Nanoclusters in catalysis: a comparison of CS(2) catalyst poisoning of polyoxoanion- and tetrabutylammonium-stabilized 40 +/- 6 A Rh(0) nanoclusters to 5 Rh/Al(2)O(3), including an analysis of the literature related to the CS(2) to metal stoichiometry issue

It is crucial in metal particle catalysis to know the true number of catalytically active surface sites; without this knowledge it is impossible (i) to know the true turnover frequency (TOF, i.e., the moles of product/(moles of active metal atoms x time)); (ii) to know for certain whether a (quantitatively) better catalyst has been made-on a per-active-metal-atom basis; (iii) to know the amount of active sites remaining in a deactivated catalyst; and (iv) to know how many active sites have been regenerated in a reactivated catalyst. For this reason, herein we report the first quantitative, more complete and fundamental study of nanocluster catalyst poisoning using the preferred CS(2) method with polyoxoanion- and tetrabutylammonium-stabilized Rh(0) nanoclusters; 5% Rh/Al(2)O(3) is also examined as a valuable comparison point. Both catalysts are examined under essentially identical conditions and while catalyzing a prototype reaction, cyclohexene hydrogenation. A number of control studies are also reported to be sure that the kinetic method used to follow the CS(2) poisoned hydrogenation reaction is reliable, to test for H(2) gas-to-solution mass-transfer limitations, to test for reversibility in the CS(2) poisoning, and to test for loss of the volatile CS(2). The results allow 10 previously unavailable insights and conclusions, including the first quantitative comparison of the active-site corrected TOF for a nanocluster catalyst (in this case Rh(0) nanoclusters) to its supported heterogeneous counterpart (the 5% Rh(0) on Al(2)O(3)). The results show that the nanocluster surface Rh(0) is between 2.3 and 23 times more active on a per-active-metal-atom basis. Overall, the results introduce to the transition-metal nanocluster area the catalyst poisoning methodology necessary for the determination of the number of active metal sites. The important literature of CS(2) catalyst poisoning studies is also cited and discussed with a focus on the previously neglected issue of the exact poison/metal stoichiometry ratio. Significantly, the single metal crystal plus CS(2) literature provides evidence that the CS(2)/metal ratio probably lies between 1/1.5 and 1/10 in most cases. The data presented herein suggest that the CS(2)/Rh ratio for the Rh(0) nanoclusters is very likely within this range and for certain is <1/17.