Femtosecond time-resolved photodissociation dynamics of methyl halide molecules on ultrathin gold films

The photodissociation of small organic molecules, namely methyl iodide, methyl bromide, and methyl chloride, adsorbed on a metal surface was investigated in real time by means of femtosecond-laser pump–probe mass spectrometry. A weakly interacting gold surface was employed as substrate because the intact adsorption of the methyl halide molecules was desired prior to photoexcitation. The gold surface was prepared as an ultrathin film on Mo(100). The molecular adsorption behavior was characterized by coverage dependent temperature programmed desorption spectroscopy. Submonolayer preparations were irradiated with UV light of 266 nm wavelength and the subsequently emerging methyl fragments were probed by photoionization and mass spectrometric detection. A strong dependence of the excitation mechanism and the light-induced dynamics on the type of molecule was observed. Possible photoexcitation mechanisms included direct photoexcitation to the dissociative A-band of the methyl halide molecules as well as the attachment of surface-emitted electrons with transient negative ion formation and subsequent molecular fragmentation. Both reaction pathways were energetically possible in the case of methyl iodide, yet, no methyl fragments were observed. As a likely explanation, the rapid quenching of the excited states prior to fragmentation is proposed. This quenching mechanism could be prevented by modification of the gold surface through pre-adsorption of iodine atoms. In contrast, the A-band of methyl bromide was not energetically directly accessible through 266 nm excitation. Nevertheless, the one-photon-induced dissociation was observed in the case of methyl bromide. This was interpreted as being due to a considerable energetic down-shift of the electronic A-band states of methyl bromide by about 1.5 eV through interaction with the gold substrate. Finally, for methyl chloride no photofragmentation could be detected at all.

: Auger electron spectra recorded from clean Mo(100) surface (black curve) and from a 2 ML Au film on Mo(100) (red curve). The inset shows the Au(N 7 VV) to Mo(M 5 N 3 N 5 ) Auger peak intensity ratio as a function of Au deposition time. The break point in the intensity ratio plot corresponds to the initial appearance of Au multilayers.
The electronic structure of the Au overlayer was also analyzed by means of fs-laser photoemission spectroscopy. The photoemission spectra measured in this laboratory have been reported previously [1,2] and exhibit similar features as the two photon photoemission spectra that have been obtained by Cao et al. from an Au (111) surface [3]. The work function deduced from the photoemission spectra was 5.02 ± 0.2 eV at an excitation wavelengths of 263 nm. This values is within 0.2 eV of that previously published by Hansson et al. (5.22 ± 0.04 eV) [4].
In the present work, gold was evaporated with a rate of 0.0125 ML/min on the Mo(100) substrate held at 400 K. Figure S2a-c shows LEED images recorded at 90 K from 1 ML gold coverage at different annealing temperatures. As can be seen, the freshly prepared surface ( Figure S2a) appears amorphous because just a pale spot originating from the 00 reflex appears in LEED image. After 1 min annealing time at 800 K, the diffraction pattern corresponding to the (1×1) Mo(100) structure can be observed in Figure S2b.

S5
Further annealing at 1000 K for 10 min leads to the appearance of a c(2×2) structure in the LEED image as can be seen in Figure S2c for the 1 ML Au on Mo(100) preparation. A simulation of the c(2×2) LEED pattern employing the software LEEDpat 2.26 [16] is shown in Figure S2d and e. A similar c(2×2) structure was reported for Cu and Ag adsorbed at submonolayer coverage on Mo(100) as well, and it was in this case attributed to a surface rearrangement [10,17]. Subsequent investigations showed that the c(2×2) structure appears also for 0.5 ML Au/Mo (100) and it was attributed to the formation of a surface alloy in this case [11,15]. First principle calculations demonstrated that a c(2×2) substitutional surface alloy is energetically favored compared to overlayer structures [14]. Charge transfer from the Mo substrate to the adatoms is responsible for stabilizing the c(2×2) vacancy array, in which the adatoms are incorporated. Furthermore, the atomic radius of Au (1.44Å) is comparable with the atomic radius of Mo (1.39Å), which allows Au atoms to insert within the surface layer and to form a planar alloy structure [11].
A further increase in the Au coverage on the Mo(100) substrate induces another structural change of the surface. Figure S3a,b shows the diffraction patterns obtained from a 10 ML Au film on Mo(100). The freshly prepared surface appears amorphous, similar to the 1 ML preparation (cf. Figure S2a). After 1 min annealing at 1000 K, a new structure appears in the LEED image, which becomes more apparent when the annealing time is increased to 10 min ( Figure S3b). In Figure S3c and d a LEEDpat simulation of a c(2×8) Au/Mo(100) structure is shown. The simulated c(2×8) structure resembles the marked structure presented in Figure S3b (see also Figure S3e). LEED and low energy electron microscopy investigations demonstrated for the case of Pd adatoms adsorbed on the Mo(100) that a change from a c(2×2) to a c(2×8) structure occurs by increasing the Pd coverage above 0.7 ML [14]. The c(2×8) structures was observed just for a deposition rate below 0.033 ML/min at substrate temperatures larger than 800 K, whereas at higher deposition rates the surface structure was significantly different. The Pd/Mo(100) c(2×8) structure persisted at a coverage of 2.5 ML. The surface structure at coverages higher than 2.5 ML was not reported.
A detailed analysis of the LEED pattern obtained in the present work reveals a more complicated structure. Nevertheless, the obtained surface structure of the 10 ML Au film could be attributed to a degenerate c(2×8) structure comparable to the one observed Wu et al. [14]. The LEED patterns presented in Figure S3 are perfectly reproducible and are not influenced by a further thermal treatment of the surface, i.e. a higher preparation temperature or a longer annealing time.