One of the major goals of modern Surface Science has been the understanding of the surface chemical bond and its role in chemical reactions in heterogeneous catalysis, which have major economic and technological importance. Since the early days, the chemisorption of gas molecules such as CO, O2, H2... on transition metals has been studied extensively and the structure of several hundreds of chemisorption systems has been determined with a variety of techniques. Practically all the structural studies have been done under Ultra-High-Vacuum or High Vacuum environments. As catalytic reactions are performed at atmospheric pressures, the question of the "pressure gap" has been open for many years: are the UHV structures the ones that actually exist in the surfaces of the catalysts under real reaction conditions? To try to answer this question, experiments involving the transfer of a crystal from a catalytic reactor to a UHV environment have been performed extensively after the initial success by Somorjai et al. in the late seventies.

There is little doubt that in situ techniques can best reveal the surface structure at industrially important pressures and temperatures. These techniques have to be optical techniques (photons in the incoming and scattered beams) since the mean free paths of electrons or ions are too small at atmospheric pressures. Third generation X-ray synchrotron sources may provide the adequate probing beams to tackle the above question thanks to the intense X-ray beams available.

With this idea in mind, an UHV/HP (Ultra-High-Vacuum/High Pressure) chamber has been designed to perform grazing X-ray diffraction experiments on single crystal surfaces (Figure 42) on the beamline ID3 [1]. The sample surfaces can be prepared in UHV with the traditional recipes of Surface Science (ion etching, annealing etc.) and they can be pressurised to several bars in order to study the surface structures of the adsorbed gases and the possible modifications of the substrates induced by the gases. The chamber is mounted in a high-precision diffractometer to perform diffraction experiments with large scattering angles allowing the exploration of extended regions of reciprocal space. The rocking scan in Figure 43 is effectively the scattering from the topmost atomic layer of the Pt(111) surface under 0.33 bar CO. The peak sharpness and intensity show that the surface can be well prepared in UHV and then characterised at elevated pressure despite the beam passing twice through the 2-mm beryllium window.

The adsorption of CO on Pt(111) has been studied extensively in the past under UHV conditions [2]. The adsorption is non-dissociative. At low temperatures, several ordered structures have been reported which disorder via order-disorder phase transitions upon reaching room temperature. The maximum coverages for the different structures range from 0.5 to 0.7 monolayers.

The description of our main experimental result follows. A Pt(111) surface having terraces of several thousand Angstroms was prepared in UHV (p= 10-9 mbar). Then, with the sample at room temperature, CO was admitted into the chamber up to a pressure of 100 mbar, the crystal temperature was increased to 330°C for several minutes and then it was cooled down to room temperature.

The above CO-temperature experiment resulted in a very large decrease of the integrated intensity of the Pt crystal truncation rods. The intensity decreased by a factor of about 10 while the width of the Pt reflections was unaffected. A 7 x 7 R 19.1 degrees structure was formed on the surface. 82 in-plane reflections were measured which were reduced to 30 non-equivalent reflections after taking into account the symmetry and the domains. Also, six rods and three crystal truncation rods were collected.

Although the structural analysis is still being done and no model has been proposed so far, several comments are appropriate. To our knowledge, the above structure has never been observed before, in vacuum experiments. However, it has been observed in an electrochemical environment, with STM measurements on an electrochemical cell consisting of a Pt(111) electrode and CO gas bubbling in an acidic solution [3]. A simple model of the structure (to be confirmed) gives a coverage of 0.57 layers which is relatively high. The formation of the 7 structure has both a temperature and pressure requirement. The threshold of pressure has not yet been determined with precision but we have evidence that at least a few mbar of pressure are required. The temperature required is of about 250°C or more depending of the duration. It is also interesting to note, that the 7 structure is different from the high coverage structures predicted theoretically with Monte Carlo calculations [4].

In summary, by exposing Pt(111) to 0.1 atm of CO we have obtained and measured a new ordered structure which may be relevant to the understanding of the surface of a real catalysts under standard operating conditions.

References
[1] P. Bernard et al., Rev. Sci. Instruments, 70,1478 (1999).
[2] G. Ertl et al., Surf. Sci., 64, 393 (1977).
[3] I. Villegas et al., J. Chem. Phys., 101, 1648 (1994).
[4] B.N. Persson et al., J. Chem. Phys., 92 , 5034 (1990).

Authors
K. Peters, P. Steadman. H. Isern, J. Alvarez, O. Robach, P. Bernard, S. Ferrer.
ESRF