Selective Oxidation by H5[PV2Mo10O40] in a Highly Acidic Medium

Dissolution of the polyoxometalate (POM) cluster anion H5[PV2Mo10O40] (1; a mixture of positional isomers) in 50% aq H2SO4 dramatically enhances its ability to oxidize methylarenes, while fully retaining the high selectivities typical of this versatile oxidant. To better understand this impressive reactivity, we now provide new information regarding the nature of 1 (115 mM) in 50% (9.4 M) H2SO4. Data from 51V NMR spectroscopy and cyclic voltammetry reveal that as the volume of H2SO4 in water is incrementally increased to 50%, V(V) ions are stoichiometrically released from 1, generating two reactive pervanadyl, VO2+, ions, each with a one-electron reduction potential of ca. 0.95 V (versus Ag/AgCl), compared to 0.46 V for 1 in 1.0 M aq H2SO4. Phosphorus-31 NMR spectra obtained in parallel reveal the presence of PO43–, which at 50% H2SO4 accounts for all the P(V) initially present in 1. Addition of (NH4)2SO4 leads to the formation of crystalline [NH4]6[Mo2O5(SO4)4] (34% yield based on Mo), whose structure (from single-crystal X-ray diffraction) features a corner-shared, permolybdenyl [Mo2O5]2+ core, conceptually derived by acid condensation of two MoO3 moieties. While 1 in 50% aq H2SO4 oxidizes p-xylene to p-methylbenzaldehyde with conversion and selectivity both greater than 90%, reaction with VO2+ alone gives the same high conversion, but at a significantly lower selectivity. Importantly, selectivity is fully restored by adding [NH4]6[Mo2O5(SO4)4], suggesting a central role for Mo(VI) in attenuating the (generally) poor selectivity achievable using VO2+ alone. Finally, 31P and 51V NMR spectra show that intact 1 is fully restored upon dilution to 1 M H2SO4.


■ INTRODUCTION
T h e m o l y b d o v a n a d o p h o s p h a t e c l u s t e r a n i o n , H 5 [PV 2 Mo 10 O 40 ] (1), is a versatile catalyst for selective aerobic oxidations of organic and inorganic compounds via a range of mechanisms, from electron transfer (ET) to electrontransfer induced oxygen transfer (ET-OT), a homogeneous liquid-phase analogue of Mars−van Krevelen type reactions. 1−4 Over the past three decades, numerous advances have been achieved using 1 in a range of solvents, including water, MeCN, and toluene, with reduced forms of 1 reoxidized by air/O 2 . 3,5,6 In general, high selectivities have been attributed at least in part to the stabilization or ET sequestration of organicradical intermediates by 1 itself, and during reoxidation, the absence of reactive intermediates from the partial reduction of O 2 that, if generated as they are in numerous aerobic oxidations, would give rise to nonselective radical-chain processes.
When functioning as an outer-sphere oxidant, the activity of 1 is controlled by its reduction potential, which at 0.4−0.45 V versus the saturated calomel electrode (SCE) has historically restricted its ET activity to transformations of more readily oxidizable substrates. 1,2 This situation changed dramatically upon discovery that in 80% aq H 2 SO 4 the reduction potential of 1 increased to 1.1−1.2 V, facilitating its use in transforming carbohydrates to synthesis gas. 7 In 50% aq H 2 SO 4 , 1 (92 mM) is capable of oxidizing benzene, facilitating its aerobic oxidation to phenol, 8 and 115 mM solutions of 1 selectively convert methylarenes to benzaldehyde derivatives. 9 While the protonation of intact 1 in 50% (9.4 M) aq H 2 SO 4 could readily account for the more positive reduction potential observed in that solvent system, 8,9 the retention of high selectivity further suggested that the cluster anion remained otherwise largely intact and thus able to sequester electrons from radical-organic intermediates. Consistent with this, after the reaction, electrochemical reoxidation, and dilution in D 2 O, 31 P NMR spectra invariably revealed the signature set of signals associated with the fi ve positional isomers of H 5 PV 2 Mo 10 O 40 . 7−9 To better understand its reactivity and selectivity, the nature of 1 (115 mM) in 50% aq H 2 SO 4 has now been systematically investigated by 51 V and 31 P NMR spectroscopy, ESR spectroscopy, cyclic voltammetry (CV), and single-crystal Xray crystallography and by a series of reactions involving oxidation of a model substrate, p-xylene, to p-methylbenzaldehyde. The herein reported data, which provide an entirely new picture of 1 in 50% aq H 2 SO 4 , reveal the reversible formation of a selective oxidative system composed of free VO 2 + (for enhanced reactivity), with high selectivity attributed to the coformation of an oxo-bridged (corner-sharing) dimeric permolybdenyl cation, [Mo 2 O 5 ] 2+ (Figure 1). Control experiments indicate that free phosphate (also observed) plays a minimal role. As previously observed, 7−9 dilution of this thermodynamically stable oxidative system leads quantitatively to hydrolytic self-assembly of the component species into fully intact 1.

■ RESULTS AND DISCUSSION
To investigate the solution-state chemistry of 1 in 50% aq H 2 SO 4 , the first set of experiments involved the use of 51 V NMR spectroscopy, in combination with cyclic voltammetry, to help explain the previously observed 9 ca. 0.5 V increase in reduction potential.
For this, 51 V NMR spectra of 1 (115 mM) were obtained for solutions containing incrementally larger concentrations of H 2 SO 4 . The use of 115 mM 1 was critically important not only because this concentration was used in previously reported reactivity studies 9 but also because of the speciation chemistry of metal cations, including those that form metal-oxide cluster anions (POMs) is a function of concentration. 10,11 As such, smaller concentrations of 1 would not provide definitive information about the reactive system under investigation.
Release of Pervanadyl Ions (VO 2 + ). The pH of 115 mM 1 in pure water is 0.06. As the concentration of H 2 SO 4 , [H 2 SO 4 ], was incrementally increased from 0 to 5 M, the set of 51 V NMR signals characteristic of the five positional isomers of 1, observed between −510 and −525 ppm, decreased in intensity and were entirely replaced by a broad signal at ca. −550 ppm ( Figure 2). This signal shifted to more negative ppm values as [H 2 SO 4 ] was increased to 9.4 M. Notably, very little change was observed when an additional 0.25 equiv (28.75 mM) of NaVO 3 was added (topmost plot in Figure 2). Moreover, an identical solvent system containing 230 mM NaVO 3 gave a 51 V NMR spectrum nearly identical to that obtained upon dissolution of 1 ( Figure S1). The sharp signal at −539 ppm, indicated by an asterisk in Figure 2, is due to the external reference, K 4 [PVW 11 O 40 ], present in a coaxial NMR tube.
The nearly identical spectrum of NaVO 3 suggested that as [H 2 SO 4 ] increased, V(V) ions were released from 1 to give pervanadyl ions, VO 2 + . Although documented for much smaller concentrations of 1 in 1−3 M acid, the partial release of V(V) from 1 has been observed in equilibrium with a proposed V-depleted anion, [PVMo 10 O 39 ] 8− . 12 In the present case, using 9.4 M H 2 SO 4 , the 51 V NMR data suggested a more extensive loss of V from 1. However, due to the broad signals associated with the quadrupolar 51 V nucleus (I = 7/2), and the possibility of dynamic processes (such as reversible protonation) leading to signal broadening, no definitive conclusions could be reached without independently quantifying the extent of V release. This was done by the cyclic voltammetric analysis of the same set of solutions as shown in Figure 2, after which the number of equivalents of VO 2 + released was quantified by amperometric titration.
First, cyclic voltammograms (CVs) were obtained as a function of [H 2 SO 4 ] ( Figure 3A). These revealed a large positive shift in reduction potential as [H 2 SO 4 ] increased from 3 to 4 M. (Between these concentrations, broad voltammograms were obtained; 9 see Figure S2.) Notably, this shift in reduction potential corresponds nicely with the [H 2 SO 4 ] values at which large changes were observed by 51 V NMR spectroscopy ( Figure 2). In the inset to Figure 3A, CVs obtained at 0 and 9.4 M H 2 SO 4 are shown as gray and red plots, respectively.
For comparison, CVs were also obtained as a function of [H 2 SO 4 ] for NaVO 3 alone ( Figure 3B). A similar positive shift in reduction-potential values was observed. In the absence of simultaneous speciation chemistry of 1, a smoother shift to more positive potentials was observed. In both cases, however, the magnitudes of the positive shifts and the final potentials     Figure 3B) was electrochemically reversible, while quasi-reversible behavior was observed for V from 1, pointing to some interactions between released VO 2 + and other products of the speciation of 1.
Amperometric titration was then used to quantify the numbers of equivalents of VO 2 + released from 1 upon dissolution in 9.4 M H 2 SO 4 . For this, 1 (115 mM) was dissolved in the acidic medium and the cathodic-current maximum was used as a starting point for the titration (blue curve in Figure 4A). Next (also in Figure 4A) CVs were recorded after incremental additions of NaVO 3 , up to 230 mM added VO 3 − (i.e., identical to the total concentration of V(V) in 115 mM 1). As VO 3 − was added, the cathodic current relative to that for 1 alone doubled from unity (for 1 alone) to two, as the "relative concentration" of free V(V) increased from 230 mM (from 1 alone) to a total of 460 mM ( Figure  4B). This doubling of current indicated that upon dissolution of 115 mM 1 in 9.4 M H 2 SO 4 , both V(V) atoms are released to form VO 2 + . To further confirm the fidelity of this method, a similar experiment was carried out starting with a solution of 230 mM VO 3 − ( Figure 4C,D). An identical doubling of the relative current ( Figure 4D) confirmed that no unexpected concentration-dependent behavior was responsible for the results shown in Figure 4A,B.
Release of Phosphate. The solution-state behavior of the phosphate heteroatom in 1 as a function of [H 2 SO 4 ] was then investigated by 31 P NMR spectroscopy ( Figure 5). As was observed in 51 V NMR spectra (Figure 2), the intensity of the characteristic set of signals associated with positional isomers of 1 decreased with [H 2 SO 4 ], while the intensity of a new (broad) signal closer to 0 ppm grew correspondingly. Signals in this region, near 0 ppm, are typical for solutions of phosphate. At the largest [H 2 SO 4 ] values, three broad signals were observed. Integration using a coaxial tube containing Na 4 P 2 O 7 as an external standard (sharp signal at −5.8 ppm in Figure 5) indicated a quantitative release of phosphate.
The intensity of the broad signals increased when an additional 0.25 equiv (28.75 mM) of H 3 PO 4 was added, but no new signals were observed (see the two topmost spectra in Figure 5). Given that 31 P is a spin 1 / 2 nucleus (giving narrowline width signals), and in light of the absence of paramagnetic species, the broadness of the 31 P NMR signals suggested the presence of a dynamic process involving relatively labile species.
This was explored further by preparing 9.4 M H 2 SO 4 solutions of H 3 PO 4 alone and in combination with NaVO 3 (plots a and b, respectively, in Figure 6A). In both cases, relatively sharp 31 P NMR signals were observed near 0 ppm. When H 3 PO 4 was combined with 10 equiv of Na 2 MoO 4 , however, three broad signals, similar to those observed for 1, were observed (plots c and d, respectively, in Figure 6A).
This suggested that the broad 31 P NMR signals were due to a dynamic process involving both P and Mo. The dynamic nature of the system was confirmed by variable-temperature   Inorganic Chemistry pubs.acs.org/IC Article 31 P NMR spectroscopy ( Figure 6B). As the temperature of a solution of 1 in 9.4 M H 2 SO 4 was increased from 298 to 333 K, the three broad signals coalesced.
Reactivity and Selectivity. The above spectroscopic and electrochemical data pointed to the formation of a new set of species upon dissolution of 1 in 9.4 M H 2 SO 4 . Although the increase in reactivity relative to 1 in pure water or organic solvents could be explained by the quantitative formation of V V O 2 + , the pervanadyl cation is known to give only modest selectivities. This raised the intriguing question as to why the constellation of species present in 9.4 M H 2 SO 4 retained the high selectivity characteristic of ET-driven oxidations by 1.
This was addressed using the ET oxidation of an arylalkane to the corresponding benzaldehyde. 9 For this, p-xylene was chosen as a representative substrate. Its conversion to pmethylbenzaldehyde (eq 1) was investigated by comparing the reactivity and selectivity of 1 with solutions of its component species in 9.4 M H 2 SO 4 ( Figure 7).
The percent conversion (mol % relative to V) and selectivity (mol % p-methylbenzaldehyde relative to other products) for the reaction of 1 (115 mM in 9.4 M H 2 SO 4 ) is shown at the far left in Figure 7.
The high conversion with high selectivity is in line with previously reported data for this reaction. 9 The experiments then listed from left to right in Figure 7, all in 9.4 M H 2 SO 4 , were designed to systematically elucidate the species responsible for the high reactivity and selectivity. In the absence of V, i.e., for H 3 PO 4 alone (P), and in combination with 10 equiv of Na 2 MoO 4 (P + 10Mo), no reactivity was observed. For 2 equiv of NaVO 3 alone (2 V) or in combination with H 3 PO 4 (2 V + P), high conversions were achieved, but with compromised selectivity relative to 1. However, when Na 2 MoO 4 was added to NaVO 3 (2 V + 10Mo,) selectivity was fully restored, i.e., identical to that obtained using 1. The same results were obtained when H 3 PO 4 was included (2 V + 10Mo + P), indicating that the dynamic species observed by 31 P NMR for combinations of H 3 PO 4 and Na 2 MoO 4 ( Figure 6) were not essential to the retention of selectivity.
Upon reduction by p-xylene, the color of the solution changed from orange-red to greenish-blue. To further confirm the central role of VO 2 + as the electron acceptor, the reduced (greenish-blue) solution was analyzed by ESR spectroscopy. The results (Figures S4−S6, Table S1) are definitive for the formation of vanadyl ion, V IV Figure 4), indicated 1 and 3 equiv of V V O 2 + , respectively, were released from the mono-and trivanadium cluster anions (Figures S9 and S10). As expected, 9.4 M H 2 SO 4 solutions containing H 4 [PVMo 11 O 40 ] (230 mM; 2 equiv relative to 1 in Figure 7) or of H 6 [PV 3 Mo 9 O 40 ] (76.7 mM; two-thirds of an equivalent relative to 1), gave conversions of p-xylene effectively identical to that shown in Figure 7 for 1, and with equally good selectivities ( Figure S11). Notably, the first one-electron reduction potentials of H 4 [PVMo 11 O 40 ] and H 6 [PV 3 Mo 9 O 40 ] are, respectively, more and less positive compared with that of 1 ( Figure S12). Nevertheless, once the stoichiometry of V V O 2 + is taken into account, the three cluster anions display identical reactivities and selectivities.
Formation of [Mo 2 O 5 ] 2+ . The results in Figure 7 identify the presence of Mo(VI) as critical to the high selectivity observed for ET oxidations by 1 in 50% H 2 SO 4 . As such, evidence for the solution-state structure of molybdate in this medium was sought. As NMR spectra of Mo nuclei are relatively uninformative, an effort was made to obtain crystals of Mo-containing species from 9.
Notably, the corner-shared Mo 2 O 5 core of 2 is a feature of several reported molybdo-organic complexes. 14 17 A complete set of lengths and angles is given in Figure S13 and Tables S2−S4. At the same time, to our knowledge, the permolybdenyl core, [Mo 2 O 5 ] 2+ , of 2 has not been identified as a product of condensation in aqueous acid. In this context, 2 may be viewed as a new member in the natural progression from titanyl, [Ti IV (Figure 9; bottom) gave rise to three broad signals, very similar to those observed for 1 itself (topmost spectrum in Figure 9). This finding is consistent with the dissolution of 1 in 9.4 M H 2 SO 4 giving rise to a dynamic equilibrium involving interactions between phosphate and 2.
Nevertheless, the data in Figure 7 point to Mo-based species alone as sufficient for retaining the high selectivity of 1. In this regard, selectivity in ET oxidations by 1 is in many cases attributed to the rapid sequestration of organic radicals by additional equivalents of V(V) in the intact POM itself. In the present case, a similar selectivity-enhancing mechanism could involve the oxidation of radical intermediates by the [Mo 2 O 5 ] 2+ core of 2, alone, or with phosphate anions in rapid exchange with SO 4 2− ligands. The impressive reactivity of H 5 PV 2 Mo 10 Figure S14). While 2 alone does not oxidize p-xylene to p-methylbenzaldehyde ( Figure S11), the Mo-containing species are nevertheless sufficiently reactive to provide for high selectivity by sequestering electrons from organic-radical intermediates.
These species act in concert to provide high reactivity and selectivity ( Figure 10B,C). A 31 PNMR spectrum of the reduced solution is shown in Figure 10C. Importantly, ESR spectra ( Figures S4−S6, Table S1) confirm the reduced species

Inorganic Chemistry
pubs.acs.org/IC Article is the vanadyl ion, V IV O 2+ , whose spectrum is distinct from that of reduced 1 (i.e., H 6 PV IV V V Mo 10 O 40 ). 21 Electrochemical reoxidation (bulk electrolysis) gives a solution with the 31 P NMR spectrum shown in panel D, very similar to that acquired before p-xylene oxidation (panel B).
In polyoxometalate chemistry, as in aqueous speciation chemistry generally, the acid condensation that occurs upon dissolution of 1 in 50% H 2 SO 4 is reversible. Upon dilution and NaOH neutralization of the solution in panel D (115 mM 1 in 50% H 2 SO 4 ) to 40 mM (pH 0.8), hydrolysis-driven assembly (the reverse of eq 4) leads to the complete re-formation of 1, as shown by the 31 P NMR spectrum provided as an inset in the center of panel A. The shift in ppm values associated with the positional isomers of 1 is due to the presence of Na 2 SO 4 (3.2 M). 22 Unique Medium for Cluster-Anion Formation and Reactivity. Viewed from a general perspective of polyoxometalate synthesis, 50% aq H 2 SO 4 is a unique environment. POMs are typically formed by stoichiometric reactions of metalate anions (e.g., MO 4 2− , M = Mo(VI) or W(VI)), or their partially condensed cluster anion forms, with mineral acids in water. Once the ratio of H + to MO 4 2+ reaches a value of 2, insoluble metal oxides such as MoO 3 (H 2 O) (a hydrated form of MO 3 ) are obtained. This ratio of 2 (referred to in the early isopolytungstate and -molybdate literature as the Z value) 23 has generally defined the limits of condensed structures that span the speciation "space" between metalates (MO 4 2− ) and solid-state MO 3 .
In the present medium, not only is the acid concentration much larger than that typically used in POM synthesis, but also the medium itself is entirely different. Rather than (relatively) dilute mineral acids in water, 50% (9.4 M) aq H 2 SO 4 contains ca. three molecules of water for each molecule of H 2 SO 4 , giving a liquid medium that may be written as, "H 8 O 3 SO 4 " ( Figure S15, Table S5). The data provided here argue that the thermodynamically controlled speciation chemistry of POMcomponent cations in this medium is dramatically different from that in water. Notably, this unique medium makes it possible to exceed the Z values of from 0 to 2 that have traditionally defined cluster-anion formation. As shown in eq 4, 1 reacts with H + in 50% H 2 SO 4 to give 5 equiv of [Mo 2 O 5 ] 2+ , representing an overall Z value of 3 with respect to MoO 4 2− . (Similarly, pervanadyl, V V O 2 + , may be viewed as the product of reacting 2 equiv of H + with insoluble V 2 O 5 .) This access to larger Z values, and the larger degree of condensation inherent to the formation of [Mo 2 O 5 ] 2+ , is also due to the relative solubility and lability of Mo(VI). This contrasts with the W(VI) analogue of 1, H 5 [PV 2 W 10 O 40 ], which upon mixing with 50% H 2 SO 4 gave insoluble WO 3 (H 2 O) as a yellow precipitate (which did not readily react with additional H + ).
It should be noted as well that 1 in 50% aq H 2 SO 4 is quite different from 1 present as a component of the "etherate" phases used synthetically to isolate this and other cluster anions. 24 Those phases are obtained after POM formation via acid condensation in water, by adding an approximately equal volume of diethyl ether, followed by acid (HCl or H 2 SO 4 ) to total concentrations of ca. 3 M. After mixing, the etherate phase forms as a highly dense (POM-rich) bottom layer. As part of the present work, that etherate phase was analyzed by 31 P NMR spectroscopy and CV (Figures S16 and S17) and, unlike in 50% aq H 2 SO 4 , intact 1 was the overwhelmingly dominant species present. This observation suggests that, unlike 1 in "H 8 O 3 SO 4 ", highly concentrated POM anions in protonated diethyl ether [H 7 C 2 O] + may be viewed as hybrid inorganic/organic ionic liquids. 25−27 Finally, as is true for speciation chemistry generally, the highly condensed species present in 50% aq H 2 SO 4 are formed under thermodynamic control. This not only provides for the high stability of the accessed oxidative system over cycles of reduction and reoxidation 9 but, as shown in Figure 10, also allows for facile hydrolytic assembly of intact 1 upon dilution to the less acidic aqueous environment historically more typical of polyoxometalate formation and reactivity.

■ CONCLUSIONS
Data provided here demonstrate that the unique medium provided by 50% aq H 2 SO 4 , in combination with the specific solution-state chemistry of Mo(VI), provides for the formation of more highly condensed species than typically encountered in POM cluster science. In the present case, the dissolution of (2) in 34% yield upon addition of (NH 4 ) 2 SO 4 ( Figure 8). Data from 31 P NMR spectroscopy additionally show that phosphate is released ( Figure 5) and interacts in dynamic equilibrium with [Mo 2 O 5 ] 2+ and possibly additional related Mo(VI) species (Figures 6 and 9).
Reactivity studies using a model reaction, the electrontransfer oxidation of p-xylene to p-methylbenzaldehyde ( Finally, the acid condensation that generates V V O 2 + and [Mo 2 O 5 ] 2+ -core complexes upon dissolution of 1 in 50% aq H 2 SO 4 is fully reversible. Upon dilution of the solution of 1 (115 mM) in 50% aq H 2 SO 4 and adjustment to pH 0.8, hydrolytically driven assembly processes (the reverse of eq 4) lead to complete re-formation of intact 1 (Figure 10).
Cyclic Voltammetry, Amperometric Titration, and Electrolysis. Cyclic voltammetry (CV) of the POMs was carried out in a three-electrodes cell setup on a CHI 760C potentiostat at 25 ± 2°C in 0.1 M Na 2 SO 4 electrolyte solutions, using a 2 mm glassy-carbon, Pt-wire, and Ag/AgCl (3 M KCl) working, counter, and reference electrodes, respectively. The scan rate was 50 mV s −1 .
Amperometric titrations were performed using cyclic voltammetry. 1 (115 mM) was dissolved in 9.4 M H 2 SO 4 in an electrochemical cell and after its CV was recorded, increasing amounts of NaVO 3 (up to 230 mM) were added and cathodic-current maxima were recorded for the V V /V IV redox couple. Finally, the relative current values were plotted against the total relative concentration of V(V). A control experiment carried out starting with 230 mM NaVO 3 in place of 115 mM 1 gave effectively identical results.
Electrolysis in a two half-cell configurations was carried out in the presence of Pt gauze as a working electrode and Pt wire as counter and reference electrodes at 1.3 V between the anode and cathode. The two half-cells were separated by a Nafion 212 membrane washed before use with 5 wt % H 2 O 2 and 8% H 2 SO 4 consecutively. Prior to use, the glassy carbon working electrode was polished with 0.3 μm α-Al 2 O 3 (Buehler), washed thoroughly with deionized water, and then exposed to ultrasound for approximately 5 min.
Substrate Oxidations. Typically, oxidations of p-xylene were carried out in 50 mL pressure tubes under N 2 . The products were extracted by dichloroethane and diluted for analysis by gas chromatography (GC). The products were quantified using Thermo Scientific Focus Gas Chromatograph furnished with a dedicated flame-ionization detector (FID), and a 5% phenylmethyl silicone 0.32 mm internal diameter, 0.25 mm coating, 7 m column (Restek 5MS), using helium as the carrier gas.
Single-Crystal X-ray Measurements. A colorless block-shaped crystal (0.087 × 0.155 × 0.161 mm 3 ) of [NH 4 ] 6 [Mo 2 O 5 (SO 4 ) 4 ] was mounted on a CrystalCap ALS HT cryo-loop mount for data collection on a Rigaku XtalLAB Synergy-S single-crystal X-ray diffractometer, which includes a Hy-Pix-6000HE detector and a standard Cu Kα X-ray radiation source (λ = 1.54184 Å). Unit cell dimensions, space group assignment, data reduction, and finalization were done by using the CrysAlisPRO software package 28 (ver. 39.49, released 2018). A total of 50845 reflections were collected, of which 4578 were used after merging by SHELXL 29 according to the crystal class and on the basis of Friedel pair equivalency for structure solution. Analytical numeric absorption correction was done using a multifaceted crystal model, 30 and empirical absorption correction was done using spherical harmonics. 31 The structure was solved in the orthorhombic C222 1 space group (no. 20) by SHELXT 32 via intrinsic phasing and refined by SHELXL using a full-matrix least-squares technique.
The final refinement cycle included the atomic coordinates and anisotropic thermal parameters of all atoms (not including hydrogen atoms), which converged toward R1 = 0.0387, wR2 = 0.1053, and S = 1.021. All non-hydrogen atoms were located; hydrogen atoms were not assigned due to structural disorder. Full details of crystal data are listed in the Supporting Information (Table S2−S4).
Frozen-State Electron Spin Paramagnetic Resonance (ESR) Measurements. ESR measurements were carried out in a frozen state (T ≤ 100 ± 0.1 K) using a Bruker EMX220 X-band (ν ∼ 9.4 GHz) spectrometer equipped with an Oxford Instrument ESR900 cryostat and an Agilent 53150A frequency counter. Spectra processing, determination of parameters, and spectral simulations were done using the Bruker WIN-EPR/SimFonia package and OriginLab software.