Anchoring Pd nanoparticles onto support structures offers an effective means to tune their physicochemical characteristics and prevent on-stream deactivation e. g. by sintering. Supports employing porous architectures, acid/base character and/or surface redox chemistry e. g. strong metal sup­port interaction (SMSI), afford further opportunities to influence cata­lyst reactivity [170-173]. Mesoporous silicas are widely used to disperse metal nanoparticles [135, 136, 171, 174, 175]. The transition from low surface area, amorphous silica (200 m2g-1) to two-dimensional non-inter­connected pore channels (SBA-15) [73] and three-dimensional intercon­nected porous frameworks (SBA-16, KIT-6) [73, 176, 177] improved the dispersion of Pd nanoparticles and hence degree of surface oxidation and thus activity in allylic alcohol selox (Fig. 16), but had little impact on the mass transport of small alcohols to/from the active site. [135, 136] The high thermal and chemical stability of such mesoporous silica [178, 179] makes such supports well-suited to commercialisation. Pd nanoparticles confined within such mesoporous silicas demonstrate good selectivity in crotyl and cinnamyl alcohol selox to their respective aldehydes (>70 %), and excellent TOFs of 7,000 and 5,000 h-1 for the respective alcohols. Similar activities are reported for secondary and tertiary allylic alcohols, highlighting the versatility of silica supported Pd nanoparticles [51, 135, 136, 180-182]. Incorporation of macropores into SBA-15 via dual hard/ soft templating to form a hierarchically ordered macroporous-mesoporous Pd/SBA-15 was recently shown to promote the catalytic selox of sterically challenging sesquiterpenoid substrates such as farnesol and phytol via (1) stabilising PdO nanoparticles and (2) dramatically improving in-pore dif­fusion and access to active sites [158].

The benefits of mesostructured supports are not limited to silica, with ultra-low loadings of palladium impregnated onto a surfactant-templated mesoporous alumina (350 m2 g-1) generating atomically dispersed Pd2+ centres [137]. Such single-site catalysts were 10 times more active in cro — tonaldehyde and cinnamaldehyde production than comparable materials employing conventional (100 m2 g-1) y-alumina, owing to the preferen­tial genesis of higher concentrations of electron-deficient palladium [134, 137], due to either pinning at cation vacancies or metal ^ support charge transfer [183]. These Pd/meso-Al2O3 catalysts exhibited similar TOFs to their silica counterparts (7,080 and 4,400 h-1 for crotyl and cinnamyl al­cohol selox, respectively) [137], consistent with a common active site and reaction mechanism (Fig. 17).

Mesoporous titania and ceria have also attracted interest as novel cata­lyst supports. The oxygen storage capacity of ceria-derived materials is of particular interest due to their facile Ce3+^Ce4+ redox chemistry [173, 184-188]. Sacrificial reduction of the ceria supports by reactively formed hydrogen liberated during the oxidative dehydrogenation of alcohols could mitigate in situ reduction of oxidised palladium, and hence maintain selox activity and catalyst lifetime, with Ce4+ sites regenerated by dissociatively adsorbed gas phase oxygen [187, 189, 190]. Due to its high density, con­ventional nanocrystalline cerias possess meagre surface areas (typically ~5 m2g-1), hence Pd/CeO2 typically exhibit poor selox behaviour due to their resultant low nanoparticle dispersions which favour (self-poisoning) metallic Pd [189, 191, 192].