Report of the Electronic Structure- and Adsorption Group

Transcription

1 Report of the Electronic Structure- and Adsorption Group The silver oxygen system is one of the most well-studied in surface science due to its importance in ethylene epoxidation. 1 Many such studies have aimed to identify the types of oxygen present on and in the silver surface at the oxygen chemical potentials relevant for ethylene epoxidation. 1-5 Experimentally it has been found that two classes of oxygen species are present under such conditions. 1-2 The first, termed nucleophilic oxygen, is characterized by an O 1s binding energy of ~528.5 ev. It is active only in ethylene combustion. 3 The second type of oxygen, electrophilic oxygen, is active in epoxidation and is characterized by an O 1s binding energy of ~530.5 ev. The structure of the former is a well-known family of surface reconstructions, while the latter has long been attributed to oxygen on the unreconstructed surface (O ads ), in part due to scanning tunneling microscopy 6 and computed phase diagrams. 5 This assignment has help motivate a large number of computational studies of the mechanism of ethylene epoxidation on silver. 7 However, recent evidence suggests this assignment to be in error. 4, In an effort to unambiguously identify the O 1s binding energy of oxygen on the unreconstructed surface and place bounds on its coverage under oxygen chemical potentials relevant for ethylene epoxidation we have performed a series of UHV and NAP-XPS experiments coupled with ab initio calculations. Through this approach we find that the O 1s binding energy of O ads is < 528 ev on the Ag(110) surface, see Figure 1a and 1b, in agreement with our earlier results on the Ag(111) surface. 8 And while the coverage of O ads can reach 0.1 ML at 120 K, Figure 1a, at temperatures relevant for epoxidation (423 K) the kinetic limitations to surface reconstructions are removed and the measured maximum coverage is less than 0.04 ML, see Figure 1b, in agreement with the computed thermodynamic limit, see Figure 1c. 9 Figure 1: (a) O 1s spectrum of clean and oxygen covered Ag(110) surface. Temperature and oxygen dosing are indicated in the figure. (b) O 1s BE evolution with time measured in-situ for Ag(110) exposed to 10-6 mbar O 2. (c) Computed maximum O ads coverage as a function of temperature.

2 Thus, our results demonstrate the electrophilic oxygen shown to be active in ethylene epoxidation is not the O ads often modeled computationally. Critically, however, we were also able to identify a spectroscopic fingerprint for O ads and develop a method of preparing high (0.1 ML) coverage of the species. With this information we are now in a position to test the reactivity of O ads under artificially clean conditions and search for its presence in in situ experiments. The first plausible models dealing with selectivity of the ethylene epoxidation over silver appeared already in the 1940s. These postulated a special form of oxygen is involved in epoxidation 1,10,11 which became known as electrophilic oxygen. This oxygen species, characterized by an O 1s binding energy (BE) of ~530.5 ev is the only species known to produce epoxide during temperature programmed reaction, 3 yet its structure remained elusive until we were able to show it is the oxygen in adsorbed SO 4 (SO 4,ad ) that is part of a non-stoichiometric two-dimensional Ag/SO 4 phase where the sulfur is formally S(+V). 12 With the structure and function of electrophilic oxygen solved we have begun performing in situ XPS experiments aimed at uncovering both its role during steady state epoxidation and understanding the mechanism of SO x management during ethylene epoxidation. These experiments have demonstrated that under mbar pressures and during surface titrations with ethylene the epoxide selectivity tracks the coverage of SO 4,ad, as expected for the selective species, see Figure 2a. Furthermore, they have exposed a rich subsurface chemistry involving the exchange of dissolved and surface SO x species, see Figure 2b and 2c. Here a species with a BE of ev has been observed and assigned to subsurface SO, based on the computed BE of ev. The computed phase diagram of sub-surface species (Figure 2c) supports this assignment, showing that at the reaction conditions sub-surface SO is the preferred SO x dissolved species. This subsurface chemistry may play a critical role in mediating the coverage of SO 4,ad and hence epoxide selectivity. a) b) c) 220 ev 580 ev 750 ev SO4 SO S Figure 2: (a) Amount of detected CO 2 and C 2 H 4 O (by gas chromatography) vs SO 4 on silver powder at 503 K in C 2 H 4 :O 2 1:1 (total pressure 0.3 mbar) and selectivity evolution during time on stream. (b) S 2p taken at the same reaction conditions for different kinetic energies (indicated in the figure) of the emitted photoelectrons. (c) Computed phase diagram of the sub-surface SO x species.

3 In the research community, the discrepancy between higher operating pressures applied in catalytic processes and lower measurement pressures accessible during surface characterization is known in the community as the pressure gap 13. To bridge this gap, we have developed recently an experimental setup that provides chemical information on a molecular level under atmospheric pressure and in presence of reactive gases at elevated temperatures 14, see Figure 3a. This approach is based on separating a reaction volume form the vacuum environment by a silicon nitride grid that contains an array of micrometer-sized holes coated with a bilayer of graphene, following the same approach than in a EC-cell developed previously in our group 15. Using this configuration, we have investigated the local electronic structure of catalysts by means of photoelectron spectroscopy, and in presence of gases at 1 atmosphere. The successful operation of this setup was demonstrated with the hydrogenation of propyne on Pd black catalyst (powder), see Figure 3b. Figure 3: a) Gas reaction cell and SEM/TEM of Pd catalyst used b) Pd 3d XP spectra and gas chromatography traces using a mol sieve MS5 column and Al 2 O 3 /KCl column: b.1) He, b.2) He/H 2, and b.3) He/H 2 /C 3 H 4. EMIL, the Energy Materials In-Situ Laboratory Berlin is a concerted effort of the Helmholtz- Zentrum Berlin and the MPG (Fritz-Haber-Institut and MPI for Chemical Energy Conversion) to install and operate a unique facility at the synchrotron source BESSY II that is dedicated to the synthesis and characterization of materials and devices for energy conversion, energy storage and energy efficiency. X-ray photoelectron spectroscopy (XPS) and soft X-ray absorption spectroscopy (XAS) are among the most versatile methods for the investigation of surfaces on the atomic scale, providing quantitative information about the elemental composition and chemical specificity. The FHI operates at EMIL a state-of-the-art ambient pressure high kinetic energy X-ray photoelectron spectrometer ( AP-XPS ). With its modular design it is capable to be equipped with reaction cells optimized to study solid-gas and solid-liquid interfaces. Several of these reactions environments and cells have been constructed and put into operation within the last 2 years and are now available at EMIL.

4 Figure 4: Comparison of the measured and calculated photon flux of the soft X-ray undulator UE48 at a fixed gap of 31mm. A first milestone of the project has been reached by the first light from the soft X-ray undulator UE48 and the inauguration of the EMIL lab at the end of The EMIL beamlines offer a broad photon energy range from 80eV to up to 8000eV allowing utilizing photoelectrons at high kinetic energy as a probe of the interfaces 16,17. Two canted undulators serve as a source for two plane grating monochromators (PGM) and one liquid nitrogen cooled double crystal monochromator (DCM) to make this wide photon energy range available in one overlapping focus. In combination with a small entrance aperture of the differentially pumped AP-XPS spectrometer a high operation pressure can be achieved that for instance allows to work at the vapor pressure of water (around 20mbar at room temperature). Currently, the soft X-ray branch of EMIL is under commissioning. First results suggest that the performance is as designed (Fig. 4). The tender X-ray undulator is scheduled to be implemented in the storage ring BESSY in November References: 1. Bukhtiyarov V. I., Knop-Gericke, A. Ethylene Epoxidation over Silver Catalysts, In: Nanostructured Catalysts: Selective Oxidations, C. Hess, R. Schlögl, Editors, Cambridge: Royal Society of Chemistry 2011, pp Rocha, T. C. R. et al., Phys. Chem. Chem. Phys. 2012, 14, Bukhtiyarov, V. I.; Boronin, A. I.; Prosvirin, et al. J. Catal. 1994, 150, Andryushechkin, B. V.; Shevlyuga, et al. Phys. Rev. Lett. 2016, 117, Li, W.-X.; Stampfl, C.; Scheffler, M. Phys. Rev. B 2003, 68, Bare, S.; Griffiths, K.; Lennard, W. N.; Tang, H. T.; Surf. Sci., 1995, 342, a) Linic, S.; Barteau, M. A. J. Am. Chem. Soc. 2003, 125, b) Kokalj, A.; Gava, P.; de Gironcoli, S.; Baroni, S. J. Catal. 2008, 254, c) Ozbek M. O.; Onal, I.; van Santen, R. A. Top. Catal. 2012, 55, a) Jones, T. E.; Rocha, T. C. R. et al., Phys. Chem. Chem. Phys. 2015, 17, b) Jones, T. E.; Rocha, T. C. R.; et al., ACS Catal. 2015, 5, Carbonio E. A.; Rocha, T. C. R.; Klyushin, A.; Pis, I., et al., to be submitted 10. van Santen, R. A.; Kuipers, H. P. C. E. The Mechanism of Ethylene Epoxidation, In: Advances in Catalysis, D.D. Eley, Herman Pines, Paul B. Weisz, Editors, Academic Press 1987, 35, pp Sachtler, W. M. H.; Backx, C.; van Santen, R. A. Catal. Rev. 1981, 23, Jones, T. E.; Wyrwich, R.; Greiner, G.; Böcklein, S. et al., in preparation.

5 13. Stoltze, P.; and Nørskov, J. K.; Phys. Rev. Lett. 1985, 55, Velasco, J. J.; et al., Rev. Scj. Inst. 2016, 87, Velasco, J. J. et al., Ang.Chem. Int. Ed. 2015, 54, Follath, R.; Hävecker, M.; et. al. B. J. Phys.: Conf. Series 2013, 425, Hendel, S.; Schäfers F.; et al., AIP conference proceedings 2016, 1741,