The control of the solution electrochemical potential as well as pH impacts products in redox reactions, but the former gets far less attention. Redox buffers facilitate the maintenance of potentials and have been noted in diverse cases, but they have not been a component of catalytic systems. We report a catalytic system that contains its own built-in redox buffer. Two highly synergistic components (a) the tetrabutylammonium salt of hexavanadopolymolybdate TBA 4 H 5 [PMo 6 V 6 O 40 ] (PV 6 Mo 6 ) and (b) Cu(ClO 4 ) 2 in acetonitrile catalyze the aerobic oxidative deodorization of thiols by conversion to the corresponding nonodorous disulfides at 23 °C (each catalyst alone is far less active). For example, the reaction of 2-mercaptoethanol with ambient air gives a turnover number (TON) = 3 × 10 2 in less than one hour with a turnover frequency (TOF) of 6 × 10 −2 s −1 with respect to PV 6 Mo 6 . Multiple electrochemical, spectroscopic, and other methods establish that (1) PV 6 Mo 6 , a multistep and multielectron redox buffering catalyst, controls the speciation and the ratio of Cu(II)/Cu(I) complexes and thus keeps the solution potential in different narrow ranges by involving multiple POM redox couples and simultaneously functions as an oxidation catalyst that receives electrons from the substrate; (2) Cu catalyzes two processes simultaneously, oxidation of the RSH by PV 6 Mo 6 and reoxidation of reduced PV 6 Mo 6 by O 2 ; and (3) the analogous polytungstate-based system, TBA 4 H 5 [PW 6 V 6 O 40 ] (PV 6 W 6 ), has nearly identical cyclic voltammograms (CV) as PV 6 Mo 6 but has almost no catalytic activity: it does not exhibit self-redox buffering.

A recent report established that the tetrabutylammonium (TBA) salt of hexavanadopolymolybdate TBA4H5[PMo6V6O40] (PV6Mo6 ) serves as the redox buffer with Cu(II) as a co-catalyst for aerobic deodorization of thiols in acetonitrile. Here, we document the profound impact of vanadium atom number (x = 0–4 and 6) in TBA salts of PV x Mo12–x O40 (3+x)– (PVMo) on this multicomponent catalytic system. The PVMo cyclic voltammetric peaks from 0 to −2000 mV vs Fc/Fc+ under catalytic conditions (acetonitrile, ambient T) are assigned and clarify that the redox buffering capability of the PVMo/Cu catalytic system derives from the number of steps, the number of electrons transferred each step, and the potential ranges of each step. All PVMo are reduced by varying numbers of electrons, from 1 to 6, in different reaction conditions. Significantly, PVMo with x ≤ 3 not only has much lower activity than when x > 3 (for example, the turnover frequencies (TOF) of PV3Mo9 and PV4Mo8 are 8.9 and 48 s–1, respectively) but also, unlike the latter, cannot maintain steady reduction states when the Mo atoms in these polyoxometalate (POMs) are also reduced. Stopped-flow kinetics measurements reveal that Mo atoms in Keggin PVMo exhibit much slower electron transfer rates than V atoms. There are two kinetic arguments: (a) In acetonitrile, the first formal potential of PMo12 is more positive than that of PVMo11 (−236 and −405 mV vs Fc/Fc+); however, the initial reduction rates are 1.06 × 10−4 s−1 and 0.036 s–1 for PMo12 and PVMo11 , respectively. (b) In aqueous sulfate buffer (pH = 2), a two-step kinetics is observed for PVMo11 and PV2Mo10 , where the first and second steps are assigned to reduction of the V and Mo centers, respectively. Since fast and reversible electron transfers are key for the redox buffering behavior, the slower electron transfer kinetics of Mo preclude these centers functioning in redox buffering that maintains the solution potential. We conclude that PVMo with more vanadium atoms allows the POM to undergo more and faster redox changes, which enables the POM to function as a redox buffer dictating far higher catalytic activity.

Our group reported that the polyoxometalate Na 10 [Co 4 V 2 W 18 O 68 ]•26H 2 O (Co 4 V 2 ) is an active water oxidation catalyst and provided characterization of this system (J. Am. Chem. Soc. 2014, 136 (26), 9268). Two recent publications called into question the stability of Co 4 V 2 , one noting the miss-assignment of a 51 V NMR peak (Inorg. Chem. 2016, 55 (11), 5343) and another providing additional stability studies (ACS Catal., 2017, 7 (1), 7). We report here solution studies that further clarify stability limitations in this system by locating the correct 51 V NMR resonance of Co 4 V 2 and the other V-containing species present. Furthermore, we demonstrate that the observed catalytic activity cannot be explained simply by Co(II) aq , but arises from multiple active WOC species in solution. Key points about investigating such complex equilibrating aqueous catalyst systems are addressed.

Mixed 3d metal oxides are some of the most promising water oxidation catalysts (WOCs), but it is very difficult to know the locations and percent occupancies of different 3d metals in these heterogeneous catalysts. Without such information, it is hard to quantify catalysis, stability, and other properties of the WOC as a function of the catalyst active site structure. This study combines the site selective synthesis of a homogeneous WOC with two adjacent 3d metals, [Co 2 Ni 2 (PW 9 O 34 ) 2 ] 10− (Co 2 Ni 2 P 2 ) as a tractable molecular model for CoNi oxide, with the use of multiwavelength synchrotron Xradiation anomalous dispersion scattering (synchrotron XRAS) that quantifies both the location and percent occupancy of Co (∼97% outer-central-belt positions only) and Ni (∼97% inner-central-belt positions only) in Co 2 Ni 2 P 2 . This mixed-3d-metal complex catalyzes water oxidation an order of magnitude faster than its isostructural analogue, [Co 4 (PW 9 O 34 ) 2 ] 10− (Co 4 P 2 ). Four independent and complementary lines of evidence confirm that Co 2 Ni 2 P 2 and Co 4 P 2 are the principal WOCs and that Co 2+ (aq) is not. Density functional theory (DFT) studies revealed that Co 4 P 2 and Co 2 Ni 2 P 2 have similar frontier orbitals, while stopped-flow kinetic studies and DFT calculations indicate that water oxidation by both complexes follows analogous multistep mechanisms, including likely Co−OOH formation, with the energetics of most steps being lower for Co 2 Ni 2 P 2 than for Co 4 P 2 . Synchrotron XRAS should be generally applicable to active-site-structure-reactivity studies of multi-metal heterogeneous and homogeneous catalysts.

Multiple, fast electron transfers between phosphovanadomolybdates and Cu nodes in HKUST-1 lead to synergism in both activity and stability.