One of the most important processes in modern chemistry is the carbon-carbon bond-forming reaction, since it represents key steps in the construction of complex molecules from simple precursors. However not until the discovery and development in the 1970s of metal-catalysed cross-coupling reactions was there a simple and direct method for carbon-carbon bond formation between sp and sp³ centres. A wide variety of metal-catalysed cross-coupling reactions now exist and have become powerful tools, finding application in the synthesis of biological molecules, liquid crystals, macromolecules and also in supramolecular chemistry.

Pd-catalyzed cross-coupling reactions, such as Heck and Suzuki-Miyaura reactions, are used extensively in synthetic chemistry, and characterised by the coupling of an electrophilic species, such as organic halides, with a nucleophile, such as an alkene or organoboron, catalysed by Pd⁰. This catalysis is not confined to organometallic Pd complexes in the homogeneous phase; palladium colloids and supported Pd particles show good activity in Heck coupling (Scheme 1) and this proposal addresses the underlying fundamental surface chemistry.

Scheme 1: Heterogeneously catalysed Heck coupling reaction

Over the past year Dr. Zhipeng Chang, the post-doctoral fellow appointed to this grant, has continued to explore the surface chemistry of several aryl halides, allylic alcohols and aldehydes pertinent to Heck coupling reactions, over catalytically relevant Pd and Pt model surfaces. These studies have involved synchrotron Fast XPS and NEXAFS measurements at both Elettra (Italy) and Daresbury (UK) synchrotrons, in conjunction with laboratory termperature-programmed mass spectrometry in our York laboratory.

The surface chemistry of bromobenzene, one of the simplest activated aryl halides used in C-C cross-couplings, has been investigated over both Pd(111) and Pt(111) surfaces, alongside the unsubstituted benzene parent molecule. Time-resolved XPS shows that benzene adopts a single chemically distinguishable environment over Pt(111) within the monolayer, with a saturation coverage of 0.2 ML. This contrasts with Pd(111) wherein discrete tilted and flat-lying chemisorbed benzene form. Around 20% of a benzene monolayer desorbs molecularly over both surfaces, with the remainder dehydrogenating to surface carbon. Bromobenzene likewise adsorbs molecularly at 90 K over palladium and platinum, giving rise to two C 1s environments corresponding to the C-H and C-Br function. The saturation C6H5Br monolayer coverage is ~0.11 ML in both cases. NEXAFS reveals bromobenzene adopts a tilted geometry, with the ring plane around 60 ° to the surface. Bromobenzene multilayers desorb at 180 K, with higher temperatures promoting competitive molecular desorption versus C-Br scission within the monolayer. Approximately 30% of a saturated bromobenzene monolayer either desorbs reversibly or as reactively formed hydrocarbons. Debromination yields a stable (phenyl) surface intermediate and atomic bromine at 300 K. Further heating results in desorption of reactively formed H₂, C₆H₆, and HBr. Although no coupling products desorbed from Pt(111), biphenyl production was observed over Pd(111) around 500 K. Both surfaces are efficient for low-temperature bromobenzene hydrodebromination to benzene and HBr, and palladium also proved capable of homo-coupling the associated phenyl surface intermediates. This latter exciting discovery provides the first unequivocal evidence that metallic Pd surfaces are able to perform the crucial oxidative addition and reductive elimination steps in the C-C bond-forming catalytic cycle. The bromobenzene/Pt(111) work has been published in Journal of Physical Chemistry C, and that on Pd is in preparation for the same journal.

We have begun to study the adsorption and surface reaction of various functionalised alkenes commonly cross-coupled with bromobenzene in commercial Heck reactions. C 1s XPS and NEXAFS over Pd(111) reveal both crotyl alcohol (H₃CHC=CHC-OH) and its corresponding aldehyde adsorb molecularly at 100 K with the C=C p-bond parallel to, and the C=O projecting away from, the catalyst surface (Figure 2). These alkenes first undergo oxydehydrogenation followed by C-O bond scission (decarbonylation for the aldehyde) around 330 K. The resulting alkylidyne fragments subsequently dehydrogenate to leave a carbidic overlayer by 600 K. Alcohol dehydration is a secondary minor pathway competing with decarbonylation, and results in the desorption of but-2-ene and water. The analogous aromatic alkene, cinnamaldehyde (Ph-H₂C=CH=O) was also studied by Fast XPS and NEXAFS and likewise underwent decarbonylation upon heating, poisoning the catalyst surface. We have just submitted a letter on crotyl alcohol over Pd(111) to the Journal of Physical Chemistry C. We recently completed analogous surface science measurements of anisole (methoxybenzene) and bromoanisole over Pd(111). These are reactants currently employed for Suzuki coupling reactions in our parallel colloidal catalyst programme. These successful preliminary experiments put us in an excellent position for further laboratory and synchrotron investigations of coadsorbed arylhalide/alkene adlayers, and associated cross-coupled surface products.