Photoionization for benchmark studies in transition-metal catalysis

Detlef Schröder¹

¹Institute of Organic Chemistry and Biochemistry, Czech Academy of Sciences, Flemingovo nám. 2, 16610 Prague 6, Czech Republic

Keywords: catalysis, photoionization, synchrotron radiation, vanadium oxide

Abstract

The concepts for employing photoionization studies with synchrotron radiation for benchmark studies in transition-metal chemistry are outlined briefly. As an illustration, the exemplarily case of trimethoxo-vanadium oxide is presented, where photoionization data essentially helped to establish the entire thermochemistry of OV(OCH₃)₃ from the neutral compound to the quasi-terminal fragments VO⁺ and VOH⁺ in the gaseous phase.

1. Introduction

Experiments with VUV photons (7 - 60 eV) stemming from synchrotron sources are of outmost importance as a linkage between modern and advanced experimentation in chemistry and physics on the one hand and the more and more improving theoretical tools on the basis of quantum mechanics. Nowadays, one may in fact state that ab initio theoretical studies of a problem in main-group chemistry may be more adequate, more accurate, require less personal and infrastructure and are faster and cheaper than conventional experimentation.

Despite the enormous progress of quantum theory within the last two decades, these methods need testing and benchmarking for keeping standards as well as to warrant a continuous improvement. Moreover, the high standards of accuracy have meanwhile only been reached for main-group elements, whereas transition-metal compounds form a considerably more challenging task.

This is the point of linkage at which experiments with well-resolved VUV photons from a synchrotron source provide a junction between experiment and theory by means of the highly accurate determination of atomic or molecular quantities (such as ionization energies, vibrational levels, excited states etc.) or - in fortunate cases - even allow the determination of activation barriers of chemical reactions. While in main-group chemistry, such experiments thus present a test for existing theoretical tools, in transition-metal research the benchmarks derived from synchrotron experiments essentially stimulate the progress in the development of new methods.

2. Methods for the delivery of benchmark data

Instead of detailed descriptions of the beamlines or the experimental end-stations used at the synchrotron facilities, only the general concepts for the establishment of benchmark data will be introduced. With the availability of tuneable VUV photons from a synchrotron source, various chemical compounds can be excited and/or ionized. If samples are used which are sufficiently volatile in ultra-high vacuum at ambient temperatures (typically up to a few hundreds °C), photoionization by synchrotron radiation can be combined with mass spectrometric techniques, which ensures a highly sensitive detection on a single-event counting basis. As an example, consider a molecular species M having an ionization energy of 10.0 eV. Below the ionization threshold, the few M⁺ cations being formed can be attributed to impact from cosmic irradiation inside the apparatus (typical count rate 0.03 - 0.1 s^-1). Slightly below the ionization threshold, Rydberg states of the neutral molecule, M*, can be formed which may eventually autoionize to the molecular ion M⁺, but the cross section of these processes is usually very low. If the photon energy reaches the very ionization threshold, however, the M⁺ signal increases very rapidly to a plateau regime with typically several 10⁴ counts per second. When the analyzer of the mass spectrometer is fixed on the mass-to-charge ratio of the molecular ion M⁺, the photoionization threshold of M can thus be determined by monitoring M⁺ as a function of the wavelength (and the flux) of the ionizing photons. The typical precision amounts to about ± 0.005 eV for atomic and ± 0.03 eV for molecular species [1]; double photoionization to dications has less favourable threshold characteristics and thus at best ± 0.1 eV [2,3]. Thresholds for dissociative ionization are broadened by Franck-Condon effects and thus also only precise by ± 0.1 eV at best [1].

3. Case study: Ion thermochemistry of trimethoxovanadium oxide OV(OCH₃)₃

Vanadium-oxide catalysts play a very important role in a number of large-scale processes such as in the oxidation of methanol to formaldehyde, the oxidative dehydrogenation of ethylbenzene, or the industrial production of maleic anhydride. A key problem in these processes, both partical oxidations, is the minimisation of competing combustion processes eventually leading to CO_x. In this respect, the knowledge of the elementary steps of such oxidation reactions is of prime importance as it can help to increase the yields of the desired products, thereby reducing the amounts of byproducts, waste, and heat production.

One way to achieve detailed insight into the elementary steps of catalytic processes are gas-phase studies of small model systems, both by experiment and modern quantum theory. These extensive efforts, on the experimental as well as the computational sides, need some dedicated benchmarks for evaluation of the performance of the different methods. In this respect, photoionization experiments with synchrotron radiation can provide essential information which cannot be achieved by any other means. As a example, we refer to the trimethoxovanadium oxide, OV(OCH₃)₃, which can be regarded as a model system for C−H bond activations by high-valent transition-metal oxides [4].

Figure 1 shows the photoion yields of the molecular ion OV(OCH₃)₃⁺ (Figure 1a) and the primary fragment HOV(OCH₃)₂⁺ (Figure 1b), where the latter is accompanied with the loss of formaldehyde and hence represents an example of an oxidation reaction. Analysis of the photoion yields reveals thresholds of (9.56 ± 0.04) eV for the photoionization of the neutral compound to the cation and (10.1 ± 0.1) eV for dissociative photoionization to HOV(OCH₃)₂⁺ [5]. Both values provide accurate benchmarks for the calibration of theoretical methods, in that the former describes the energy demand for removal of one electron from the vanadium (V) compound, thus resembling defect formation in the solid state, and the latter turns out to be not due a thermochemical limit imposed by the exit channel but rather represents the height of the activation barrier for C−H bond activation.

Figure 1. Photoionization yields of (a) the molecular ion OV(OCH₃)₃⁺ and (b) the fragment ion HOV(OCH₃)₂⁺ as a function of the energy of the photons used to ionise neutral, gaseous HOV(OCH₃)₂⁺ [5].

The usefulness of the synchrotron data is demonstrated by the success of subsequent work [6], in which starting from neutral OV(OCH₃)₃ the combined expertise of experiment and theory could be used to establish the thermochemistry of trimethoxovanadium oxide from the bulk, neutral compound to the quasi-terminal fragments VO⁺ and VOH⁺, respectively (Figure 2).

Figure 2. Ion thermochemistry from the molecular ion OV(OCH₃)₃⁺ to the quasi-terminal fragments VO⁺ and VOH⁺.

4. Conclusions

The above example as well as related work [7,8] demonstrate that photoionization experiments using synchrotron radiation provide accurate reference data for the reliable testing and calibration of other experimental methods and for the critical evaluation of modern theoretical approaches. In this respect, the exploitation of synchrotron radiation for essays in transition-metal chemistry is just in its infancy and thus likely to essentially contribute to the future success in this field.

References

1. D. Schröder, J. Loos, H. Schwarz, R. Thissen, O. Dutuit, J. Phys. Chem. A 108, (2004), 9931.

2. J. Roithová, D. Schröder, J. Loos, H. Schwarz, H.-C. Jankowiak, R. Berger, R. Thissen, O. Dutuit, J. Chem. Phys. 122, (2005), 094306.

3. J. Roithová, J. Žabka, D. Ascenzi, P. Franceschi, C.L. Ricketts, D. Schröder, Chem. Phys. Lett. 423, (2006), 254.

4. M. Engeser, D. Schröder, H. Schwarz, Chem. Eur. J. 11, (2005), 5975.

5. D. Schröder, J. Loos, M. Engeser, H. Schwarz, C. Jankowiak, R. Berger, R. Thissen, O. Dutuit, J. Döbler, J. Sauer, Inorg. Chem. 43, (2004), 1976.

6. D. Schröder, M. Engeser, H. Schwarz, E. C. E. Rosenthal, J. Döbler, J. Sauer, Inorg. Chem. 45, (2006), 6235.

7. D. Schröder, J. Loos, H. Schwarz, R. Thissen, O. Dutuit, Inorg. Chem. 40, (2001), 3161.

8. Á. Révész, Cs. I. Pongor, A. Bodi, B. Sztáray, T. Baer, Organometallics, 25, (2006), 6061.

Acknowledgements.

This work was supported by the Czech Academy of Sciences (Z40550506).