Building and identifying highly active oxygenated groups in carbon materials for oxygen reduction to H2O2

The one-step electrochemical synthesis of H2O2 is an on-site method that reduces dependence on the energy-intensive anthraquinone process. Oxidized carbon materials have proven to be promising catalysts due to their low cost and facile synthetic procedures. However, the nature of the active sites is still controversial, and direct experimental evidence is presently lacking. Here, we activate a carbon material with dangling edge sites and then decorate them with targeted functional groups. We show that quinone-enriched samples exhibit high selectivity and activity with a H2O2 yield ratio of up to 97.8 % at 0.75 V vs. RHE. Using density functional theory calculations, we identify the activity trends of different possible quinone functional groups in the edge and basal plane of the carbon nanostructure and determine the most active motif. Our findings provide guidelines for designing carbon-based catalysts, which have simultaneous high selectivity and activity for H2O2 synthesis.

The onset potential was defined as the potential, at which the current density of ORHP reached 0.15 mA cm −1 (about 5% of the limiting current). c The potential is unknown.

Supplementary Methods:
Preparation of partially reduced graphene oxide (pRGO): Graphene oxide was prepared by the Hum-mers′ method. 12 In detail, the graphene oxide was synthesized by charging purified natural graphite (1 g, Alfa Aesar, 100 mesh, 99.9995 %) and potassium nitrate (0.6 g, Alfa Aesar) in sulfuric acid (25 mL, Sigma Aldrich, 99.999%) in a one-neck round bottom flask (50 mL). The reaction flask was placed in an ice-bath. Then, potassium permanganate (3 g, Sigma Aldrich) was slowly added into the suspension, while vigorously agitating. Safety Note: the temperature of the suspension should be kept lower than 20 °C.
After adding potassium permanganate, the ice-bath was removed and the suspension was held at room temperature for 30 min. Finally, ultra-pure water (50 mL, 18.2 MΩ cm, Direct-Q ® 3UV, Millipore Corporation) was slowly added to dilute the suspension, and maintained for 15 min. The suspension was then further diluted with ultra-pure water (150 mL) and treated with diluted hydrogen peroxide (3%, Sigma Aldrich, 30 wt% in H2O) to remove the residual permanganate and manganese oxides to colorless manganese sulfate. 12 After further careful rinsing with ultra-pure water, the suspension was stored in a refrigerator for subsequent usage.
The chemical conversion of graphite oxide to partially reduced graphene oxide (pRGO) was performed by adopting the method in Li et al. 13 The typical procedure involved treating a homogeneous suspension (150 mL, approximately 0.6 wt% in H2O) with ammonia solution (18 mL, Sigma Aldrich, 28-30 wt% in water) and hydrazine monohydrate (2.05 mL, Alfa Aesar, 98 wt%). The calculated weight ratio of hydrazine to GO was around 7:10. 13 After sonication for 20 minutes, the suspension was placed in a water-bath at 95 °C for 1 h. 13 The color of the suspension changed from brown to dark black, which indicated that the reduction had occurred. Finally, after copiously rinsing and freeze-drying in tert-butyl alcohol, the resultant RGO was further dried in a vacuum oven at 80 °C for 10 h.

Structural characterization:
The soft XANES spectra were recorded with a resolution of 0.2 eV for the K-edge of carbon and oxygen in the total electron yield (TEY) mode. The calibration of photon energy was done using the π* resonance position in the carbon K-edge spectrum of graphite (285.4 eV) 14 . For S29 the oxygen K-edge, the photon energy was calibrated using the π* resonance position of carbonate (534 eV) within the high emitting regions 14 . For easy comparison, the spectra were first normalized to the intensity of 280 eV for carbon and 525 eV for oxygen, respectively, and then by the maximum intensity of each spectrum again, so that each spectral line ranged from 0 to 1 15 .

Electrochemical measurements:
The calibration of the reference electrode potential was performed by measuring the onset potential of the hydrogen oxidation reaction (HOR) for commercial Pt black (Sigma Aldrich). The electrolyte was first bubbled with pure H2 for 30 min, and a positive scan rate was set to as low as 1 mV s −1 . The onset potential of HOR was defined as the point, where the current was bigger than zero. The potential was determined to be −0.

S30
Determination of quinone by CV: The quinone redox was measured by CV method in N2-saturated 0.5 M aq. H2SO4. Here, H2SO4 was selected as the supporting electrolyte, because the proton has a much higher mass transfer speed than OH − , and the Debye length of its electrical double layer (EDL) is smaller than that in a base. The CV curves were recorded at a scan rate of 50 mV s −1 . For fair comparison, all CV curves were normalized with the 1 F.

Determination of collection efficiency (N):
The collection efficiency was determined using the previous report. 16  The resulting collection efficiency (N) was determined to be 37 ± 1 %, which is consistent with the value (37 %) provided by the manufacturer.

S31
Polarization curves were obtained in the O2-saturated 0.1 M aq. KOH solution at a rotation speed of 1600 RPM. Disk current density (JD) was collected by CV between 0.05 V and 1.00 V at a scan rate of 10 mV s −1 . Ring current density (JR) was recorded at a constant applied potential of 1.15 V. To rule out the interference of capacitance current, polarization curves were calculated by averaging the forward and backward scans. True activity was calculated by normalizing the current with the capacitance. Here, we did not employ electrochemical specific surface area (ECSA), because it is difficult to find a proper conversion factor. This is due to the presence of pseudocapacitance caused by the functional groups.
Hydrogen peroxide current density (JH2O2) was calculated using the following equation:
The current density of the byproduct H2O was determined using the following equation: The ratio of H2O2 yield (H2O2 %) was determined from RRDE data, using the following equation: Stability test in H-type cell: The stability test was conducted in an H-type cell. A 115 Nafion membrane (Sigma Aldrich) separated the anodic and cathodic cells. The counter electrode and reference electrode were placed in the anodic cell, and the working electrode was immersed in the cathodic cell. The electrolyte was 0.1 M aq. KOH. MgSO4 (400 ppm) was added to the cathodic cell as a stabilizer. The working electrode was prepared by drop-casting 40 μL ink (10 mg mL −1 with 0.2 wt% Nafion) 4 times on carbon paper (JNT20, JNTG). The typical working area was 1 cm × 2 cm. The other part of the carbon paper was sealed with epoxy resin. Stability was tested using a chronoamperometry method. The applied potential S32 was 0.65 V. After every 30 h measurement, the electrolyte was replaced with fresh, for a total of four times.

H 2 O 2 concentration measurement and the Faraday efficiency:
The concentration of yielded H2O2 was determined by permanganate titration, which was considered one of the most precise and reliable analytical methods. However, since KMnO4 is partially reduced to MnO2 in basic solution, which could act as a catalyst for H2O2 decomposition, the titration was conducted after acidifying the H2O2 solution. Into the 20 mL as-prepared solution, 0.5 M aq. H2SO4 (5 mL) was added. The titration reaction follows the following stoichiometry: 2 MnO4 − + 5 H2O2 + 6 H + → 2 Mn 2+ + 8 H2O + 5 O2 (6) The titration can be observed the purple solution of MnO4 − is reduced by H2O2 into colorless Mn 2+ .
Here, we selected a commercially available potassium permanganate solution (0.02 M, equivalent concentration 0.1 N, Reag. Ph Eur, Sigma Aldrich) as the titration reagent. The molar concentration of H2O2 (CH2O2) can be determined by the following equation: where CKMnO4 is the molar concentration of the potassium permanganate solution, which is equal to 0.02 M; VKMnO4 is the titration volume of the potassium permanganate solution; the VH2O2 is the volume of hydrogen peroxide solution used for analysis.
Faraday efficiency during the stability test was determined using the following equation: Faraday efficiency % 8 where n is the electron transfer number, which is 2 for the dioxygen reduction into H2O2; F is the Faraday constant (96485.3 C mol −1 ); is the consumed quantity of electric charge (C).