Ethanol formation mechanism from carbon monoxide + molecular

Antonin Deluzarche, Jean Paul Hindermann, Roger Kieffer, Raymond Breault, and Alain Kiennemann. J. Phys. Chem. , 1984, 88 (21), pp 4993–4995...
3 downloads 0 Views 339KB Size
J. Phys. Chem. 1984, 88, 4993-4995

Ethanol Formation Mechanism from CO

+ H, on a Rh/TiO,

4993

Catalyst

Antonin Deluzarche, Jean-Paul Hindermann, Roger Kieffer, Raymond Breault, and Alain Kiennernann* Laboratoire Chimie Organique AppliquZe, UniuersitZ Louis Pasteur, Strasbourg, France (Received: January 26, 1984; In Final Form: May 7, 1984)

Ethanol formation mechanisms via CO insertion in a methyl-metal bond or in a methoxy group is discussed with regards to Takeuchi and Katzer's results concerning the isotopic repartition of 13Cand I8O in ethanol formed from a mixture of l3CI6Oand '*C'80. In their analysis Takeuchi and Katzer concluded that CO was inserted into a surface carbene rather than into a surface alkyl since this last hypothesis could not yield the right isotopic distribution in the ethanol. But their results could be compatible with such a mechanism if we consider the possible exchange of I6O with I8O and vice-versa by a reversible reaction between water and adsorbed formaldehyde or acetaldehyde.

In a paper published in this Journal,' Takeuchi and Katzer used labeled carbon monoxide (I3Cl6Oand 1zC'80)to study the ethanol formation mechanism on a Rh/TiOz catalyst. The result were compared with two different models for the statistical repartition of the I3Cand lSOin the ethanol thus produced. Model A supposes the breaking of the C - 0 bond for all the molecules while in model B, the first C O molecule is dissociated and the second reacting molecule is not. This last hypothesis corresponds to the mechanism generally proposed for the ethanol synthesis from C O H2. It suggests CO insertion into a methyl group M-CH3. The experimental isotopic distribution obtained by Takeuchi and Katzerl for the ethanol produced is in accordance with hypothesis A but, as they mentioned,' it does not prove that a completely dissociative mechanism is real. Their results though enabled Takeuchi and Katzer to raise serious doubts (1) on the enolic condensation mechanism of AndersonZ

H

+

H

\C-CJoH

2H.d

-5

-

CH3\C/0H

II MII

2H.d

II

M

C,H,OH

C=O

1.

- iH3

--

CH3-CH-0

C2H50H

0

I

or (b) from a surface carbon obtained by a dissociated C O molecule with subsequent insertion of undissociated C O or an oxymethylene group as suggested by Biloen and S a ~ h t l e r . ~ None of these mechanisms can explain the formation of 12C1zC'60and 13C'3C180starting from a 12C'80 and I3CI6O mixture or with an initial mixture like the one of Takeuchi and Katzer' which gives too low concentrations of the former species. This led Takeuchi and Katzer to propose a complex mechanism where a M=CH2 species is formed and reacts as follows:

M

(2) or on the insertion into a methyl group. The M-CH3 could be obtained (a) from formyl species according to Henrici-Oliv6 and OlivC3 H

H

H

H

or according to Pichler and Schulz4 -tix 0

(1) Takeuchi, A.; Katzer, J. R. J . Phys. Chem. 1982, 86, 2438. (2) Storch, H. H.; Golumbic, N.; Anderson, R. B. "The Fischer-Tropsch and Related Synthesis"; Wiley: New York, 1951. Anderson, R. B. "Catalysis"; Reinhold: New York, 1956; Vol. 4. (3) Henrici-Olivt, G.; Olivt, S. Angew. Chem., Int. Ed. Engl. 1976, 15, 136. (4) Pichler, H.; Schulz, H. Chem.-Ing.-Tech.1970, 18, 1162. Schulz, H. Erdol Kohle, Erdgas, Petrochem. Brennst.-Chem. 1977,30, 123. Schulz, H.; El Deen, Z. Fuel Process Technol. 1977, 1, 45.

0022-3654/84/2088-4993$01 S O / O

d

CH3CH0d

These reactions explain the isotopic exchange of oxygen on the 'C and 13C,the 'zCH31zCH2160H,and the '3CH,13CHz'80H FO,

70-

obtained by reaction of the CH=CH and CHz-CH,

species as

(5) Biloen, P.; Sachtler, W. M. H. Adu. Cutal. 1981, 30, 165.

0 1984 American Chemical Society

4994

The Journal of Physical Chemistry, Vol. 88, No. 21, 1984

Deluzarche et al.

TABLE I: Calculated Isotopic Composition of Ethanol Synthetized from an Isotopic CO Mixture According to a Partially Dissociative Mechanism after Exchange with Water number of water-aldehyde exchanges possible

ethanol Oa

isotopes

12C12C'60 1.2 1ZCl3C160 13C12C160 25.4 13CI3Cl60 lzClzC,ao 48.5

48.2b

0

2 48.2

0

3

48.2

48.2

0

0

48.2

0

48.2

0

48.2

resultsC

13.3

13.5

23.6

12.8 26.2

23.6

12.9 26.2

13.7 23.5

13 26.2

18.4 t 1.0 24.2 t 0.5

31.1

28.8

28.3

26.9

26.8

26.0

26.1

25.4

25.7

25.2

25.5

25.0

26.1 t 0.4

23.8

25.8

23.8

25.9

23.8

26.0 23.8

26.1

23.8

26.1

23.8

26.1

23.8

21.8 .t 1.4

8.2

8.5

10.1

9.9

11

10.6

11.5

10.9

11.8

11.1

11.9 11.2

40.1

37.0

32.6

24.4

23.9

25.2

13C13C180 0.5

4.4

5.9

10.6

At 48.2%conversion.

+

aldehydes could be adsorbed. This was shown by CH&O or CH3CH=@ If we take into account such adsorbed intermediates, it is then possible to explain Takeuchi and Katzer's results by a mechanism consisting of CO insertion into a methyl-metal group or via carboxylate species. Furthermore, van den Berg and Glezer8 showed that acetaldehyde was formed with Rh catalysts. Indeed, the adsorbed aldehyde can react with a water molecule formed during the reaction (HZi60or H2I80)by the rupture of the C-0 bond:

12 + H 312C-CH=

18

0

H2180

,A H313CL3CH " ' 0 H 3I3C L 3 C H d 6 0

9.5 i 0.3

TABLE 11: Calculated Isotopic Composition of Ethanol Synthetized from an Isotopic CO Mixture According to a Carboxylate Pathway number of water-formaldehyde and -acetaldehyde exchangesa possible Takeuchi ethanol no and Katzer's isotopes exchange 1 t 1 2 + 2 5 + 5 resultsb

+

H,'2CL2CHA60

12

Reference 1.

suggested by Katzer.' This mechanism leads to the correct isotopic distribution for the ethanol formation, but we can formulate the following remarks: it makes use of a ketene species proposed and then abandoned by Blyholder and Emmett6 and not reused since; ethylene oxide is generally not observed as a product in the CO Hzsynthesis; and, finally, it cannot explain the formation of ethylene glycol only at high pressure. We presented7 a C O Hz reaction mechanism which takes into account the possibility of producing ethanol via carboxylate species or by CO insertion into a methyl-metal group. That mechanism has been supported by a number of surface species found and their reaction evolution using the chemical trapping methode7 Although, among the postulated oxygenated species, we could not find experimental evidence for the species corresponding to the chemisorbed aldehydes we suggested that these

16

48.2 12.5 26.2

24.0 26.2

H313CL3CH=0

and Katzer's

8

6

12.9 16.2 23.7

9.2 26.2

At 0% conversion.

_-

Takeuchi 5

12.1

7.5 24.4

13CIZCl80

0

4

11.1 12.2 23.7

5.5 26.1

izci3ciao a

1

0

H

+ H 313 C 1 3 C H d

eo

The acidic character of the TiOz usedg promotes that exchange and the kinetics of the reaction is so fast that it cannot be measured. In the example just mentioned, the exchange during the reversible addition of water on physisorbed aldehyde explains the and '3CH313CHz180H which could formation of 'ZCH312CH2160H not be explained by Oliv6, Pichler and S c h ~ l zor , ~ Biloen and Sachtler5 without an exchange of oxygen between water and (6) Blyholder, G.; Emmett, P. H. J . Phys. Chem. 1959, 63, 962; 1960, 64, 470. (7) Deluzarche, A.; Hindermann, J. P.; Kieffer, R. J . Chem. Res., Synop. 1981, 72; J . Chem. Res., Miniprint 1981, 934. Hindermann, J. P.; Deluzarche, A.; Kieffer, R.; Kiennemann, A. Can. J . Chem. Eng. 1983, 61, 21. (8) van den Berg, F. G. H.; Glezer, J. H. E. Proc. K. Ned. Akad. Wet., Ser. B Palaeontol.,Geol., Phys., Chem. 1983, 86, 227. (9) Katzer, J. R.; Sleight, A. W.; Gajardo, P.; Michel J. B.; Gleason, E. F.; McMillan, S . Faraday Discuss. Chem. Soc. 1981 No. 72, 8.

12C'ZC1b0 12CIJCl60 l3~1zc]60

5.7

11.3

12.6

13

18.4 f 1.0

28.7

27.7

26.9

26.2

24.2.t 0.5

40.1

28.7

26.0

25.0

26.1 t 0.4

27.0

23.8

23.9

23.8

21.8 z 1.4

10

11.4

12

13C13~160 12ClzC180 12C13ClSO

13C12Clao

13C13C'80

4.2

At 48.2% conversion.

9.5 z 0.3

Reference 1.

acetaldehyde. The participation of the support's oxygen cannot be important since Takeuchi and Katzer'O showed that the exchange l60-l8O with the catalyst is too slow. Table I shows the statistical results obtained for an increasing number of water-aldehyde exchanges. They were calculated from two different initial CO mixtures given in ref 1: [12C160]= 2.3 & 0.2 or 11.0 f 1.4; [13C160]= 45.1 f 0.2 or 41.4 f 4.0; [12C'80] = 51.5 0.4 or 38,8 4.0; [13C'80] = 1.1 0.2 or 8.8 f 1.0. These two repartitions corresponding to 0 and 48.2% conversion, respectively. It clearly shows that without any water-aldehyde exchange, the statistical isotopic distribution for the ethanol is far from the isotopic distribution obtained by Takeuchi and Katzer, but when equilibrium is obtained we find the same distribution as for their mechanism. The results of Takeuchi and Katzer are not in contradiction with an other proposition in our mechanism concerning the chain growth which is done by insertion of a CO molecule in a methoxy group.' In that mechanism there are two possible water-aldehyde equilibria: one with adsorbed formaldehyde and the other with acetaldehyde. There is also a l60-l8O exchange through the acetate species formation. For example

*

*

*

FH3

A-

I

I

iH3 - iH3 I

C H ~ C H ~ ~ ~ O c=160 H

i=lEo

-

c H3~ H 2 (10) Takeuchi, A,; Katzer, J. R. J . Phys. Chem. 1981, 85, 937.

1 8 ~ ~

J. Phys. Chem. 1984, 88, 4995-5004 This exchange at the acetate level cannot explain the formation without an 12CH3'2CH2160H and '3CH313CH21sOH exchange with water of the formed acetaldehyde or surface acetyl species. Table I1 gives the statistical repartition of the labeled carbon and oxygen in the ethanol obtained after reaction with the same CO mixture and for the same number of equilibria for C H 2 0 and CH3CHO. The isotopic repartition after five (5 5) water-aldehyde equilibria shows that the mechanism by carboxylate species cannot be rejected. Even if the results of Takeuchi and Katzer" show that there is a minor contribution to the ethanol production by a methanol homologation they cannot exclude the intermediacy of methanol precursors.

+

Conclusions It is shown with the results of Takeuchi and Katzer' that the mechanism consisting of a CO insertion into a M-CH3 group (1 1) Takeuchi, A,; Katzer, J. R.; Crecely, R. W. J. Catal. 1983,82, 474.

4995

cannot be rejected if we take into account a rapid reversible water-aldehyde reaction which is not the case in the previously cited work^.^-^ The methyl group can be formed in different ways: surface carbon, aldehyde, or methoxy species with the participation of the ~ u p p o r t .The ~ work of Takeuchi and Katzer, showing that the methanol cannot come from a surface carbon, combined with the recent results of Makambolz which showed a direct way from a methoxy to a methyl group and, finally, the presence of formyl species as evidenced in our laboratory on a 2% Rh-2% Ce02/Si02 catalyst lead us to suggest a mechanism through nondissociative adsorption of CO.

Acknowledgment. A bursary from the Hydro-Quebec Research Center to R.B. is gratefully acknowledged. We are indebted to one of the referees for printing out ref 8. Registry No. CO,630-08-0; Rh, 7440-16-6;ethanol, 64-17-5. (12) Makambo, L. Thesis Docteur es Sciences Universitd Poitiers, 1983.

Analysis of Slow-Motional Electron Spln Resonance Spectra in Smectic Phases in Terms of Molecular Configuration, Intermolecular Interactions, and Dynamics E. Meirovitchtl and J. H. Freed*$ Baker Laboratory of Chemistry, Cornell University, Ithaca, New York 14853, and Isotope Department, The Weizmann Institute of Science, 76100 Rehovot, Israel (Received: February 21, 1984)

ESR spectra from two oxazolidine derivative spin probes (CSLand 1,14-stearic acid) dissolved in the smectic phase of S2 were carefully analyzed and the results interpreted in terms of ordering characteristics, molecular conformation, and dynamics by using powerful and comprehensive spectral simulation techniques. The rigid CSL is highly ordered in the smectic A phase as expected for strong interactions with the rigid aromatic cores of the liquid-crystal molecules and shows reorientational motion well approximated as Brownian in the mean potential of the solvent molecules. The principal axes of ordering of CSL are found to be tilted with respect to the principal axes of the magnetic tensor, and the Euler angles specifying this tilt could be determined, because of the sensitivity of the slow-motional spectra to these parameters. The considerablesensitivity of these spectra to the shape of the orienting potential is also demonstrated, and values for coefficients in the expansion of the potential in the spherical harmonics through L = 4 are estimated. The flexible 1,14-stearic acid probe is only weakly ordered in the smectic phase. It shows an anisotropy in its rotational diffusion tensor that is smaller than predicted for an extended all-trans conformation, suggesting an average configuration with a decreased length-to-width ratio. However, the end segment containing the N-O group appears to be in an all-trans configuration,judging by the observed collinearity between the principal symmetry axis of diffusion and ordering with that of the hyperfine tensor. Its low ordering and reduced activation energy for reorientation in the smectic phase suggest primary coupling to the alkyl chain region of the solvent molecules. The spectral simulations are improved by introducing asymmetry in the viscosity, but related anomalies suggest the likelihood of some other mechanism such as a fluctuating torque model. A comparison of the results of this and other studies is made to show how, by the use of different spin probes, one can obtain insights into the intermolecular interactions operating in anisotropic fluids.

I. Introduction Anisotropic fluids differ from isotropic liquids in that there is a mean ordering potential in the former case. It leads to preferred spatial orientations being imposed on the constituent molecules. Various spectroscopic methods, including nuclear magnetic resonance (NMR) and electron spin resonance (ESR), have been employed to elucidate ordering characteristics in liquid-crystalline phases.'J The anisotropic parts of the N M R interaction are at most of the order of several hundred kilohertz. Molecular reorientation in liquids, occurring at rates of the order of 107-1010 s-l, is, therefore, rapid on the time scale of this technique, and the N M R spectra are determined by motionally averaged Hamiltonians. Thus, the appearance of the spectrum is not directly + Weizmann Institute.

* Cornell University.

0022-3654/84/2088-4995$01.50/0

affected by the dynamic rates, so the experimentally measurable spectral shifts are related to products of mean geometric parameters (e.g., averaged conformations of chain segments) and of ordering tensors representing the molecular orientation in the mean potential due to the surrounding molecules. It is not possible to decouple these in a straightforward manner. Assumptions regarding molecular conformation need to be made to extract in(1) (a) Doane, J. W. 'Magnetic Resonance of Phase Transitions"; Owens, F. J., Poole, C. P., Farach, H. A., Eds., Academic Press: New York, 1979; p 171. (b) "Introduction to Liquid Crystals"; Priestly, E. B., Wojtowicz, P. J., Sheng, P., Eds.; Plenum Press: New York, 1969. (c) Diehl, P.; Khetrapal, C. L. In "NMR, Basic Principles and Progress"; Diehl, P., Fluck, E., Kosfeld, R., Eds.; Springer-Verlag: West Berlin, 1969; p 3. (d) Mantsch, H. H.; Saito, H.; Smith, I. C. P. Prog. Nucl. Magn. Reson. Spectrosc. 1976, 11, 211. (2) "Spin Labeling: Theory and Applications"; Berliner, L. J., Ed.; Academic Press: New York, 1976 and references therein.

0 1984 American Chemical Society