Direct 2H NMR Observation of the Proton Mobility of the Acidic Sites of Anhydrous 12‐Tungstophosphoric Acid

Flip and hop before you′re caught! The proton dynamics of solid 12-tungstophosphoric acid (TPA) is probed by solid-state 2H NMR in a temperature range from 293–503 K. Protons of TPA are shown to be involved in two types of molecular motion (see picture): the anisotropic local two-site flipping between the two possible orientations of the OH bond at bridged oxygens of the Keggin anion, and the isotropic diffusion by hopping between neighboring surface oxygens of the anion. Heteropolyacids (HPAs) represent a promising class of solid acids with a particularly high catalytic activity in acid-type reactions such as hydration, dehydration, alkylation, isomerization, and sulfhydration.1 As strong solid acids HPAs provide an environment-friendly alternative to traditional homogeneous catalytic processes, for example, alkylation or isomerization,1a that involve large quantities of anhydrous hydrofluoric or sulfuric acids. HPA complexes are also intensively studied as proton-conducting materials for proton exchange membrane fuel cells.2 In both cases, the catalytic activity and the proton conductivity of HPA is governed by the structure and the mobility of its protonic species that form active Brønsted acid sites on the surface of the HPA anion. Among the family of HPAs, 12-tungstophosphoric acid with Keggin-type heteropolyanion structure (H3PW12O40) exhibits the strongest acidity and overall stability, thus being of major interest for industrial applications.1a, 3 The structure of 12-tungstophosphoric acid (TPA) is established as being composed of a Keggin unit with a central PO4 tetrahedron surrounded by twelve WO6 octahedra.4 The negative charge of the Keggin unit (−3) is neutralized by acid surface protons in the form of hydroxy groups. In addition to the four central oxygen atoms there are three types of oxygens on the TPA surface: twelve terminal oxygens (Ot), bound to a single tungsten atom, and two types (12+12) of bridging oxygens, both linking a pair of tungsten atoms. In one type the tungsten atoms are sharing a central oxygen atom, whereas in the other they are not. The crystalline structure of dehydrated TPA is stable up ∼570 K.5 Upon hydration TPA may crystallize into a secondary structure with six water molecules per Keggin unit being the most stable under normal conditions.5 Despite good knowledge of the TPA primary and secondary structures, both in the hydrated and dehydrated state,5–6 the precise location of the acidic protons on the surface of the Keggin anion in anhydrous TPA is still actively debated.7 As neutron scattering failed to give substantial information on the locationof the acidic protons, other techniques were applied. Kozhevnikov et al.7a stated that the protons are located on the terminal oxygen atoms by using 17O NMR spectroscopy. The same conclusions were made by Ganapathy et al.7b based on the results of 1H/31P and 31P/1H REDOR NMR experiments and density functional theory (DFT) calculations of the proton affinity (for a single proton). In opposition to these results, infrared spectroscopy,7c solid-state NMR,7d and extended Huckel molecular orbital calculations7e indicated that TPA protons are located on the bridging oxygen atoms. The most recent findings partially clarify this controversy by DFT calculations (considering the position of all three protons) and solid-state NMR,7f, 7g pointing out the presence of protons on both terminal and bridging oxygen sites. While in the literature major attention was paid to the localization of the acid protons of anhydrous TPA, their dynamics, which is vital to solid-acid catalysis,8 was scarcely studied. In the anhydrous state, the elementary step for proton mobility is the process in which the proton passes from one oxygen atom to another along the Keggin-unit surface or moves between neighboring Keggin units. The only experimental evidence of such motion was reported by Uchida et al.,9 who by means of 1H magic angle spinning (MAS) NMR spectroscopy estimated the hopping rate at ∼200 Hz at 298 K. The upper threshold for the hopping rate at 298 K was estimated to be 1 kHz with 2H solid-state NMR by Ueda et al.7d A recent advance in the problem of proton mobility in TPA was a DFT computational study by Janik et al.,3c, 10 in which the possible mechanisms of the proton mobility were discussed. The activation barrier for the proton transfer between neighboring surface oxygens in the Keggin anion was estimated at ∼100 kJ mol−1. Below 450 K the computed rates were <100 Hz and they became almost immobile at 298 K. The only experimental results on proton mobility at temperatures close to that of catalytic reactions were reported by Chuvaev, et al.11 However, the method used to characterize the mobility was based on the analysis of the second momentum of the wide-line solid-state 1H NMR spectrum, which provided no insight on the types and geometry of the molecular motions. Thus, it remains a great challenge to provide detailed information on the proton dynamics in the temperature range of reactions mediated by solid TPA. In this work, we clarify the details of proton motion (deuteron in our case of deuterated TPA) and its rate in solid TPA. We used solid-state 2H NMR to directly evidence and quantify the mobility of the acidic surface deuterons in anhydrous TPA. To the best of our knowledge, this is the first experimental report clarifying the details of proton mobility in anhydrous TPA over a wide temperature range. The experimental 2H NMR line shape for solid D3PW12O40 at 293–403 K shows (Figure 1) a single Pake-powder pattern with a quadrupolar constant Q=205 kHz7d and an anisotropy parameter η∼0. 2H NMR spectra of D3PW12O40 at 293–423 K. In the absence of dynamics (τC, characteristic time of deuteron reorientation, >10−3 s) the 2H NMR line shape depends only on the nature of the chemical bond (OD).12 The observed line shape reflects the fact that all deuterons have similar bonding. The spectrum remains static and unchanged up to 403 K, which underlines the fact that the rate of any possible motion is below 1 kHz within the temperature range from 293–403 K. Above 403 K, the spectrum is broadened but remains otherwise virtually static: the effective quadrupolar constant decreases to 176 kHz with η∼0. The broadening of the spectrum and the decrease of the quadrupolar constant indicate that deuterons become involved in some local motion by fast low-amplitude liberation of OD bonds and/or their isotropically reorientation with a characteristic time τC≤2π Q−1≈ 5×10−6 s.13 Further heating of TPA results in a notable transformation of the 2H NMR spectrum, which is related to the particular dynamics of deuterons on the surface of the Keggin anion (Figure 2). Visually, the spectrum represents a superposition of two signals; one with an anisotropic line shape while the second one is characterized by an isotropic pattern. The anisotropic line shape prevails at 433–463 K, indicating the involvement of deuterons in some anisotropic low-symmetry motion. In contrast, at 503 K the isotropic pattern is prevailing with a Lorentzian line shape,12b, 14 showing that the deuterons are involved in an isotropic movement, presumably diffusion over the surface of the Keggin anion. Temperature dependence of experimental (left) and simulated (right) 2H NMR spectra of deuterated anhydrous TPA. Effective contribution of a) diffusing deuterons, b) localized flipping deuterons, and c) total simulated line shape (See alsoFigure S1). The observed anisotropic line shape corresponds best to deuteron exchange between two equivalent positions (two-site flipping), with an angle 2θ of 126° between the directions of the OD bond in each of the positions.15 This anisotropic two-site flipping motion reflects the local dynamics of the deuterons at the surface oxygens of the Keggin anion, which could be assigned to the exchange of a deuteron between the two possible orientations of the OD bond at a bridged oxygen of the Keggin anion. Alternatively, this motion could correspond to the exchange of a deuteron between its possible positions on a bridged and a terminal oxygen of the Keggin anion. In both cases, the angle 2θ between the OD bond directions at two adjacent adsorption sites should be 126°. This follows from the simulated line shape for a two-site exchange, which fits perfectly the experimental one, provided that the angle 2θ=126°. Taking into account that the angle between the two possible orientations of the OD bond at a bridged oxygen should be close to a tetrahedral one,16 we conclude that the detected anisotropic motion corresponds to the deuteron exchange between the two possible orientations of a OD bond of a bridged oxygen in the Keggin anion (Figure 3 a). Different deuteron motions occurring on the Keggin anion of TPA. a) Anisotropic two-site flipping of the OD bond of a deuteron localized on a single bridging oxygen. b) Migration of deuterons over the Keggin anion by hopping between neighboring surface oxygens. The isotropic motion could be represented by deuterons migrating over the whole surface of the Keggin anion by hopping between neighboring oxygen atoms3c (Figure 3 b). As a first approximation, the Keggin unit can be considered as a spherical surface and deuterons migrating from one oxygen to another would cover the whole sphere leading to a real isotropic spectrum.17 However, the observed transformation of the line shape (see the Supporting Information, Figure S1) did not follow the expected pattern for a true isotropic rotation at 423–453 K, but rather exhibited the pattern typical for a pseudo-isotropic motional mechanism.18 That is, the reorientation is represented by a jump-exchange process of high symmetry such as the exchange over the vertices of a tetrahedron or octahedron. This finding indicates that the symmetry of the OD bond migration over the Keggin anion is high enough to average the broad quadrupolar pattern by deuteron jumps over the surface oxygens of the Keggin anion. This isotropic diffusional motion may be realized by different pathways, but evidently, all surface oxygens of the Keggin anion can assist the motion. The two different motional modes, anisotropic and isotropic coexist simultaneously in solid TPA. At any given temperature, each mode is characterized by the rate of the motion and the relative population factor. An increase in temperature results in the increase of the rates of both motions. The population of isotropically moving deuterons increases whereas that of the anisotropic two-site flipping motion decreases as the temperature rises. The temperature dependences of the two-site fliping (rate constant kf) and the isotropic reorientation by diffusion over the Keggin anion (rate constant kD) show that both motions follow a normal Arrhenius law (Figure 4). The activation barrier for the localized two-site flips is Ef=72±5 kJ mol−1 with a pre-exponential factor of kf0=(3.2±0.4)×1013 s−1; the isotropic deuteron migration is characterized by an activation barrier of ED=78±6 kJ mol−1 with a pre-exponential factor of kD0=(1.2±0.6)×1013 s−1. Arrhenius plots for the rate constants for deuteron two-site flipping, kf (□) and isotropic diffusion, kD (○). The temperature dependence of the relative population factor for each motion shows (Figure 5) that the population of the deuterons in the hopping state is always dominating the population in the flipping state. While at 423 K the two populations are comparable, virtually all deuterons are involved into diffusion over the Keggin anions above 503 K. Relative population of the motional modes of the TPA surface deuterons: two-site flipping (•), isotropic diffusion over the Keggin anion (□). Such observation evidences that in anhydrous TPA deuterons exhibit motional modes with rates higher than 1 kHz (τC<10−3 s) already at 423 K. Moreover, at the temperatures of chemical reactions mediated by TPA (T≥373 K) the major fraction of deuterons is mobile and migrating (τC<10−4 s) over the Keggin-anion surface. This means that the acidic protons are not localized at certain terminal or bridging surface oxygens at the respective reaction temperatures, that is, the acidic protons can be regarded as being delocalized on the time scales of the catalytic reactions.9 In summary, the 2H NMR method allowed us to directly evidence and quantify the mobility of the acidic protons (deuterons, in our case) in anhydrous 12-tungstophosphorous acid. We were able to follow the dynamics of protons up to temperatures where chemical reactions mediated by solid TPA occur. There are two types of surface protons, the ones rapidly flipping, being localized on the bridged oxygen sites, and the ones involved into relatively fast migration over the surface oxygens of the Keggin anion. While below 423 K the population of fast migrating protons is low, almost all surface protons are rapidly diffusing over the Keggin-anion surface at T>503 K by hopping between neighboring oxygens. Deuterated 12-tungstophosphoric acid (TPA), D3PW12O40, was prepared by recrystallization from a D2O solution in a dessicator under argon atmosphere. The anhydrous TPA sample for NMR studies was prepared by thermal treatment under vacuum at 430 K for 21 h. Gravimetric analysis showed <0.1 molecules of crystallization water per TPA Keggin unit in the sample. 2H NMR experiments were performed at 61.432 MHz on a Bruker Avance-400 spectrometer, using a high power probe with 5 mm horizontal solenoid coil. The solid-state 2H NMR spectra were acquired using the Exorcycled quadrupole-echo sequence (90X-τ1-90ϕ-τ2-acq-t),19 where τ1=20 ms, τ2=22 ms, and t is the repetition time for the sequence during accumulation of the NMR signal. The duration of the 90° pulses was 1.9–2.1 ms. The typical acquisition time required for the accumulation of a spectrum at each studied temperature was 60–70 h. The NMR measurements were performed over a broad temperature range from 298–503 K. Detailed descriptions of the approaches simulate the experimental 2H NMR spectra are provided in the Supporting Information. The authors would like to thank Dr. Nadine Essayem for helpful discussions. This work was supported by the Russian Foundation for Basic Research (grant no. 12–03–31108). As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.