Metal-Free Homogeneous O2 Reduction by an Iminium-Based Electrocatalyst

The oxygen reduction reaction (ORR) is important for alternative energy and industrial oxidation processes. Herein, an iminium-based organoelectrocatalyst (im+) for the ORR with trifluoroacetic acid as a proton source in acetonitrile solution under both electrochemical and spectrochemical conditions using decamethylferrocene as a chemical reductant is reported. Under spectrochemical conditions, H2O2 is the primary reaction product, while under electrochemical conditions H2O is produced. This difference in selectivity is attributed to the interception of the free superoxide intermediate under electrochemical conditions by the reduced catalyst, accessing an alternate inner-sphere pathway.


General Considerations
All chemicals and solvents (ACS or HPLC grade) were commercially available and used as received unless otherwise indicated.For all air-sensitive reactions and electrochemical experiments, HPLC-grade solvents were obtained as anhydrous and air-free from a PPT Glass Contour Solvent Purification System.Gas cylinders were obtained from Praxair (Ar as 5.0; O2 as 4.0) and passed through activated molecular sieves prior to use.Gas mixing for variable concentration experiments was accomplished using a gas proportioning rotameter from Omega Engineering.UV-vis absorbance spectra were obtained on a Cary 60 from Agilent using a quartz cuvette with 1 cm pathlength.The concentration of O2 saturation in MeCN is reported to be 8.1 mM and the saturation concentration in MeCN with added electrolyte to be 6.3 mM. 1 Flash column chromatography was performed using silica gel or alumina gel (230 -400 mesh) purchased from Fisher Scientific.Elution of compounds was monitored by UV. 1H and 13C NMR spectra were measured on a Varian Inova 600 (600 MHz) or Bruker Avance III 800 (800 MHz) spectrometer and acquired at 300 K.Chemical shifts are reported in parts per million (ppm δ) referenced to the residual 1H or 13C resonance of the solvent.The following abbreviations are used to indicate the multiplicity of signals: s -singlet, d -doublet, t -triplet, q -quartet, m -multiplet and br -broad.

Synthesis of im + Catalyst
Procedures taken and adapted from "Improved Parent Iminium Synthesis Procedure." 2 Under N2 atmosphere, benzyl cyanide (S7, 1 equiv) was added into a flame-dried round-bottom flask equipped with a stir bar.Tetrahydrofuran (0.4 M) was added and the suspension was cooled to 0 °C with stirring.60% sodium hydride (dispersion in paraffin liquid) (3 equiv) was added to the mixture and stirred for 1 hour.Iodomethane (3 equiv) was then added dropwise to the solution at 0°C and the solution was heated to 70 °C for 4 hours.Upon completion, the crude reaction mixture was cooled to room temperature then quenched with ice.The organic layer was extracted 3 times with ethyl acetate, and the combined organics were washed with brine, dried over anhydrous MgSO4, and concentrated in vacuo.The crude product was carried on to the next step.
Under N2 atmosphere, crude methylated benzyl cyanide (1 equiv) was dissolved in tetrahydrofuran (0.4 M) in a round bottom equipped with a stir bar and cooled to 0 °C with stirring.The solution was cooled to 0 °C and 2.4 M solution of lithium aluminum hydride in tetrahydrofuran (2.5 equiv) was added dropwise by addition funnel.The solution was heated to 70 °C and stirred for 4 hours.Upon completion, the crude reaction mixture was worked up following the Fieser Method: The white suspension was cooled to 0 o C and diluted to roughly twice its volume with diethyl ether."x" mL water was slowly added to the reaction mixture, where "x" is the amount of lithium aluminum hydride used for the reduction in grams."x" mL 15% aqueous sodium hydroxide was then added, followed by "3x" mL water.The mixture was then warmed to rt and stirred 15 minutes, followed by addition of anhydrous MgSO4.Upon stirring for an additional 15 minutes, the mixture was filtered over a pad of celite and concentrated in vacuo.The crude product was carried onto the next step.
Under N2 atmosphere, the crude amine (1 equiv) was dissolved in anhydrous dichloromethane (0.1 M) in a round-bottom flask equipped with a stir bar.Pyridine (1.2 equiv) was added with stirring, followed by dropwise addition of trifluoroacetic anhydride (1.1 equiv) via syringe.The reaction mixture was stirred at room temperature overnight (ca.16 h).Upon completion, the orange-brown solution was quenched with a brine wash, dried over MgSO4, and concentrated in vacuo.The residue was purified by flash chromatography with isocratic 20% ethyl acetate in hexanes to give acetamide S8 a white crystalline solid (32% yield over 3 steps).NMR spectra are consistent with literature reports. 3der N2 atmosphere, phosphorus pentoxide (1.5 equiv) was added to a 2-neck round-bottom flask equipped with a stir bar.Half the volume of phosphorus oxychloride (5 equiv) was added to the solution and heated to 70 °C, then immediately heated to 120 °C.The acetamide (S8, 1 equiv) was dissolved in the remaining half volume (5 equiv) of phosphorus oxychloride and added to the solution slowly.The solution turned brown and then black, and was further heated to 150 °C and allowed to stir for 5 hours.Upon completion, the mixture was cooled to room temperature, carefully diluted to twice the original volume with dichloromethane, and transferred to a large Erlenmeyer flask.In the flask, the brown-black solution was quenched slowly with excess water, then saturated aqueous sodium bicarbonate with stirring.The mixture was treated with base until it was light tan in color and the pH measured at 8 or greater.The resultant solution was extracted 3 times with dichloromethane and the combined organic layers were washed with brine, dried over MgSO4, and concentrated in vacuo.The crude residue was purified by flash chromatography with 0 -5% ethyl acetate/hexanes to give the cyclized imine as a yellow oil (S9, 63% yield).NMR spectra are consistent with literature reports. 3 an N2 glovebox, imine (S9, 1 equiv) was dissolved in anhydrous dichloromethane (0.5 M) in a vial equipped with a stir bar.Trimethyloxonium tetrafluoroborate (0.9 equiv) was added to the solution and stirred at room temperature overnight (ca.16 h) and removed from the glovebox.Solvent was removed in vacuo and the resulting solid was washed with anhydrous diethyl ether, then recrystallized from dichloromethane and diethyl ether to yield the iminium catalyst as a white crystalline solid (im + , 90% yield).NMR spectra are consistent with literature reports. 3

Electrochemical Analysis
All cyclic voltammetry experiments were performed using a Metrohm Autolab PGSTAT302N potentiostat.Glassy carbon working (⌀ = 3 mm) and non-aqueous silver/silver chloride pseudoreference electrodes behind PTFE tips were obtained from CH Instruments.The pseudoreference electrodes were obtained by depositing chloride on bare silver wire in 10% HCl at oxidizing potentials and stored in a 0.1 M tetrabutylammonium hexafluorophosphate solution in acetonitrile in the dark prior to use.The counter electrode was a glassy carbon rod (⌀ = 3 mm).All CV experiments were performed in a modified scintillation vial (20 mL volume) as a singlechamber cell with a cap modified with ports for all electrodes and a sparging needle.Tetrabutylammonium hexafluorophosphate (TBAPF6) was purified by recrystallization from ethanol and dried in a vacuum oven before being stored in a desiccator.All data were referenced to an internal ferrocene standard (ferrocenium/ferrocene reduction potential under stated conditions) unless otherwise specified.All voltammograms were corrected for internal resistance.Ferrocene was purified by sublimation prior to use.In the event that the presence of electrochemical features precluded ferrocene addition, ferrocene was added to the electrochemical cell at the end of analysis for reference.All CVs were scanned to negative potentials before sweeping to positive potentials.Rotating ring-disk electrode electroanalytical experiments were performed using a BioLogic VSP Bipotentiostat and a Pine Research MSR Rotator.Glassy carbon working electrode (⌀ = 5 mm) with a gold ring were obtained from Pine Research. A  0  G (*) -at 25 ºC. 4 For Figures S1 and S2.To ensure that species adsorbed to the electrode, a rinse test was performed with im + and TFAH (Figure S8).A CV was taken under catalytic conditions after which the electrode was removed and the sides were wiped and placed in a blank solution containing TFAH and a CV was taken.

Rotating Ring-Disk Electrode Methods
Description of Au Ring Roughening Procedure.The Au ring electrode was roughened according to a previously reported method. 5The electrodes were polished first on a felt polishing pad with 0.3 micron alumina, then with 0.05 micron alumina and rinsed with water and ethanol.Cyclic voltammograms were obtained in 0.5 M H2SO4 by scanning from 0 to 1.6 V vs. Ag/AgCl at 100 mV/s, then at 20 mV/s for an additional 2 cycles to obtain the pre-roughening, surface oxide reduction charge.The electrode was then pulsed between 2.4 and 0.2 V vs Ag/AgCl for 2.4 ms each and repeated for 250,000 cycles.Bubbles formed during electrolysis pulses were dislodged by contacting with a large bubble from a glass pipette.After electrolysis, the electrode was held at 0.3 V vs. Ag/AgCl for 2 minutes and the roughening was evaluated by CV.
Description of RRDE Collection Efficiency.7][8] Conditions: Ar saturation, 0.1 M TBAPF6, 0.5 mM ferrocene in MeCN, glassy carbon disk electrode (5 mm), roughened Au ring electrode, glassy carbon rod counter electrode, Ag/AgCl pseudoreference electrode; scan rate 0.1 V/s.To calculate the collection efficiency of the RRDE, the ratio of the ring current (ir) to the disk current (id) at each rotation rate was used to determine Nempirical (Eq S1).The Nempirical value at each rotation rate was multiplied by a factor of 100 to determine the collection efficiency % at each rotation rate (~15%).The solution was sparged until saturation was achieved.Im + (0.5 mM) was dissolved in solution and 0.1 M TFAH was added.A standard CV was taken of the solution to confirm the potential window to be used for the experiment.The roughened Au ring was set to +1.2 V. LSVs were obtained for various rotation rates between 400 and 2400 under the described conditions.In between each scan, the solution was sparged for 3 minutes.The reproducibility of scans was confirmed by repeating scans at the same rotation rate, producing exact overlays.The same procedure was repeated for air saturation conditions, which were achieved by sparging the solution with air for 3 minutes.Disk (id) and ring (ir) currents were corrected by subtracting the current observed under Ar to ensure that the current observed was a result of H2O2 formation.
The arithmetic mean of the number of electrons received by O2 (ncat) during the ORR was calculated from the disk current (id) and ring current (ir) according to Eq S2: The H2O2 ratio (p) is defined as the fraction of O2 reduced to H2O2 and relates to ncat by Eq S3: Multiplying p by 100% provides the %H2O2 selectivity of the ORR.It was determined that under electrochemical conditions, this system shows a 7.50 ± 1.3% selectivity for H2O2.Over the course of the reaction, 2 mL aliquots of the catalytic solution were removed and extracted with 10 mL of DCM and 5 mL of DI H2O.][9][10][11] Aliquots were taken at ~15 s, ~30 s, ~1 min, and ~2 mins.Experiments were done in triplicate.A calibration curve was used to establish Eqs S4-S5 and were used to calculate the % selectivity of H2O2, which was determined to be 102 ± 8.4% after 2 min.

Disproportionation Control
To determine the stability of H2O2 under catalytic conditions, control studies were conducted in the presence of im + , TFAH, and O2.Generally, solutions containing 8 µM [im + ] and 50 mM TFAH were sparged with O2 gas and rapidly mixed in a rapidly mixed in a 1:1 ratio with a N2 saturated urea•H2O2 solution (final concentrations: 4 µM im + , 0.93 mM urea•H2O2, 25 mM TFAH, 4.05 mM O2).As the solution was allowed to react, 2 mL aliquots were removed at 0 s and after 2 min, extracted with 10 mL DCM and 5 mL DI H2O.Then, 3 mL of the aqueous layer was removed and added to the cuvette.A UV-vis spectrum was taken before and after the addition of 0.1 mL of 0.1 M Ti(O)SO4 solution and the difference at 408 nm was used to determine the amount of H2O2 present ([H2O2]detected).The % recovery was determined according to Eq S6 from measured [H2O2]expected of the H2O2 stock solution.After 2 min, 104 ± 3.6 % H2O2 was recovered.

H2O2RR Control
To determine the stability of H2O2 in the presence of im 0 and TFAH, control studies were conducted in the presence of im + , TFAH, and Cp*2Fe, under an N2 atmosphere.In a N2-filled glovebox, urea•H2O2 was added to a solution containing im + , TFAH, and Cp*2Fe (final concentrations: 4 µM im + , 25 mM TFAH, 1 mM Cp*2Fe, and 1.5 mM urea•H2O2).After 2 min, a 2 mL aliquot was removed from the 'catalytic' solution and extracted with 10 mL of dry, degassed DCM and 5 mL of degassed water.Then, 3 mL of the aqueous layer was removed, and a UV-vis spectrum was taken before and after the addition of 0.1 mL of Ti(O)SO4 solution.The difference in the absorbance at 408 nm was used to quantify the amount of H2O2 present according to Eq S5, Eq S6 was used to calculate % H2O2 recovered relative to the stock H2O2 solution.

Computational Methods
Geometry optimization was done with the Gaussian 16 package 12 at the B3LYP-D3(BJ)/def2-TZVP level [13][14][15][16][17][18][19][20] with a complete structural model.Dispersion and bulk solvent effects (acetonitrile = MeCN; ε = 35.688)were accounted for at the optimization stage, by using Grimme's D3 parameter set with Becke-Johnson (BJ) damping 19,20 and the SMD continuum model, 21 respectively.The stationary points and their nature as minima (no imaginary frequencies) were characterized by vibrational analysis using the IGRRHO approach as implemented by default in the software package, which also produced enthalpy (H), entropy (S) and Gibbs energy (G) data at 298.15 K.The minima connected by a given transition state were determined by perturbing the transition states along the TS coordinate and optimizing to the nearest minimum.Free energies were corrected (ΔGqh) to account for concentration effects and for errors associated with the harmonic oscillator approximation.Thus, according to Truhlars's quasi-harmonic approximation for vibrational entropy and enthalpy, all vibrational frequencies below 100 cm −1 were set to this value. 22These anharmonic and concentration corrections were calculated with the Goodvibes code. 23Concentrations were set at 0.001 M for all species unless otherwise indicated, 0.004 M for O2, 0.500 M for TFAH, and 18.9 M for MeCN.Single point calculations for refining energy differences were completed with Orca 5.0 24 at the DLPNO-CCSD(T1)/cc-pVTZ level. 18,25,26 E[29][30][31] The stability of the wavefunction and spin contamination were studied at the double-and triple-zeta levels of theory.Reduction potentials from computational data were obtained according to our previous methodology by using the calculated free energy of reduction of the species of interest by [phenazine] -, corrected to the experimental potential of phenazine reduction vs Fc + /Fc. 32

Figure S5 .
Figure S5.(A) CVs of im + under Ar (black) and O2 (red) saturation with 0.065 M TFAH added (green).(B) First derivative of current density of im + under Ar saturation (from black trace in A) Conditions: Conditions: 1.3 mM im + , 0.1 M TBAPF6/MeCN; 100 mV/s; glassy carbon working electrode, glassy carbon counter electrode, Ag/AgCl pseudoreference electrode; referenced to an internal ferrocene standard.

Figure S7 .
Figure S7.(A) CVs of im + under Ar saturation in the presence of 0.261 M TFAH at varying concentrations.(B) Logarithm of im + concentration versus the reduction peak potential in (A).Conditions: varying [im + ], 0.261 M TFAH, 0.1 M TBAPF6/MeCN; glassy carbon working electrode, glassy carbon counter electrode, Ag/AgCl pseudoreference electrode; referenced to an internal ferrocene standard.

Figure S25 .
Figure S25.(A) spin density plots (0.025 iso) and (B) Kohn-Sham Orbitals (0.05 iso) of the neutral radical im 0 (S = ½) showing localization at C with contributions from N. Generated from the EPR calculation at the ωB97M-D4/def2-TZVPPD level of theory.