, min.
4 10 12 14 16 18
%
%
..
0.64
1.2 12.6) 1.7 2.0 2.3
1.23 1.49 1.53 2.03 2.20
% 1.8 4.6 6.1 5.9 0.5 7.4
min.-l
min. -1
4.8 4.5 4.0 4.4 4.6 4.5
2.7 2.3 2.5 2.2 2.6 2.5
Inter-exchange and Excess Exchange.g-In spite of the presence of added deuterium, excess exchange was negligible under the conditions of the run in Table I; in runs with higher pressures of deuterium or at higher temperatures (>60°), excess exchange became appreciable (Table 11). The sample data of Table I1 refer t o runs made with a catalyst of es(9) The difference, ethylene-drethylene-di, will be called exaess exchange and twice ethylene-di will be called inter-exchange: "hydrogen" and "ethylene" will be used as general terms without regard to isotopic content.
TEDB. FLANAGAN AND B. S. RABINOVITCH
666
3.4
3.2
3.0
3.6
I / T x IO?
Fig. 1.-Temperature de endence of isomerization and inter-exchange of Irans-ethyfke-dp with deuterium at equal concentrations (2.1 X 1018 molec. om.-().
sentially constant activity; the catalyst activity was determined experimentally by the activity toward the reactions being studied at particular standard concentrations and temperature. Under given conditions, the relative proportions of excess and inter-exchange depend somewhat upon catalyst activity, although this effect was not examined in detail. It has been shown by previous workers that below 60" excess exchange is slower than hydrogenation and faster above 60°12which is a consequence of the high heat of desorption of hydrogen. Interexchange, however, still takes place much faster than hydrogenation at the lower temperatures.
Vol. 61
ried to about 8% isomerization; a t this (constant) catalyst activity and extent of reaction, both hydrogenation and excess exchange were unimportant. The data are given in Fig. 1. The activation energies are 6.9 and 7.3 kcal./mole for isomerization and exchange respectively, Mechanism.-From the results given above it is seen that for both isomerization and exchange the time reaction orders arc the same, the absolute magnitudes of the rates are closely similar, the variation of the catalyst activity affects each reaction similarly, both reactions are enhanced equally by the addition of deuterium and the activation, energies for isomerization and exchange are of the same magnitude (the small difference may be readily explained by the isotope effect for C-H, C-D bond rupture). It is concluded that in the presence, as in the absence of added hydrogen, isomerization and inter-exchange take place by the same mechanism. The observed ratios of isomerization to exchange may be explained by a mechanism in which the double bond is broken to permit isomerization. As before, a number of mechanisms cited in our earlier work, and which need not be listed again, may be immediately eliminated as not accounting properly for the product ratios. Some variations which have recently been proposed1° may also be eliminated because they predict either no isomerization or an incorrect product ratio. A mechanism involving a half-hydrogenated intermediate is successful in interpreting the present results for exchange and isomerization in the presence of deuterium. Such a mechanism was applied to these reactions in the absence of hydrogen and the values of 01, the C-H/C-D bond rupture ratios in the half-hydrogenated intermediate, were determined over a wide range of temperaturesa6 Using these values of a,the percentage of isomerircation may be calculated from the observed amounts of exchange and are compared with some experimental values in Table 111."
TABLE I1 EFFECTOF TEMPERATURE AND PRESSURE OF DEUTERIUM (IO) B. M. W. Trapnell, "Chemisorption," Butterworths Pub. Co., London, 1055, pp. 247, 248. ON EXCESS EXCHANQE AT CONSTANT CONCENTRATION OF (11) An associatively adRorbed vinyl radical intermediate is nomETHYLENE-& (2.1 X loi8MOLEC./CM.') inally Temp.,
oc. 0 0 22 22 36 36 56 56 90 90 157 157 157
Concn. deuterium/ ooncn. ethylene-
CI,
%
C:,
da
1 1 1 1 1 0.05 0.28 3.50
1.21 2.00 0.76 1.89 1.63 0.63 1.30 2.22 0.76 3.50 2.90 2.40 2.0
1.23 2.03 1.09 2.00 1.17 0.81 1.50 2.06 1.77 4.74 2.94 4.00 12.5
1 1 1 1 1
%
CI
-% CI, 0.02 0.03 0.27 0.11
(-0.46)
0.18 0.20 (-0.16) 1.01 1.24 0.04 1.60 10.5
Activation Energies.-Ra te constan ts have been determined from 0 to 56" for the isomerization and inter-exchange reactions a t equal and constant concentrations (2.1 X 10l8molec. cc.-l) of tronsethylene-& and deuterium. Most runs were car-
CZHd
H2YCH.
+H
compatible with the ratio8 of isomerination and exchange observed here and in our earlier work; however, this niechanism does not appear to be oorrect. The value of a', the (C-H)/(C-D) rupture probnbility necessary to explain the observed ratios if CHD=CHD is adsorbed as an associatively adsorbed vinyl radical, is related to P in the following manner a' = ( a P'
+ 1/a)+ daz + 1;
i n approximatcly 20 a t l o w temperatures. The values of u' required
to predict the proper ratios of product a t high temperature.+ are much higher than the rlanRialt1limit of 1.4 (t!iz., u' = 4.1 and u = 1.4at 430'); u' docs not aplironch the claReical limit if extrapolated to infinitely high temi,erature. Value8 of (C-H)/(C-D) ruptiirp probability determined for the gas plinse decomposition of 1,2-didariteroetliyl radicals at low temperature (€3. S. Rabinovitch and D. H. Dillfl, unpublished results) agree quite well with o valueR from the half-hydrogenated state intermediabe. Tho fact that u values appear to be independent of catalyet type, e.g.. Pd on charcoal gives tho Bamc u values ea nickel, rliigRest8 the comparison is valid. The usual arguments against diasociative-type mechanisma apply, c.0.. inability to explain double bond migration and the close relation of hydrogenation to exchange. The experimental evidence is vastly in favor of a half-hydrogenated state &B a n intermediate, but a n unequivocal differentiation between
L
t
Ir
May,
EXCHAN(:F, OF trUnS-ETHYIANP:-d2 ON
NICKELIN
THE PRESENCE O F DEUTERIUM
667
led ,Jenkins and Ridea14 to the conclusion that et,hylene is not afisociativcly adsorbed, but ad( h k l P A R I S O N OF S O M E EXPERIMENTAL A N I ) CALCULATETI sorbed irreversibly as acetylenic complexes. How1'P:nCENTAnES OF ISOMERIZATIONR ever, our earlier work, i.e., exchange in the absence Teaip., c1, Ca, CO(calc.), CO(exp.). C. LI % % % % of added hydrogen on supported nickel, nickel, 0 4.!1 0.69 0.7.5 3.6 3.2 films and nickel wire, shows that some reversible 22 3.9 1.7 2.1 7.2 0.2 adsorption of ethylene on a small area must take 22 3.9 0.76 1.09 31 3 . :J place on a fully-covered ethylene surface. This 36 3.3 0.63 0.81 2.4 2.6 reversible adsorption of et(hy1ene could still be of 56 2.8 1.3 1.5 4.2 4.8 t h e form C2H4(g) GHs, although Kemball14 56 2.8 2.1 2.0 0.8 7.2 To illustrnte the validity of the mechanism for excesa has pointed out that, associative adsorption may be exchange, some rum were especially chosen where excem a factor in the adsorption of ethylene in the presexchange was highest. The catalyst activity was in some ence of hydrogen, and that the Rideal mechanism nascs quite different from the runs of Table 11. in any case cannot explain the initial multiple exchange found under his low temperature (- 100") The tremendous increase in rate (100-1000 fold) conditions; spectral evidence has been' found8 for when deuterium is used is readily explained in adsorption when ethylene is exposed to terms of a half-hydrogenated intermediate since aassociative supported nickel catalyst. deuterium species, chemisorbed from the gas phase Remarks.-It is of interest to point on the surface, furnished many more sites for half- outConcluding t,hat the isotope effect is of importance in deh'ydrogenated state formation than is the case in termining the proportions of exchange in the reacthe absence of gaseous deuterium. If no equili- tions of ethylene and deuterium. The three inbration of the surface and gaseous hydrogen takes vestigations where initial product analyses on nickel place, excess exchange is negligible (the amount of have Seen made are those of Turkevich, et U Z . , ' ~ excess deuterium adsorbed on the surface being at 90", I
