322 Drs. E. F. Armstrong and T. P. Hilditch .
the ' Philosophical Transactions/ and a number of others in the ' Proceedings of the Royal Society .
' Perhaps the most notable result obtained is that called by Dr. Chree the " acyclic change .
" This manifests itself on taking the averages of quiet days , when it appears that the mean value of the magnetic force is not the same at the end as it was at the beginning of the 24-hourly period , but shows a difference which is always in the same direction .
Owing to his knowledge and unique experience in the interpretation of magnetic curves , Dr. Chree is invariably consulted by explorers in the reduction anc publication of observations obtained in Arctic and Antarctic Expeditions .
A Study of Catalytic Actions at Solid Surfaces .
II.\#151 ; Transference of Hydrogen from Saturated to Unsaturated .
Organic Compounds in the Liquid State in Presence oj Metallic Nickel .
By Dr. E. F. Armstrong , F.I.C. , and Dr. T. P. Hilditch , F.I.O. ( Communicated by Prof. H. E. Armstrong , F.R.S.\#151 ; Received August 18 , 1919 .
) It has been shown in the first part of this series that hydrogenation in the liquid state at a surface of metallic nickel is to be considered as effected by the temporary union of the unsaturated organic compound and of hydrogen with the nickel , followed by a breakdown of this intermediate system into nickel and the saturated compound .
This explanation is a development of the " intermediate compound " theory of catalysis first put forward by De la Rive* and differs from the older view mainly in that it postulates an intermediate system of a very loose unstable type , similar to that produced between the natural enzymes , water ( or oxygen ) , and the compounds attacked by the latter , during enzymic catalysis .
It evidently applies also to other cases of organic compound catalysis , such as the ketonisation of organic acids in presence of alumina , thoria , or other metallic oxides , and the dehydration of alcohols to ethers or olefines in presence of the same class of oxides .
These catalytic actions have , indeed , been so explained by Sabatier and Mailhe , f on the basis of the simple " intermediate compound " theory .
* 'Ann .
Chim .
Phys./ vol. 39 , p. 328 ( 1828 ) .
t 'Compt .
Rend .
, ' vol. 150 , p. 823 ( 1910 ) ; ' Bull .
Soc. Chim./ vol. 13 , p. 319 ( 1913 ) .
A Study of Catalytic Actions at Solid Surfaces .
323 The dehydrogenation of hydrocarbons by finely divided nickel or palladium emed at one time to be more difficult to explain by theories of inter-lediate chemical association of the metal with the hydrocarbon , until the cperiments described below had been undertaken , when we were able to jrrelate this type of action with those which have just been mentioned .
Sabatier and Senderens* showed that cyclic hydrocarbons such as cyelo-exane or methylcyclohexane are converted by means of nickel at 70-300 ' C. into benzene or toluene and hydrogen with a certain amount E methane as by-product .
They explained the action as the withdrawal E hydrogen by the nickel on account of its tendency to form an unstable ydride .
In a later paper , Sabatier and Gfaudion f have extended their work to a umber of similar compounds at .350-360 ' C. , and state that the presence f hydrogen is necessary for the decomposition to proceed satisfactorily .
They have also observed that , when pinene vapour is carried by a current f hydrogen over nickel at 350-360 ' C. , the product consists of a mixture if cymene and cumene with a saturated hydrocarbon , probably a menthane , o that both dehydrogenation and hydrogenation were taking place simultaneously .
It is to be observed that , contrary to the hydrogenation process , catalytic dehydrogenation of these hydrocarbons is not by any means a complete operation , whilst the temperature at which it proceeds best , 270-300 ' C. , varies from the optimum range given by Sabatier for the dehydrogenation of ethyl alcohol in presence of nickel , 180-230 ' C. , and these facts suggest that the mechanism of each of the actions considered may not be of the same nature .
The evidence which we have obtained from time to time that metallic catalysts , during the hydrogenation process , interact primarily with the unsaturated organic compound , together with the resemblance of the whole process to enzyme action , led to the consideration whether the catalytic action , like that of certain enzymes , might not be reversible ; in other words , whether specific compounds , although " saturated " in the ordinary sense , might be capable of interacting with the metal to form a system which would break down into a more stable equilibrium , consisting of hydrogen and a less saturated compound .
This speculation has to some extent been substantiated by the observation that hydrocyclic compounds of the nature of those studied by Sabatier , Senderens and Gaudion ( loc. tit .
) , when heated in the liquid state with * 'Compt .
Rend .
, ' vol. 133 , p. 568 ( 1901 ) .
t ' Count .
Rend .
, ' vol. 168 , p. 670 ( 1919 ) .
Drs. E. F. Armstrong and T. P. Hilditch .
nickel and an unsaturated compound at a temperature much below that employed by these workers and in absence of hydrogen , partially yield up hydrogen which is absorbed by the unsaturated component of the system .
The compound which , amongst the few we have studied , lends itself most readily to the action is cyclohexanol ; at 180 ' C. a mixture of equi-molecular parts of cyclohexanol and methyl cinnamate in presence of nickel may be transformed into one in which about 10 per cent , of the cinnamic ester has become hydrogenated to methyl / 3-phenylpropionate , according to the ultimate equation :\#151 ; C6HnOH + C6H5.CH : CH.COO CH3 = C6H10O + C6H5.CH2.CH2.COO CH3 .
Cyclohexanol .
Methyl cinnamate .
Cyclohexanone .
Methyl \#163 ; -phenylpropionate .
According to the results of our experiments , it is necessary that both components of the system should be present in the liquid state , and no methyl / 3-phenylpropionate , or only very small traces , resulted when cyclohexanol was distilled through a mixture of methyl cinnamate and catalytic nickel .
The dehydrogenation of cyclohexanol can be regarded as removal of hydrogen either from the hydrobenzenoid nucleus or from the secondary-alcohol group present , and we have therefore examined other cases of hydrocyclic compounds in which the latter group was absent .
Cyclohexane could unfortunately not be employed in the apparatus available owing to its high vapour pressure at the necessary temperature ; distillation of cyclohexane through a mixture of methyl cinnamate and catalytic nickel led to a negative result , as in the similar distillation of cyclohexanol through the same mixture .
Dimethylcyclohexane ( hexahydroxylene ) and dihydropinene possess , however , sufficiently high boiling-points to admit of their employment in our apparatus , and it was found that , although little or no interaction with methyl cinnamate and nickel took place at 180 ' C. , a certain amount occurred at 230 ' C. Similarly , a mixture of ethyl stearate and methyl cinnamate , when heated at 230 ' C. with catalytic nickel , gave a product containing small quantities of methyl / 3-phenylpropionate and ethyl oleate .
It was also found that a mixture of the cinnamic ester and catalytic nickel was absolutely unaffected by passage of a current of ethane , either at 180 ' or 230-240 ' C. , but it is not possible to state definitely whether this is due to the incapacity of ethane to participate in the interaction , or to the fact that the ethane was not present in the liquid state .
Employment was , however , found for the ethane , which had been used in the last experiment , by utilising it as the inert gaseous medium in the reaction vessel in many of the other experiments .
In others carbon dioxide I A Study of Catalytic Actions at Solid Surfaces .
v'as employed , but the interaction seemed to proceed much better in iresence of ethane .
It cannot be definitely stated that carbon dioxide is ess suitable as an inert medium , since the difference may well be due to races of some " anticatalytic " impurity in the carbon dioxide ; the ethane ivould be free from the latter , having been prepared from ethylene by iatalytic hydrogenation over nickel , and then passed through liquid methyl jinnamate at 180 ' C. in presence of nickel .
Another case of simultaneous dehydrogenation and hydrogenation was observed by Zelinski and Glinka* in the action of spongy palladium on methyl tetrahydroterephthalate , when a mixture of methyl terephthalate and hexahydroterephthalate resulted .
This is obviously of the same type as the action now described , differing only in that the transference of hydrogen is effected between different molecules of the same , instead of varying , compounds:\#151 ; 06H8(C00 CH3)2 + 2C6H8(COO CH3)2=C6H4 ( COO CH3)2 + 2C6H10(COO CH3)2 .
Tetrahydro- Tetrahydrotere- Terephthalate .
Hexahydrotereterephthalate .
phthalate .
phthalate .
The most novel feature of the present experiments would appear to be that simultaneous dehydrogenation and hydrogenation have now been effected at temperatures not far removed from the general optimum hydrogenation range of 170-180 ' C. , and that hydrogen has been transferred from one compound to another , instead of from one molecule to another .of the same species .
Although we have no absolute proof that the mechanism of the change is not dependent on the production of hydrides of nickel , we consider that it is more natural to regard it as a further case of the catalytic equilibria which we have discussed above , depending on the formation of an intermediate system comprising the organic compounds , nickel and hydrogen :\#151 ; Saturated \#153 ; \#151 ; * I Saturated corn- compound + 1 *\#151 ; pound , Ni The equilibrium ultimately arrived at will depend on the resultant of the varying affinity for nickel of the saturated and unsaturated compounds involved .
The above doctrine of catalytic hydrogenation and dehydrogenation in the liquid state affords some explanation of the products obtained during the hydrogenation of unsaturated glycerides .
Mooref has shown that partial hydrogenation of ethyl oleate causes the formation of ethyl stearate * \#163 ; Ber .
, ' vol. 44 , p. 2305 ( 1911 ) .
t ' J. Soc. Chem. Ind. , ' vol. 38 , p. 320 ( 1919 ) .
Unsaturated compound , Ni hydrogen Unsaturated compound + Ni + hydrogen .
326 Drs. E. F. Armstrong and T. P. Hilditch .
and also of ethyl esters of isomeric forms of ordinary oleic acid , these being most probably ethyl elaidate and an ethyl " iso-oleate " derived from A11:12 oleic acid .
The appearance of the elaidic ester , the stereo-isomeric form of ordinary oleic ester , is readily explained by the assumed equilibrated action between catalytic nickel and the ethylenic linkage , for on reformation of the constituents during the balanced action it is plain that either or both stereo-isomeric ethylenic derivatives might result .
The displacement of the " double bond " along the carbon chain is more difficult to understand , but since we have established that catalytic nickel is capable of bringing about dehydrogenation simultaneously with hydrogenation , it is possible to offer a tentative explanation of this abnormal isomerisation .
As already stated , we have observed that transference of hydrogen from ethyl stearate to methyl cinnamate in presence of nickel proceeds to a very small extent at 230 ' C. , and probably in the penultimate phase of our conception of the hydrogenation process , represented on p. 325 as [ Saturated compound , TSTi ] , the nickel , being already associated with the freshly produced ethyl stearate , is capable of active dehydrogenation at temperatures lower than 230 ' C. There would then be formed a dehydrogenated ethyl stearate , the hydrogen liberated being transferred to more ethyl oleate .
The " dehydrogenated ethyl stearate " is , of course , according to this view , the ethyl A11:12 oleate isolated by Moore , dehydrogenation proceeding preferentially at this point of the stearic acid chain of carbon atoms .
We are not concerned here with the reasons for ' this preferential action , and unfortunately the amount of transformation effected , even at 230 ' C. , when starting from ethyl stearate and methyl cinnamate , is so small that it does not appear possible at present to produce sufficient of the dehydrogenated ester to isolate it in quantity and determine its constitution .
It is nevertheless suggestive that , as Moore has pointed out , the production of the isomeric form of oleic acid takes place to a greater extent the higher the temperature bf hydrogenation , and is also more marked , at equal temperatures , with palladium , generally a more vigorous catalyst than nickel .
Moreover , his observation that the final ratio of " iso-oleate " ( .
, mixed elaidate and A11:12 oleate ) , to ordinary ethyl oleate is different when ordinary ethyl oleate is the subject of hydrogenation from that obtained when ethyl " iso-oleate " is hydrogenated , is to be expected if this explanation is correct .
) A Study of Catalytic Actions at Solid Surfaces .
327 Experimental .
Methyl cinnamate was selected as the material to receive hydrogen for the following reasons:\#151 ; ( i ) It absorbs hydrogen in presence of nickel in the usual hydrogenation process smoothly and rapidly over the temperature range 120-240 ' C. , without appreciable formation of by-products , the ethylenic linkage only being acted upon .
( ii ) The separation of a mixture of methyl cinnamate and / 8-phenyl-propionate is somewhat mgre readily effected than that of most respective pairs of ethylenic and saturated compounds ; after removal of the other component of the system a preliminary separation was made by distillation of the esters in a vacuum , when with methyl / 8-phenylpropionate present , a mixture of esters relatively rich in the latter distilled at about 105-110 ' C./ 5 mm. , the remaining methyl cinnamate being recovered at 115-120 ' C./ 5 mm. The enriched mixture was hydrolysed with caustic soda and a mixture of cinnamic and / 3-phenylpropionic acids thus obtained .
/ 3-phenylpropionic acid is freely soluble in cold petroleum ether , whereas cinnamic acid is very sparingly soluble , and a final separation was made by this means ; the / 3-phenylpropionic acid so obtained was almost , but not completely , free from cinnamic acid , and melted in general at about 50 ' C. instead of 48 ' C. , the true melting-point .
The catalytic operation was carried out by mixing at about 50 ' C. , the desired proportions of methyl cinnamate and cyclic derivative with nickel catalyst , and displacing all air from the vessel by a current of ethane or carbon dioxide .
The apparatus was then closed and heated at the experimental temperature whilst intimate mixing of the solid catalyst with the liquid was maintained .
Subsequently the product was filtered from catalyst and then distilled , as described above .
The following experiments illustrate the results obtained during the investigation:\#151 ; Cyclohexanol.\#151 ; The cyclohexanol was prepared by hydrogenation of phenol , and had been freed from simultaneously formed cyclohexanone by treatment with sodium bisulphite solution .
50 grm. of cyclohexanol and 80 grm. of methyl cinnamate were heated with nickel catalyst in an atmosphere of ethane at 180 ' C. for hours .
On fractionation of the filtered product at 10 mm. there were obtained:\#151 ; ( 1 ) 36 grin .
, b.p. 58-60 ' C. ( 2 ) 30 grm. of liquid esters , b.p. 113-114 ' C. ( 3 ) 40 grm. of crystalline methyl cinnamate , b.p. 128-130 ' C. vol. xcvi.\#151 ; a. - 2 a Drs. E. F. Armstrong and T. P. Hilditch .
, Fraction ( 1 ) was extracted with a concentrated solution of sodium bisulphite , when the characteristic bisulphite compound of cyclohexanone , moderately soluble in the sodium bisulphite solution , was obtained ; on regeneration of the ketone there resulted about 3 grm. of cyclohexanone .
A portion ( 10 grm. ) of fraction ( 2 ) was hydrolysed , and eventually 1*5 grm. of / 3-phenylpropionic acid , melting at 50 ' C. , was produced .
This corresponds to a total production of 4'5 grm. of ^-phenylpropionic acid , or 5 grm. of methyl / 3-phenylpropionate .
In another experiment with 50 grm. each of methyl cinnamate and cyclo-hexanol , in presence of nickel catalyst , at 180 ' C. for 5 hours in an ethane atmosphere , about 12 grm. of the esters were hydrogenated .
It may be mentioned that the bromine absorption of the / 3-phenylpropionic acid obtained was in general about 2-3 per cent. Dimethylcycloliexane.\#151 ; This was obtained by repeated hydrogenation of xylene , followed by removal of unchanged xylene by a mixture of sulphuric and fuming sulphuric acid : it then boiled at 129-132 ' C. 50 grm. of methyl cinnamate and 25 grm. of dimethylcyclohexane were heated with nickel catalyst in an ethane atmosphere at 180 ' C. for 1|- hours , when it was found that no action had taken place .
A similar experiment was then made at 230 ' C. for about 1 hour .
In this case slightly more than 2 grm. of methyl / 3-phenylpropionate were found in the filtered product .
Dihydropinene.\#151 ; The pinene fraction of turpentine oil ( b.p. 155-157 ' C. ) was hydrogenated in the customary manner , the product being the desired dihydropinene ( b.p. 166 ' C. ) .
On heating equal weights of dihydropinene and methyl cinnamate with catalytic nickel in a carbon dioxide atmosphere at 180 ' C. , no formation of phenylpropionate was observed ; but at 230 ' C. , after analysis of the product in the usual manner , 3 grm. of / 3-phenylpropionic acid , melting at 50 ' C. , were obtained from 40 grm. of methyl cinnamate used in the experiment .
The hydrocarbons recovered fi ; om this experiment were soluble to the extent of 30 per cent , of their volume in concentrated sulphuric acid/ the original dihydropinene containing 8 per cent , soluble in this reagent .
Ethyl stearate.\#151 ; The ethyl stearate was prepared by hydrogenation of pure ethyl oleate , and melted at 34 ' C. 100 grm. of ethyl stearate and 50 grm. of methyl cinnamate were heated in an ethane atmosphere at 180 ' C. for 3 hours without formation of any / 3-phenylpropionate .
The recovered methyl cinnamate crystallised at once whilst the 90 grm. of ethyl stearate recovered absorbed 0'3 litre of hydrogen measured at 18 ' C. , on rehydrogenation .
I A Study of Catalytic Actions at Solid Surfaces .
329 A similar result was observed when the same proportions of the esters and catalyst were heated at 230 ' C. for 1 hour .
On heating these proportions at the higher temperature for 5 hours , however , a small transference of hydrogen from stearate to cinnamate was evident .
The products separated under 5 mm. pressure as follows :\#151 ; .
B.P. 105-108 ' C. , 5 grm. mixed cinnamate and / 3-phenylpropionate .
B.P. 115-128 ' C. ( mainly 125-126 ' C. ) , 24 grm. crystalline methyl cinnamate .
Residue , 97 grm. fatty acid ethyl esters , which re-absorbed 1*0 litre of hydrogen measured at 19 ' C. on rehydrogenation .
From the lower boiling fraction 0 8 grm. of / 3-phenylproprionic acid was obtained .
In another experiment , 100 grm. of ethyl stearate and 25 grm. of methyl cinnamate were heated with nickel catalyst in ethane at 250 ' C. for 1 hour , when a slight action occurred , about 0-5 grm. of phenylpropionic acid being isolated , whilst the residual ethyl stearate ( after removal of all methyl cinnamate by distillation in a vacuum ) absorbed P05 litres of hydrogen on rehydrogenation .
An experiment was also made at 180 ' C. , the apparatus being initially filled with hydrogen , in order to ascertain whether a slight initial hydrogenation might induce the transference action at a lower temperature ; the result was negative , about 1 grm. of / 3-phenylpropionic acid being obtained\#151 ; an amount less than that possible to be obtained from the 400 c.c. of hydrogen present .
A similar attempt to induce the process with hydrogen in the case of dihydropinene and methyl cinnamate was equally unsuccessful .
Errata in Part I ( p. 137 Page Line For Read 138 9 and 10 from bottom 25 per cent , palmitin 5-10 per cent , of saturated glycerides .
144 14 from top ( Ni lead\#151 ; CH : CH\#151 ; ) ( Ni , \#151 ; CH : CH\#151 ; ) .
144 In table , 2nd column 0-25 0-6 .
2 a 2

