Catalytic Methane Monofunctionalization by an Electrogenerated High-Valent Pd Intermediate

Abstract

Electrophilic high-valent metal ions are potent intermediates for the catalytic functionalization of methane, but in many cases, their high redox potentials make these intermediates difficult or impossible to access using mild stoichiometric oxidants derived from O₂. Herein, we establish electrochemical oxidation as a versatile new strategy for accessing high-valent methane monofunctionalization catalysts. We provide evidence for the electrochemical oxidation of simple PdSO₄ in concentrated sulfuric acid electrolytes to generate a putative Pd₂^III,III species in an all-oxidic ligand field. This electrogenerated high-valent Pd complex rapidly activates methane with a low barrier of 25.9 (±2.6) kcal/mol, generating methanol precursors methyl bisulfate (CH₃OSO₃H) and methanesulfonic acid (CH₃SO₃H) via concurrent faradaic and nonfaradaic reaction pathways. This work enables new electrochemical approaches for promoting rapid methane monofunctionalization.

Short abstract

Electrochemical oxidation of Pd^II in sulfuric acid generates a dinuclear Pd^III species that rapidly converts methane to CH₃OSO₃H and CH₃SO₃H and supports continuous catalyst regeneration.

Introduction

Methane is an abundant, low-cost, carbon-based feedstock, but its chemical inertness and propensity for uncontrolled oxidation impedes its widespread utilization as a precursor to liquid fuels and commodity chemicals.¹⁻⁴ Existing technologies for the conversion of methane to liquid products proceed via a two-step indirect route that involves a high-temperature, capital-intensive process.⁵ An alternative to this scenario is the use of homogeneous catalysts that employ transition metal and main group ions to activate the localized, low-energy filled orbitals within the C–H bonds of methane at milder temperatures.⁶ In concentrated CF₃CO₂H and H₂SO₄ media, electrophilic ions mediate the two-electron oxidation of methane to methyl esters, which are protected from overoxidation by the electron-withdrawing effect of CF₃CO₂– and HSO₄– groups.⁷ These methyl esters can be subsequently hydrolyzed to produce methanol. In these schemes, catalytic methane functionalization proceeds via two-electron redox cycling of the metal center. Apart from Pt-based catalysts that are known to activate the methane C–H bond in the low-valent Pt(II) state,⁸⁻¹¹ most catalysts activate methane in their high-valent form and generate the methanol product following reductive elimination to a lower valent species (Figure 1, left).^12,13 Irrespective of the mechanistic details, the two-electron redox potential of the metal center crudely approximates the driving force for oxidative methane functionalization (Figure 1, right, top arrow), and, invoking a linear free-energy correlation, metal ions with higher redox potentials are expected to functionalize methane more rapidly. Herein lies a central challenge for catalyst design: increasing the redox potential of the metal center accelerates methane functionalization¹⁴ but simultaneously impedes reoxidation by the stoichiometric oxidant in the system (Figure 1, right, bottom arrow). As O₂ is the only viable terminal oxidant for large-scale methane functionalization, the redox potential of the metal catalyst cannot significantly exceed the O₂/H₂O redox potential, E° = 1.23 V. This constraint relegates metal centers that possess two-electron redox potentials positive of the O₂/H₂O couple, such as Tl^III, Pb^IV,¹⁵ and Au^III,¹⁶ to principally stoichiometric methane functionalization reactivity. Additionally, as O₂ is too sluggish and unselective to be used directly, the SO₃/SO₂ couple is commonly employed.¹³ This oxidant is attractive because it can be regenerated via aerobic combustion over a V₂O₅ catalyst,¹⁷ but the SO₃/SO₂ couple possesses an even lower redox potential, E° = 0.86 V (Tables S14 and S15), further reducing the driving force for catalyst reoxidation. This low driving force, combined with the kinetic sluggishness of S–O bond cleavage,^18,19 makes catalyst reoxidation rate-limiting in nearly all methane functionalization cycles that employ SO₃.^18,20 Clearly, practical methane functionalization schemes require the development of alternative approaches that provide for rapid catalyst reoxidation with a tunable driving force.

Figure 1.

Redox potentials of methane oxidation catalysts. (Left) Simplified catalytic cycle for electrophilic methane oxidation using stoichiometric oxidant YO. (Right) Estimated redox potentials of electrophilic methane functionalization catalysts/reagents²¹ compared to the values for the CH₄/CH₃OH, O₂/H₂O, and SO₃/SO₂ redox couples. Derivation of SO₃/SO₂ and CH₄/CH₃OH redox potentials is detailed in Tables S14 and S15.

In principle, electrochemical methods for regenerating electrophilic high-valent catalysts could overcome this central reoxidation challenge. By varying the electrode potential, the driving force for reoxidation can be systemically controlled to maintain a nonequilibrium population of highly reactive high-valent species, even if their redox potentials are more positive than that of viable terminal oxidants such as O₂ and SO₃. By decoupling the metal redox potential from that of the terminal chemical oxidant (e.g., O₂), we envisioned that electrochemical methods can access high potential metal ions that would be capable of functionalizing methane at exceptional rates. Thus, continuous electrogeneration of these high-valent species in a concentrated acid medium could, in principle, enable new catalytic and electrocatalytic cycles for methane functionalization. The electrons extracted to maintain this pool of high-valent ions could be used to drive the reduction of O₂ to water at a separate electrode, allowing for continuous electrochemical methane functionalization. Previous efforts toward this goal were hampered by significant overoxidation on heterogeneous electrocatalysts, leading to <5% methanol selectivity on nickel oxide based systems.^22,23 In addition, the development of molecular electrocatalysts was impeded by the sluggish interfacial electron transfer kinetics typical of Pt-group ions.²⁴⁻²⁶ The latter prevented direct electrochemical reoxidation on a timescale commensurate with methane functionalization.²⁷ Herein, we establish an electrochemical strategy for selective catalytic methane functionalization that employs electro-oxidation of Pd^IISO₄ in concentrated sulfuric acid to generate a putative Pd₂^III,III intermediate. This species rapidly reacts with methane to generate precursors to methanol—CH₃OSO₃H and CH₃SO₃H—via concurrent electrocatalytic and nonfaradaic pathways, respectively.

Results and Discussion

Shown in Figure 2A is the cyclic voltammogram (CV) of PdSO₄ in concentrated, 95–98%, sulfuric acid electrolyte. Autoionization of the sulfuric acid removes the need for added electrolytes.²⁸ Fluorine-doped tin oxide (FTO) coated glass slides²⁹ served as corrosion-resistant working electrodes (Figure S1) under these conditions (see Materials and Methods). Background double layer charging current is observed between 0.42 and 1.46 V (Figure 2A, black; all potentials are reported vs the Ag₂SO₄/Ag (SSE) electrode³⁰⁻³²), whereas scanning to more positive potentials (Figure 2A, red) reveals an oxidative wave at E_p,a = 1.96 V prior to solvent oxidation beyond 2.40 V. On the return scan, there is a pronounced hysteresis in the CV, with the forward and backward traces crossing at 1.75 V. Scanning into this oxidative wave is both necessary and sufficient to generate a broad reduction wave at 0.76 V on the cathodic scan. The hysteretic crossing of the forward and reverse CV traces is diagnostic³³ of an overall two-electron oxidation sequence that proceeds via an initial electron transfer step (E₁), a chemical reaction step (C), and a subsequent second electron transfer step (E₂) that occurs at a lower redox potential than the initial oxidation process. Together, these steps are denoted as an E₁CE₂ process.³³ The hysteretic behavior arises from the large inversion in redox potentials between the E₁ and E₂ steps; during the forward sweep, the E₁ and C steps form an intermediate that is easier to oxidize than the starting material, giving rise to additional current on the reverse trace. Although, in general, an E₁CE₂ mechanism can give rise to a disproportionation reaction following the C step, a high rate of disproportionation serves to eliminate the hysteretic behavior³³ (Figure S8), indicating that its role in the Pd oxidation sequence is minimal. The high-valent product of the E₁CE₂ sequence is then back-reduced to Pd^II in the broad wave at 0.76 V. CV data collected at varying scan rates (Figure 2B) further support an E₁CE₂ oxidation mechanism. At scan rates greater than 200 mV/s, we observe another cathodic wave at E_p,c = 1.41 V, consistent with back-reduction of the mono-oxidized species, the reverse of E₁, occurring prior to the C step. Concomitantly, the hysteretic behavior centered at 1.75 V disappears at faster scan rates, consistent with this back-reduction beginning to outcompete the C step. Together the data establish that Pd^II solutions in concentrated sulfuric acid electrolyte undergo a two-electron oxidation via an E₁CE₂ sequence to generate a high-valent Pd species.

Figure 2.

Evidence for an electrogenerated Pd₂^III,III species. (A) CVs (50 mV/s scan rate) of PdSO₄ (∼25 mM) in concentrated H₂SO₄. Arrows indicate potential of scan initiation and direction of scan. (B) Varying scan rate CVs of PdSO₄ (∼24 mM) in concentrated H₂SO₄. Arrows indicate direction of scan and progression of waves at 1.41 and 1.75 V with increasing scan rate. (C) UV−vis spectra of PdSO₄ before (black) and after (red) electrolysis in concentrated H₂SO₄. (D) Proposed mononuclear and binuclear E₁CE₂ mechanisms. (E) Return scans of CVs (200 mV/s scan rate) recorded in four concentrations of PdSO₄ depicting the integrated charges, Q₁ and Q₂, of the back-reduction waves. (F) Simulated (red and blue) and experimental (black) percent integrated charge (Q₁ %) in the first back-reduction wave at E_p,c = 1.41 V vs the concentration of PdSO₄. Dotted lines serve as guides to the eye.

Spectroscopic data provide insight into the electronic structure of the electrogenerated species. The high-valent Pd species was generated by preparative bulk-electrolysis and displays a strong UV–vis absorption band (ε = 2.0 × 10⁴ M^–1 cm^–1; Figure S4B) centered at 300 nm (Figure 2C, red) that is absent in the Pd^II spectrum (Figure 2C, black). In situ UV–vis spectroelectrochemistry (Figure S3) reveals that this 300 nm band arises as a direct result of polarization beyond the oxidative wave. Notably, the high-valent species displays no detectable absorption bands at lower energies out to 1400 nm. Mononuclear, dinuclear, and polynuclear Pd complexes with unpaired spins display strong absorption bands (ε = 0.4–5.0 × 10³ M^–1 cm^–1) between 500 and 1000 nm;^34,35 thus the spectroscopic data obtained here suggest that the product generated via electro-oxidation is a ground-state singlet species. This postulate is supported by Evans method magnetic susceptibility measurements (Figure S5), which reveal that electro-oxidation generates a diamagnetic species at room temperature. Together, the spectroscopic data are consistent with either a mononuclear Pd^IV or a dinuclear Pd₂^III,III complex. Both of these species can be formed via E₁CE₂ mechanisms that are depicted in Figure 2D and abbreviated as Pd^II/Pd^III//*Pd^III/Pd^IV or Pd^II/Pd^III//Pd₂^III,II/Pd₂^III,III, where / and // denote E and C steps, respectively.

To distinguish between the foregoing mechanistic possibilities, we collected CV data at four Pd^II concentrations ranging from 0.34 to 6.8 mM (Figure 2E). The magnitude of the back-reduction wave at ∼1.4 V serves as a measure of the concentration of Pd^III at the electrode after scanning through the oxidative wave. We compared this result to the magnitude of the broad wave at ∼0.75 V, a measure of the concentration of the two-electron oxidized product, Pd₂^III,III or Pd^IV, generated at the electrode. As the Pd^II concentration increases, the magnitude of the wave at ∼1.4 V decreases relative to the magnitude of the wave at ∼0.75 V, indicating that the rate of the C step is enhanced by increasing [Pd^II]. The observed concentration dependence excludes a purely mononuclear E₁CE₂ mechanism, Pd^II/Pd^III//*Pd^III/Pd^IV. To further verify the mechanism, we simulated cyclic voltammograms (see Supporting Information for full simulation details) for both mononuclear (Figure S6) and dinuclear (Figure S7) mechanisms and found that the experimental percentage of charge passed in reducing the Pd^III intermediate, Q₁ % = 100 × Q₁/(Q₁ + Q₂) (Figure 2F, black squares), matches closely with the simulated Q₁ % for the dinuclear mechanism (Figure 2F, blue circles) and is in stark contrast to the roughly concentration-independent behavior observed in simulations of the mononuclear mechanism (Figure 2F, red triangles). Although these simulations exclude more complex mechanisms that might involve transient surface adsorption or oligomeric intermediates, the formation of an oligomeric product is inconsistent with the spectroscopic data detailed above.³⁵ Together these results suggest that electro-oxidation of Pd^II proceeds via an E₁CE₂ mechanism involving a dinuclear C step to generate a putative Pd₂^III,III species. We note that well-characterized Pd₂^III,III complexes are known to be key intermediates in C–H functionalization catalysis,^36,37 and our data suggest that similar species may be electrogenerated under these conditions. However, we acknowledge that the data presented here would also be consistent with a mixed-valent Pd₂^II,IV species, and investigations aimed at parsing between these valence tautomers and further characterizing the structure of the complex are currently ongoing.

The putative Pd₂^III,III species formed via electro-oxidation is highly active for electrocatalytic methane functionalization at elevated temperatures. At 140 °C, the CV scan of PdSO₄ in concentrated H₂SO₄ (Figure 3A, black) is similar to that recorded at ambient temperature (Figure 2A, red). As in the room temperature CV, a hysteretic anodic peak is observed at E_p,a = 1.82 V along with a broad reduction feature at E_p,c = 1.03 V, indicating that the same E₁CE₂ mechanism to generate the putative Pd₂^III,III species is operative at elevated temperatures. Upon charging the electrochemical cell with 100 psi of methane (Figure 3A, red), the anodic peak at 1.82 V transforms into a catalytic wave, the magnitude of which continues to rise with increasing methane pressure up to 500 psi (Figure 3A, violet). Consistent with electrocatalytic regeneration of the Pd^II via methane oxidation, the broad reduction peak at 1.03 V disappears completely in the presence of methane. Electrocatalytic methane oxidation is observed at lower temperatures as well. At 100 °C, increasing the methane pressure leads to a less pronounced rise in the anodic current beyond 1.5 V, but nonetheless leads to a diminution of the Pd₂^III,III back-reduction peak (Figure 3B), indicating consumption of the high-valent species via methane oxidation. Notably, there is a decline in the magnitude of this back-reduction wave at temperatures as low as 80 °C (Figure S9), revealing the potency of the putative Pd₂^III,III in carrying out methane oxidation catalysis.

Figure 3.

Methane functionalization by an electrogenerated Pd₂^III,III species. CVs (20 mV/s scan rate) of PdSO₄ (∼23 mM) in concentrated H₂SO₄ recorded at 140 °C (A) and 100 °C (B) with varying CH₄ pressure. (C) Double potential step chronoamperometry of PdSO₄ (∼23 mM) in concentrated H₂SO₄ recorded at 140 °C with varying CH₄ pressure and step potentials of 2.0 V followed by 0.50 V. (D) Plot of k_obs vs CH₄ pressure at 140 °C. (E) Arrhenius plot of electrochemical methane oxidation rate constants, k_cat (psi^–1 s^–1), between 80 and 140 °C. (F) ¹H NMR of the reaction mixture after treating a 4.2 mM Pd₂^III,III (black) and 8.4 mM Pd^II (red) solution in 20% SO₃/H₂SO₄ with 500 psi of CH₄ at 100 °C for 20 min.

To extract the rate constant for methane oxidation by the electrogenerated Pd₂^III,III species, we collected chronoamperograms at 2.0 V, well beyond the 1.82 V anodic peak potential corresponding to formation of this species. At 140 °C, in the absence of methane (Figure 3C), a diffusion-limited decay of the anodic current is observed, whereas, in the presence of methane, steady-state electrocatalysis is observed with anodic currents that are invariant with time. The high electrochemical driving force for generating the putative Pd₂^III,III and the time invariance of the chronoamperogram indicate that the current flow is limited, in this regime, by the activation-controlled rate of methane oxidation by the putative Pd₂^III,III species. Under pure kinetic conditions, the following relationship holds:³⁸

1

where t is time, j_[CH4] and j₀ are the current densities in the presence and absence of methane, respectively, and k_obs (s^–1) is the apparent rate constant for methane functionalization. Working curves were produced by plotting j_[CH4]/j₀ vs t^1/2 (Figure S10 and S11), and linear regions of these plots, which correspond to activation-controlled kinetics, were used to extract k_obs. At 140 °C and 500 psi of methane, the turnover frequency of electrocatalytic methane oxidation is 2000 h^–1. Under identical conditions, Pd^II(SO₄) catalyzes methane functionalization by itself, but with a turnover frequency of 0.4 h^–1, indicating that electrical polarization serves to increase the rate of methane functionalization by 5000-fold (Table S16). Indeed, the Pd₂^III,III species functionalizes methane at rates >20-fold faster than state-of-the-art nonelectrochemical catalysts, K₂PtCl₄ and (2,2′-bipyrimidyl)PtCl₂,^9,39 which, at the same temperature and methane pressure, display turnover frequencies of 90 h^–1 and 6 h^–1 (Table S16) respectively in fuming sulfuric acid. These comparisons illustrate the power of electrical polarization to drive methane functionalization at extremely high rates while maintaining modest reaction temperatures that may be more amenable to process integration.

Concentration-dependent studies were used to further characterize the mechanism of methane functionalization. We observed a first-order dependence of k_obs on methane concentration (Figure 3D) that indicates a bimolecular reaction between the putative Pd₂^III,III intermediate and methane prior to or during the rate-limiting step of catalysis. Slopes of these plots were used to extract bimolecular rate constants, k_cat (psi^–1 s^–1), for methane electro-oxidation, and an Arrhenius plot of these values over the 80–140 °C temperature range provides an activation energy barrier of E_a = 25.9 (±2.6) kcal/mol (Figure 3E), which is consistent with the very high rate of catalysis and is among the lowest experimental values reported for oxidative methane functionalization (Table S17). Additionally, the foregoing discussion highlights the power of electrochemical methods for extracting the kinetics of methane functionalization, data that have thus far been difficult to obtain using stoichiometric oxidants.

To identify the products of methane functionalization, we subjected a 4.2 mM solution of Pd₂^III,III in 20% SO₃/H₂SO₄, generated via preparative bulk electrolysis of Pd^II, to 500 psi of methane at 100 °C for 20 min. The added SO₃ suppressed the slow thermal decay of the Pd₂^III,III species via water oxidation. UV–vis spectroscopy of the sample following the reaction (Figure S12) establishes the quantitative regeneration of Pd^II over this time period, and the ¹H NMR spectrum of the reaction mixture reveals the formation of 0.5 (±0.1) equivalents of CH₃OSO₃H, indicating that, under these reaction conditions, all of the oxidizing equivalents in the Pd₂^III,III species go toward the two-electron oxidation of methane to a methanol derivative (Figure 3F). Because methane oxidation is a two-electron process, this reaction stoichiometry provides further evidence in support of an average Pd oxidation state of 3+ in the electrogenerated species. Remarkably, CH₃OSO₃H is not the only product of the reaction; we also observe 34 (±12) mM CH₃SO₃H, the product of the net redox-neutral insertion of SO₃ into the C–H bond of methane.⁴⁰ The same reaction, conducted in the presence of Pd^II, displays no methane functionalization products at this temperature, indicating that the electrogenerated Pd₂^III,III species is uniquely responsible for both carrying out oxidative methane functionalization to generate CH₃OSO₃H and catalyzing nonfaradaic methane sulfonation to generate CH₃SO₃H. In further support of this conclusion, we do not observe additional CH₃SO₃H if the reaction time is extended beyond the timescale of Pd₂^III,III consumption (Table S12). Given the ∼7-fold excess of CH₃SO₃H relative to CH₃OSO₃H that we observe under these conditions, these data indicate that the Pd₂^III,III catalyzes methane sulfonation at rates significantly higher than the rate of electrocatalytic methane oxidation measured electrochemically. Since CH₃SO₃H is known to convert to CH₃OSO₃H upon thermolysis, this sulfonation reaction pathway would not compromise net reaction selectivity in a well-engineered system.^41,42

In an optimally engineered electrochemical cell, continuous electroregeneration of the putative Pd₂^III,III species would permit methane sulfonation and electro-oxidation to proceed indefinitely. Bulk electrolysis of Pd^II at 70 °C in a stirred 20% SO₃/H₂SO₄ electrolyte in the presence of 500 psi of methane led to a sustained catalytic current of 0.65 mA/cm² that is unchanged over the course of 5 h (Figure S14). A relatively low temperature was chosen for this measurement to minimize convolution from a slow background reaction with Pd^II (see Supporting Information for details). ¹H NMR spectroscopic analysis of the reaction mixture following electrolysis reveals the formation of both CH₃OSO₃H and CH₃SO₃H in a ratio of ∼1:14. Remarkably, owing to simultaneous faradaic and nonfaradaic methane functionalization reactions, ∼3.4 molecules of methane are functionalized per electron passed, with a low ∼7% of the product subject to overoxidation to CO₂. The low electron stoichiometry of methane functionalization serves to improve the electrical energy efficiency of this system while maintaining extremely high rates of catalysis.

Concluding Remarks

These results demonstrate a simple strategy for selective methane monofunctionalization at modest temperatures via electrochemical oxidation of simple Pd^II salts in concentrated sulfuric acid solutions. Together, the data are consistent with a mechanistic model (Figure 4) for electroinduced methane functionalization catalysis under these conditions. Electrochemical oxidation of Pd^II proceeds via an E₁CE₂ sequence to generate a putative Pd₂^III,III intermediate, which rapidly functionalizes methane via concurrent faradaic and nonfaradaic pathways to generate CH₃OSO₃H and CH₃SO₃H, respectively. Both CH₃OSO₃H and CH₃SO₃H can be thermally and hydrolytically converted to methanol,^41,42 providing, in net, a viable electropromoted pathway for selective methane functionalization. The power of electrochemistry to generate and maintain a nonequilibrium population of highly reactive high-valent metal species opens the possibility of using this methodology to access a wider range of metal ion catalysts for the functionalization of diverse chemically inert substrates. As we demonstrate here, the electrogenerated high-valent species can carry out both faradaic and nonfaradaic reactions, reducing the electron stoichiometry of the process and enhancing its electrical efficiency. Combining this methodology with established methods for carrying out oxygen reduction catalysis provides a new electrochemical strategy for accelerating kinetically challenging aerobic C–H bond functionalization reactions.

Figure 4.

Proposed mechanism for electrochemical methane functionalization by a putative Pd₂^III,III intermediate. Green and blue arrows indicate faradaic and nonfaradaic reaction pathways, respectively.

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscentsci.7b00342.

• Materials and methods, simulations of electrochemical data, additional spectroscopic data, and original data for rate measurements (PDF)

The authors declare no competing financial interest.