Supporting Information Electrochemical Reduction of Carbon Dioxide on Nitrogen-Doped Carbons: Insights from Isotopic Labeling Studies Dorottya Hursán 1,2 and Csaba Janáky 1,2* 1 Department of Physical Chemistry and Materials Science, University of Szeged, Rerrich Square 1, Szeged, H-6720, Hungary 2 MTA-SZTE Lendület Photoelectrochemistry Research Group, Rerrich Square 1, Szeged, H- 6720, Hungary *Corresponding author: janaky@chem.u-szeged.hu S1
1. Materials and methods 1.1 Synthesis of the HPG-PPy catalyst HPG-PPy was synthesized by a two-step pyrolysis of a polypyrrole precursor. The synthesis procedure was adopted from the literature (ref #14 in the main text), where an HPG carbon was prepared from a polyaniline hydrogel precursor. An aqueous solution (10 ml) containing 1.5 M ammonium peroxydisulfate (98%, Acros Organics) was added to a solution (13 ml) of 2.8 M freshly distilled pyrrole (98%, Sigma Aldrich) and 0.58 M phytic acid (50 w%, Sigma-Aldrich) at 0 C, under continuous stirring. In a few minutes the appearance of a black, dense precipitate indicated the polymer formation. After the addition of 50 ml water into the polypyrrole dispersion, stirring was continued for one hour at 0 C. The polymer dispersion was kept in the refrigerator for 24 hours, then freeze-dried to remove the water. Pyrolysis was performed in a tube furnace in N 2 -atmosphere (110 cm 3 min -1 flow rate) at 900 C for two hours (2 C min -1 heating rate until 600 C and 5 C min -1 until 900 C). The formed carbon was mixed with 7 M KOH (Fluka) with a KOH / carbon ratio of 3, and vacuum-dried at 60 C. The second heat treatment was performed at 800 C for one hour (5 C min -1 heating rate) in N 2 -atmosphere (110 cm 3 min -1 flow rate). The obtained catalyst was washed several times with water to remove the KOH residues until ph = 7 was reached. The HPG-PPy catalyst was spray-coated on a glassy carbon substrate with an automated spray-coating system. A catalyst suspension containing 5 mg/ml HPG-PPy and 250 µl Nafion dispersion (10%, Fuelcell store) was prepared in a water-ethanol mixture (50 V / V%). Spray coating was performed at 110 C until 2.1 (± 0.1) mg cm -2 loading was reached. 1.2 Electrochemical measurements and analysis of the reduction products For electrochemical measurements an Autolab PGSTAT204 potentiostat/galvanostat was used. In the electrochemical experiments we applied a three-electrode setup with the HPG- PPy catalyst coated on a glassy carbon substrate as the working electrode, a Pt foil as the counter- and an Ag/AgCl/3M NaCl as the reference electrode. We used a two-compartment sealed electrochemical cell (35 cm 3 electrolyte and 29 cm 3 headspace) where the cathode and anode were separated with a Nafion-117 membrane to prevent the mixing of the products. As we used electrolytes with different ph values, all the potentials are given versus the reversible hydrogen electrode (E RHE = E Ag/AgCl + 0.2 V + 0.059 V ph). CO 2 reduction experiments were performed in a KHCO 3 (Reanal) electrolyte saturated with CO 2 (99.95 %, Messer). Some control experiments were performed in KH 2 PO 4 - K 2 HPO 4 (Reanal) phosphate buffer solutions saturated with either CO 2 or Ar (99.96% Messer). Gas-phase products were analyzed with a Shimadzu-2010 Plus gas chromatograph equipped with a barrier discharge ionization detector (BID) and a Shincarbon ST column. The cathode compartment of the electrochemical cell was directly connected to the injection unit of the GC, which allowed the on-line detection of products. During electrolysis 0.5 ml of the headspace was injected into the GC in every 20 minutes of the electrolysis. Analysis S2
parameters were the following: carrier gas: helium; oven program: T start = 35 C (2.5 min), ΔT ramp = 20 C min -1, T end = 270 C (3 min); injection temperature: T = 150 C; linear velocity was controlled by the pressure p start = 250 kpa (2.5 min) Δp ramp = 15 kpa min 1 p end = 400 kpa (7.5 min); and split ratio: 10. 1.3 Isotopic labeling experiments In joint isotopic labeling experiments 13 CO 2 (99 atom% 13 C, Sigma-Aldrich) and NaH 13 CO 3 (98 atom% 13 C, < 3 atom% 18 O, Sigma-Aldrich) were used as the carbon source. In the selective labeling experiments either the CO 2 or the NaHCO 3 was 13 C labeled, while the other component was unlabeled. In the selective labeling experiment where bicarbonate was the labeled component, 35 cm 3 ultrapure water (in the electrochemical cell) was saturated with 12 CO 2, then the appropriate amount of NaH 13 CO 3 was dissolved in it. After the addition of the salt, the cell was immediately sealed. When the CO 2 gas was the labeled component, first the 35 cm 3 water was purged with Ar to remove all the dissolved gases, which was followed by the introduction of 13 CO 2 (42.6 cm 3 min -1 flow rate for 11 minutes). Finally, the NaH 12 CO 3 was dissolved in the 13 CO 2 -saturated water and the cell was sealed similarly. In the joint labeling experiment, the 13 CO 2 was introduced to the Ar-saturated 35 cm 3 0.1 M NaH 13 CO 3, similarly to the previous experiment. Non (isotopic) equilibrium condition means that electrolysis was started right after the addition of the bicarbonate salt to the cell, that is when the isotopic composition of the CO 2 gas is different from that of bicarbonate. Under (isotopic) equilibrium conditions, the electrolysis was started two hours after the addition of the bicarbonate to the cell. During this two hours, the electrolyte was continuously stirred. The isotopic composition of the CO product was monitored with GC MS (Shimadzu GC MS QP2010 S, RT Molsieve 5 Å column), in selective ion monitoring mode. The signal of the m/z = 28 and m/z= 29 molecule ions of the CO were used in the analysis. In every 10 minutes of the electrolysis 200 µl of the headspace gas was injected into the GC-MS with a gastight syringe. Analysis parameters were the following: T start = 90 C (0 min), ΔT ramp = 10 C min -1, T end = 110 C (0 min), ΔT ramp = 30 C min -1, T end = 150 C (3 min); injection temperature: T = 200 C; linear velocity was controlled by the pressure p = 50 kpa; and split ratio: 50. In the experiment where the isotopic composition of the CO 2 gas in the headspace was monitored under selective labeling conditions (bicarbonate is labeled), a ZB-WAX Plus column was applied. In this column CO 2 is not retained and by the monitoring of the m/z = 44, 45, 13, 12 ions the isotopic composition of the CO 2 gas can be determined. The same procedure was used for the preparation of the 12 CO 2 -saturated 0.1 M NaH 13 CO 3 as stated above. S3
I / ma cm -2 Intensity / a.u. I / ma cm -2 I / ma cm -2 2. Electrochemical characterization of the HPG-PPy catalyst and analysis of the reduction products Cyclic voltammogram of the HPG-PPy electrode (Fig. S1A) was recorded in a potential range where no Faradaic process takes place. The large capacitive current is indicative of a porous, high surface area electrode. Comparing the charge capacitance of the HPG-PPy electrode to the bare glassy carbon substrate with known geometrical surface area (1 cm 2 ), we can estimate the electrochemical surface area of the HPG-PPy electrode, which is around 75 cm 2 /1 cm 2 geometrical surface area. Based on the linear sweep voltammograms (Fig. S1B) recorded in the presence and absence of CO 2, we decided to perform the CO 2 reduction experiments at 0.61 and 0.71 V. At these potentials higher currents can be observed in the presence of CO 2. A representative chronoamperometric curve of the long-term potentiostatic electrolysis at 0.71 V is shown in Fig. S1C. The quite fast current decay in the first 5 minutes of the electrolysis is probably because of the decrease in the capacitive current contribution. After this period, the current reached a steady-state value around 0.7 ma cm -2, without a major decrease within two hours. During the electrolysis the amount of the generated H 2 and CO was continuously increasing (Fig.S1D). A 15 10 HPG-PPy glassy carbon substrate B 0 5 0-5 0.5 M phosphate buffer/ Ar 0.5 M KHCO 3 / CO 2-5 -10 C -10 50 mv s -1-15 -15-0.2-0.1 0.0 0.1 0.2-1.0-0.8-0.6-0.4-0.2 0.0 E (vs. RHE) / V -0.6-0.8-1.0 E (vs. RHE / V) 0.5 M KHCO 3 / CO 2 D 1.0x10 6 8.0x10 5 6.0x10 5 18.9 min 42.5 min 64.7 min 87.2 min 109.8 min H 2 5 mv s -1 CO -1.2-1.4 0.1 M KHCO 3 / CO 2-0.71 V vs. RHE 4.0x10 5 2.0x10 5 2000 4000 6000 8000 0.6 0.7 0.8 2.4 2.5 2.6 2.7 t / s t R / min Fig. S1. A: Cyclic voltammogram of a HPG-PPy electrode in a CO 2 -saturated 0.5 M KHCO 3 solution (50 mv s -1 scan rate). B: Linear sweep voltammograms (5 mv s -1 ) of a HPG-PPy electrode in an Arsaturated 0.5 M phosphate buffer and in a CO 2 -saturated 0.5 M KHCO 3 electrolyte. Dashed lines indicate the potentials at which the potentiostatic electrolysis experiments were performed. C: Chronoamperometric curve recorded during potentiostatic electrolysis at -0.71 V (vs. RHE) on a HPG- PPy electrode in a CO 2 -saturated 0.1 M KHCO 3 solution. (The noise is because of the pumping of the gas phase before the sampling for GC analysis). D: Chromatograms (GC-BID) of the products with increasing electrolysis time. S4
MS intensity / a.u. [ 12 CO] / [ 13 CO] [ 13 CO 2 ] / [ 12 CO 2 ] 3. Isotopic composition of the CO 2 gas in time in selective labeling experiments 1.6 1.2 0.1 M NaH 13 CO 3 / 12 CO 2 0.8 0.4 0.0 0 2000 4000 6000 8000 t / s Fig. S2. Change in the ratio of the concentration of the 13 CO 2 to the 12 CO 2 in time in the headspace of the electrolysis cell containing 0.1 M NaH 13 CO 3 saturated with 12 CO 2. We performed a separate experiment (i.e., without electrolysis) to determine the isotopic composition of the CO 2 gas in the headspace of the electrolysis cell in a selective labeling experiment where the bicarbonate was the isotopically labeled compound. At the beginning, most of the CO 2 was unlabeled (only around 1.5 % is labeled after 1 minute). The concentration ratio of the 13 CO 2 to the 12 CO 2 is continuously increasing in time, as equilibrium exists between bicarbonate and CO 2,aq as well as CO 2,aq and CO 2,g. After two hours the 13 C: 12 C ratio in the CO 2 gas approaches the theoretical ratio. The fact that it takes around two hours to reach the isotopic equilibrium gives a handle to discriminate between CO 2 and bicarbonate if solely one of them is labeled. 4. Joint isotopic labeling experiments A 1.6x10 6 1.2x10 6 8.0x10 5 CO 13 CO 2 / 0.1 M NaH 13 CO 3 m/ z = 28 m / z = 29 B 0.07 0.06 0.05 0.04 0.03 0.1 M NaH 13 CO 3 / 13 CO 2 4.0x10 5 0.0 3.85 3.90 3.95 4.00 t R / min 0.02 0.01 0 2000 4000 6000 8000 t / s Fig. S3.: A: GC-MS signal in selective ion monitoring mode of the CO product formed during potentiostatic electrolysis on an HPG-PPy electrode in a joint labeling experiment. Traces of the m/z = 29 and m/z = 28 ions are shown, corresponding to the 13 CO and 12 CO, respectively. B: Concentration ratio of the 12 CO to the 13 CO in time during electrolysis. S5
To prove that the detected CO indeed originates from the reduction of CO 2 /bicarbonate and not from the incidental degradation of the carbon catalyst itself, we performed the electrolysis with all the carbon sources being isotopically labeled (0.1 M NaH 13 CO 3 saturated with 13 CO 2 ). Under such conditions, around 98% of the generated CO was 13 C labeled and this ratio remained stable during the time of the experiment (130 min). The 2% 13 CO originates from the 12 C content of the 13 C-labeled CO 2 gas and bicarbonate (note the 98% purity). These results confirm that CO is produced from CO 2 and the degradation of the catalyst is not evidence. 5. Approximate concentrations of dissolved CO 2 in the different applied electrolytes and the respective ph values Electrolyte [CO 2,aq ] / mm ph (before ph (after electrolysis) electrolysis) (i) 0.1 M phosphate buffer / CO 2 35 3.81 3.99 (ii) 0.13 M KHCO 3 / Ar 0.8 8.78 8.81 (iii) 0.1 M KHCO 3 / CO 2 33 6.64 6.72 For electrolyte (i) it was assumed that all the dissolved carbon is CO 2,aq and the bicarbonate concentration is negligible. Furthermore we suppose that he CO 2,aq concentration under 1 atm CO 2 partial pressure is independent of the ionic strength of the electrolyte. The approximate aqueous CO 2 concentration of electrolyte (ii) was calculated by extrapolation of the data presented in Fig. 3D of ref #1 in the SI 1. For electrolyte (iii) data in Fig. 5. of ref #16 in the main text was used. Note that the concentrations stated in the table are rough estimates, not based on precise calculations. Data are used only to qualitatively explain some trends. The ph values are shown for the as is solutions and for those after 2h electrolysis. The ph slightly increased during the electrolysis in all cases, which can be explained by the reduction of H +. The ph value, however, remained in the range where only one species of the CO 2 - bicarbonate system is dominant in case of the (i) and (ii) electrolyte. References: (1) Min, X.; Kanan, M. W. Pd-Catalyzed Electrohydrogenation of Carbon Dioxide to Formate: High Mass Activity at Low Overpotential and Identification of the Deactivation Pathway. J. Am. Chem. Soc. 2015, 137, 4701 4708 S6