Acetic acid decomposition on nickel(100 ... - ACS Publications


Acetic acid decomposition on nickel(100...

6 downloads 60 Views 1MB Size

J. Phys. Chem. 1987,91, 5531-5534 spectroscopy in a domain that is Fourier-conjugate to the spectral domain. The results of Figure 5 show that FTCRS, like other nonlinear spectroscopies employing interferometric schemes,"I0 also has potential in studies of dynamical processes. In effect, this potential exists because the Fourier transform of an FTCRS interferogram is related in a straightforward way to the Raman resonances of the sample (eq 4c and 5), and the Raman resonances contain information about dynamics (e.g., the r, in eq 6 ) . Given this, one notes that obtaining dynamical information with FTCRS is subject to some of the same problems that obtain in conventional spectral domain studies of dynamics (e.g., contributions from inhomogeneous broadening and pure dephasing). What is important is that FTCRS provides a means by which to obtain the information in situations where other methods may be difficult to apply, e.g., in studies of the ground-state dynamics of gaseous species in bulbs or in ultracold molecular beams. It has been pointed out in connection with eq 3, above, that there is a fundamental difference between nonlinear FTCRS and linear interferometric spectroscopies. Nevertheless, it is clear that there is also a fundamental similarity between the methods; each of the techniques makes use of a Michelson interferometer to obtain spectroscopic information in a domain that is Fourier conjugate to the spectral domain. In light of this similarity, we would make two points. First, much research has been aimed at characterizing and improving the spectral resolution, wavenumber accuracy, signal-to-noise ratio, instrumental distortions, etc., of linear interferometric methods.12 One expects that much of what has been learned through such work can be applied directly, or with minor modification, to the characterization and improvement of FTCRS. Second, since an interferometric version of Raman spectroscopy already exists (Le., FT-RamanI6), one might well question the need for FTCRS. To answer this, one notes that FTCRS is complementary to FT-Raman spectroscopy in roughly the same way that CRS is complementary to conventional Raman techniques.' For situations that require high spatial resolution,

5531

high spectral resolution, very efficient rejection of background light, high signal levels, and/or the study of Raman resonances with frequencies less than several hundred cm-l, FTCRS should be superior to FT-Raman spectroscopy. In closing, it is pertinent to point out that a consideration of the general basis underlying FTCRS (section 11) suggests the possibility of developing a variety of spectroscopic techniques based on nonlinear interferometry. One expects that if a nonlinear signal (1) depends on some product of laser spectral densities of the form [S1(wI)]"[S2(w2)]", and (2) there is some resonance condition(s) corresponding that allows only those products [S1(w1)]n[S2(02)]m to particular values of (ol- w2)to give rise to any signal, then by performing an interferometric experiment in which both lasers are directed through an interferometer before exciting the sample (Figure 1) one will obtain an interferogram that is modulated at the resonant difference frequencies of the sample. Indeed, results from this obtained by using an interferometric version of stimulated emission pumping spectro~copy,~' substantiate this expectation. Further work involving the application of FTCRS and other nonlinear interferometric techniques is in progress.

Acknowledgment. We thank Prof. M. A. El-Sayed for generous loans of equipment and Prof. P. Bernath for sending a copy of ref 12b to us. We also thank L. Connell, Dr. T. Corcoran, and B. Henson for their help. This work was supported, in part, by the donors of the Petroleum Research Fund, administered by the American Chemical Society, and by Research Corporation, du Pont de Nemours and Co., and the Academic Senate of UCLA. (26) (a) Felker, P. M.; Henson, B. F.; Corcoran, T. C.; Connell, L. L.; Hartland, G. V. Chem. Phys. Lett., submitted for publication. (b) Felker, P. M.; Hartland, G. V.; Connell, L. L.; Corcoran, T. C.; Henson, B. F. In Proceedings of the NATO Conference on Atomic and Molecular Processes with Short, Intense Laser Pulses, Bishop's University, Lennoxville (Quebec) Canada, July 20-24, 1987, to be published. (27) For example: Kittrell, C.; Abramson, E.; Kinsey, J. L.; McDonald, S.A,; Reisner, D. E.; Field, R. W.; Katayama, D. H. J . Chem. Phys. 1981, 75, 2056.

Acetic Acid Decomposition on Ni(100): Intermediate Adsorbate Structures by Reflection Infrared Spectroscopy Eric W. Scharpf and Jay B. Benziger* Department of Chemical Engineering, Princeton University, Princeton, New Jersey 08544 (Received: June 19, 1987)

Temperature-programmedreflection absorption infrared spectroscopy (TPRAIS),in concert with reflection absorption infrared spectroscopy (RAIS), was used to follow the sequence of stable surface intermediates for the decomposition of acetic acid monomer and dimer on Ni(100). After acetic acid monomer adsorbs molecularly at 170 K, the acid hydrogen is irreversibly lost at 240 K and a bridge-bonded acetate is formed. The bridge-bonded acetate then undergoes a reversible transformation to a monodentate acetate above 320 K which eventually decomposes to COz, C(ad), and H2at 435 K. Acetic acid dimer adsorbs molecularly at 170 K with the hydrogen-bonded ring approximately parallel to the surface. The dimer decomposes by dehydration at 255 K to adsorbed CO, a bridge-bonded acetate, and an adsorbed methyl group. The acetate decomposes to COz, C(ad), and H2 at 440 K. The key step in the acetate decompositions is the C-C bond scission. The dynamic infrared study shows the importance of performing the spectroscopy at reaction conditionsto identify the stable molecular configurations involved during the reaction.

Introduction Recently, there has been some interest in nickel as a catalyst to carbonylate methanol to form acetic acid.'q2 The decomposition of acetic acid on nickel is related to the reverse of this carbony(1) Fujimoto, K.; Omata, K.; Shikada, T.; Taminaga, H. 0. P r e p . Am. Chem. Soc., Diu. Pet. Chem. 1986, 3 1 , 85. ( 2 ) Rizkalla, N . P r e p . Am. Chem. Soc., Diu. Pet. Chem. 1986, 3 1 , 79.

lation reaction. Decomposition of acetic acid on nickel surfaces has been studied with temPerature-Progra"ed reaction (TPR) using the (1 1 and (1 Planes. Recent work by Schoofs and Benziger3 indicated that acetic acid monomer reacts by dehydrogenation to adsorbed acetate with subsequent decomposition (3) Schoofs, G. R.; Benziger, J. B. Surf. Sci. 1984, 143, 359. (4) Madix, R. J.; Falconer, J. L.; Susko, A. M. Surf. Sci. 1976, 54, 6 .

0022-3654/87/2091-5531$01.50/0 0 1987 American Chemical Society

Letters

5532 The Journal of Physical Chemistry, Vol. 91, No. 22, 1987

to C 0 2 , H2, and adsorbed carbon and oxygen while acetic acid dimer reacts by dehydration to adsorbed acetate, methyl, and CO. The following mechanism for monomer decomposition was also proposed

- +

CH,COOH(gas)

CH,COO(ad)

2H(ad)

CH,(ad)

C02(ad)

-

H2(gas)

CH,(ad)

CH3COO(ad)

+ H(ad)

C(ad)

+ COz(ad)

,/*H2(gas)

(1 - x)COz(gas) + xCO(ad) CO(ad)

-

+ xO(ad)

CO(gas)

with decomposition of adsorbed acetate as the limiting step. The current work attempts to clarify the structural detail of adsorbed intermediates present in the decomposition of acetic acid monomer and dimer on Ni(100) by reflection absorption infrared spectroscopy (RAIS) and examine the dynamics of the reaction with temperature-programmed RAIS (TPRAIS). RAIS and TPRAIS techniques allow the infrared spectra of species adsorbed on a reflecting surface to be measured down to submonolayer coverages. The theory, originally developed by Gree~~ler,~-' states that at high angles of incidence (-88") only p-polarized light can be absorbed. By taking the difference in intensity between reflected s- and p-polarized infrared light, the infrared absorption spectrum of the surface species can be measured, giving information about those vibrational modes with a component of dipole motion perpendicular to the surface. The TPRAIS technique developed by Benziger and SchoofsS can follow changes in infrared absorption of adsorbed monolayers during programmed heating. This paper reports the first application of the technique to follow the formation and decomposition of surface intermediates during the reaction of a molecule on a well-defined metal surface.

Experimental Section The reagent used in these experiments was glacial acetic acid of 100.0% nominal purity from J. T. Baker Chemical Co., lot 546818. This acid was further purified by several freezepumpthaw cycles before use. In a stainless steel ultra-highvacuum system, temperature-programmed reaction (TPR) spectra were taken with a UTI lOOC mass analyzer with surface cleanliness verified by Auger electron spectroscopy (AES). As with Schoofs and B e n ~ i g e rthe , ~ masses monitored were mlq = 2, 15, 18, 28, and 44, assigned to H2, CH,, HzO, CO, and COz respectively. Sample heating at a constant 7.1 K/s for TPR and 4 K/s for TPRAIS was provided by a tungsten filament positioned in close proximity to the back of the nickel crystal. The RAIS and TPRAIS experiments were performed in situ with a custom-built spectrometer capable of