Catalysis and Inhibition in the Electrochemical Reduction of CO2 on

Useful Lessons from a Methodological Misadventure. Cyrille CostentinJean-Michel SavéantCédric Tard. ACS Energy Letters 2018 3 (3), 695-703. Abstract...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/JACS

Catalysis and Inhibition in the Electrochemical Reduction of CO2 on Platinum in the Presence of Protonated Pyridine. New Insights into Mechanisms and Products Hachem Dridi,† Clément Comminges,‡ Claudia Morais,‡ Jean-Claude Meledje,† Kouakou Boniface Kokoh,‡ Cyrille Costentin,*,† and Jean-Michel Savéant*,† †

Sorbonne Paris Cité, Laboratoire d’Electrochimie Moléculaire, Unité Mixte de Recherche Université - CNRS no. 7591, Université Paris Diderot, Bâtiment Lavoisier, 15 rue Jean de Baïf, 75205 Paris Cedex 13, France ‡ IC2MP UMR-CNRS 7285, Université de Poitiers, 4 rue Michel Brunet, TSA 51106, 86073 Poitiers Cedex 9, France S Supporting Information *

ABSTRACT: In the framework of modern energy challenges, the reduction of CO2 into fuels calls for electrogenerated lowvalent transition metal complexes catalysts designed with considerable ingenuity and sophistication. For this reason, the report that a molecule as simple as protonated pyridine (PyH+) could catalyze the formation of methanol from the reduction of CO2 on a platinum electrode triggered great interest and excitement. Further investigations revealed that no methanol is produced. It appears that CO2 is not really reduced but rather participates, on the basis of its aquation into carbonic acid, in hydrogen evolution. Actually, the situation is not that straightforward, as revealed by scrutinizing what happens at the platinum electrode surface. The present study confirms the lack of methanol formation upon bulk electrolysis of PyH+ solutions at Pt and provides a detailed account of the Faradaic yield for H2 production as a function of the electrode potential, but the main finding is that CO2 reduction is accompanied by a strong inhibition of the electrode process taking place when it is carried out in the presence of acids such as PyH+ and AcOH. Cyclic voltammetry and in situ infrared spectroscopy were closely combined to investigate and understand the nature and consequences of the inhibition process. Constant comparison between the two acids was required to decipher the course of the reaction owing to the fact that the IR responses are perturbed by PyH+ adsorption. It finally appears that inhibition is caused by the reduction of CO2 into CO, whose high affinity with platinum triggers the formation of a Pt−CO film that prevents the reaction process. Thus, a paradoxical situation develops in which the high affinity of Pt for CO helps to decrease the overpotential for the reduction of CO2 and therefore blocks the electrode, preventing the reaction process.



INTRODUCTION Many researchers all over the world are actively involved in the quest for solutions to modern energy shortcomings.1,2 Among these, the reduction of CO2 into fuels, ultimately using solar energy, is one of the promising routes toward this goal.3 Injection of one electron into CO2, to form the CO2 anion radical, is energetically very costly (standard potential −1.97 V vs SHE),4 calling for catalytic pathways that could avoid going through this intermediate. Electrogenerated low-valent transition metal complexes have been used for this purpose, with a wealth of ingenuity and sophistication in the choice of metals and ligands.5−13 In this context, considerable interest and excitement were triggered by the report that a molecule as simple as protonated pyridine could catalyze the formation of methanol in a sizable amount (20% Faradaic yield) according to a 6e−+6H+ process, despite the use of a precious metal electrode (Pt or Pd).14 In view of the disagreements that subsequently appeared as to the nature and yields of the electrolysis products, it is worth © 2017 American Chemical Society

recalling the exact content of the initial accounts on the matter. The first of these involves a “hydrogenated” palladium electrode operated in the presence of protonated pyridine (PyH+) that indicates the formation of methanol in a significant Faradaic yield (up to 30%).14a This reaction is reminiscent of previous reports on the reduction of CO2 on carbon−palladium composite electrodes (a mixture of Pd with carbon deposited onto a methylviologen coating),15 where formate, rather than methanol, was the reaction product. Unfortunately, fewer details were given for the preparative-scale electrolysis carried out under similar conditions on a platinum electrode, where the yields are reported to be “practically the same” as on the “hydrogenated” palladium electrode.14a We note that none of these reports quantitated the Faradaic yield of the major product of electrolysis, hydrogen, leading to uncertainties concerning the overall Faradaic yield. Received: July 30, 2017 Published: September 11, 2017 13922

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928

Article

Journal of the American Chemical Society In contrast with the above-mentioned results, it has been reported that preparative-scale electrolysis carried out under various conditions (Table 1 in ref 16), including some very close to those of ref 14b, did not show the formation of any methanol or any formate, hydrogen being the only reduction product.17 It was also noted that the cyclic voltammetric (CV) responses obtained on a platinum electrode in the presence of PyH+‑and CO2 obey, on quantitative grounds, the mechanism depicted in Scheme 1.16 Attempts to fit the CV responses with mechanisms in which PyH+ would work as catalyst for the reduction of CO2 into methanol, as in Scheme 1 of ref 14b, were shown to be incorrect.16

Table 1. Production of Bulk Products upon Electrolysis at a Pt Electrodea Faradaic yields (%)b applied potential (V vs SHE)

gas

pH

charge (C)

H2

CH3OH

HCO−2

−0.46 −0.26 −0.46 −0.56 −0.56 −0.61

Ar CO2 CO2 CO2 CO2 CO2

4.7 4.5 4.5 4.65 4.7 4.5

73.5 6.23 31.2 31.8 148 168.3

96 106.6 95.3 99.3 104.5 95.8

b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.

b.d.l. b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.

a

10 mM PyH+/Py solutions in the presence of 0.1 M supporting electrolyte (KCl) under 1 atm of either argon or CO2. bb.d.l. indicates below detection limit.

Scheme 1

At this stage, the situation would seem rather clear: on platinum, in the presence of a weak acid, such as PyH+ or AcOH, CO2 is not really reduced but rather participates, on the basis of its conversion into carbonic acid, in hydrogen evolution. Is this the end of the story? Actually, we have found that the situation is not as straightforward as it may appear at first glance when what happens at the platinum electrode surface is taken into account more carefully. In the present study, we confirm the absence of methanol and formate formation upon bulk electrolysis of PyH+ solutions at a Pt electrode and provide a detailed account of the Faradaic yield for H2 production as a function of the electrode potential. Our main finding is however that the CO2 reduction is accompanied by a strong inhibition of the electrode when it is carried out in the presence of weak acids such as PyH+ and acetic acid. Our strategy to investigate and understand the nature and consequences of the inhibition process was a close combination of cyclic voltammetry and in situ infrared spectroscopy. A crucial item of our approach is the permanent comparison between the effect of PyH+ and AcOH. Deciphering the IR responses is indeed easier in the latter case than in the former on account of perturbing PyH+ adsorptive effects.

Figure 1. (a) Current traces for a potentiostatic bulk electrolysis at −0.33 V vs SHE of a 10 mM pyridine/pyridinium 55 mL solution at pH 4.65 with 0.1 M KCl during 23 min with ohmic drop compensation on a basket Pt electrode (47.15 cm2) under CO2 (blue line) and argon (green line). (b) Cyclic voltammograms at 0.1 V s−1 on a basket Pt electrode in 10 mM pyridine/pyridinium 55 mL solution at pH 4.65 under CO2 with 0.1 M KCl before (black) and after (blue) a 22 s electrolysis at −0.46 V vs SHE.

observed for an electrolysis run under argon but to a much lesser extent (Figure 1a, green curve). To get further insights into the electrode deactivation process, cyclic voltammograms were recorded on a Pt disk electrode. Under argon, pyridinium leads to a quasi-reversible wave corresponding to the reduction of a weak acid into H2 (red curve in Figure 2a). As opposed to the case of acetic acid reduction (red curve in Figure 2b), no clear underpotential deposition (UPD) wave is observed with the pyridinium (red curve in Figure 2a). This is because pyridine is adsorbed on Pt and thus presumably blocks sites for underpotential deposition of hydrogen atoms.19 However, pyridine adsorption does not prevent quasi-reversible pyridinium reduction into H2, thus confirming that underpotential-deposited hydrogen atoms are not intermediates in the hydrogen evolution reaction.20 Repetitive scanning under argon shows no surface inhibition phenomena for acetic acid reduction and only a slight current decrease for pyridinium reduction (Figure 2a and 2b). However, repetitive scanning under CO2 shows an important decrease in the current in the presence of either pyridinium or acetic acid (Figure 2c and 2d) in agreement with the rapid inhibition process observed during preparative-scale electrolysis experiments under CO2 unlike that observed under an argon atmosphere. In the case of acetic acid as a weak acid, as already discussed in ref 16, the quasi-reversible wave corresponding to hydrogen evolution (at ca. −0.3 V vs SHE) increases upon addition of CO2 (Figure 3a, comparison of blue and black CVs) due to the



RESULTS AND DISCUSSION Electrolysis of 10 mM PyH+/Py solutions in the presence of 0.1 M supporting electrolyte (KCl) under 1 atm of either argon or CO2 at a pH value of ca. 4.5 was carried on a Pt basket mesh electrode at various constant potentials (see Table 1). In all cases, H2 was the only product detected in the bulk of the solution with a Faradaic yield of 100(±5)% (see Supporting Information for details on analytical procedures).18 Importantly, besides the absence in solution and gas phases of any traces of product resulting from CO2 reduction, we note a systematic rapid decrease in the current when electrolysis is conducted under CO2 atmosphere (Figure 1a, blue curve). It is attributed to a blockage of the electrode surface as also demonstrated by cyclic voltammetry runs in the electrolysis solution on the working electrode used for electrolysis before and after only 22 s of electrolysis, showing a large shift of the hydrogen evolution wave (Figure 1b). A current decrease is also 13923

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928

Article

Journal of the American Chemical Society

Figure 3. (a) Cyclic voltammograms on a Pt disk electrode (7 mm diameter) of a 10 mM acetic acid/acetate solution at pH 5.05 in the presence of 0.4 M K2SO4 at 0.1 V s−1: black: under argon. Blue, red, green: under CO2. (b) Same as part a but zoomed in on the UPD and hydrogen evolution potential range. (c and d) Same as part a but with a black arrow indicating initial scan potential and direction.

Figure 2. (a) Fifty successive cyclic voltammograms on a Pt disk electrode (1 mm diameter) of a 10 mM pyridinium/pyridine solution at pH 5.13 in the presence of 0.4 M K2SO4 at 0.1 V s−1 under argon atmosphere. (b) Fifty successive cyclic voltammograms on a Pt disk electrode (1 mm diameter) of a 10 mM acetic acid/acetate solution at pH 5.08 in the presence of 0.4 M K2SO4 at 0.1 V s−1 under argon atmosphere. (c) Fifty successive cyclic voltammograms on a Pt disk electrode (1 mm diameter) of a 10 mM pyridinium/pyridine solution at pH 5.13 in the presence of 0.4 M K2SO4 at 0.1 V s−1 under CO2. (d) Fifty successive cyclic voltammograms on a Pt disk electrode (1 mm diameter) of a 10 mM acetic acid/acetate solution at pH 5.08 in the presence of 0.4 M K2SO4 at 0.1 V s−1 under CO2 atmosphere. In each case, the red curves represent the first scan.

adsorbed CO and its reoxidation during the anodic sweep as well as the absence of either methanol or formate formation (reference FTIR spectra as well as FTIR spectra recorded at the end of the electrolysis are given in Supporting Information, Figure S1). In both cases, an upward directed band at 2343 cm−1 related to CO2 indicates that CO2 is consumed during the electrolysis. A part of this consumption goes to the HCO3− form due to the acidic proton reduction of carbonic acid. For both electrolysis at −0.33 (H2 evolution wave) and −0.14 (UPD wave) V vs SHE, a band at ca. 2050 cm−1 appears, which can be attributed to CO linearly bonded to the electrode surface (COL). Figure 4 indeed shows data displayed as R/R0, with R0 being the signal at the electrolysis potential recorded at the end of the electrolysis and R the signal at a potential E during the anodic scan after electrolysis (spectra are recorded with 50 mV intervals). Because the linearly adsorbed CO absorption band is shifting with potential, a bipolar band around 2050 cm−1 is initially observed as the potential is scanned anodically. Then once the potential reaches a value positive enough for CO oxidation to occur (at ca. 0.4 V vs SHE, see Figure 4b), a monopolar upward directed band is observed. This oxidation of CO to CO2 is concomitant to the disappearance of HCO3− species (upward absorption bands of Figure 4a) and SO42− to HSO4− conversion (up and downward absorption bands, respectively, in the 1050−1250 cm−1 region (Figures 4a,c), indicating a pH decrease near the electrode. Interestingly, although CO2 reduction is shown to occur through reaction of CO2 with an adsorbed hydrogen atom, the final product is CO rather than formic acid or formate, thus pointing to a complex mechanism. CO2 reduction into CO is made possible at a potential as low as −0.14 V vs SHE at pH 5 because of the strong driving force for CO adsorption on Pt.

additional H2CO3 reduction. An anodic wave peaking at ca. 0.6 V vs SHE is observed in the reverse scan when the cyclic voltammogram is run under CO2 (Figure 3a, comparison of blue and black CVs). Note that the shallow anodic wave around 1 V vs SHE and the cathodic wave at 0.45 V vs SHE correspond to Pt oxide formation and its reduction. Observation of a new anodic wave under CO2 does not actually require scanning for the hydrogen evolution wave but only for the underpotential deposition region as shown in Figure 3a (comparison of blue, red and green CVs) or Figure 3d and as already known in the case of CO2 reduction on Pt in strongly acidic media.21 The products resulting from reaction between the underpotentialdeposited hydrogen atoms and CO2 in strongly acidic conditions were referred to as “reduced CO2” adsorbates and characterized by in situ infrared (IR) spectroscopy as being, at least partially, bonded CO, an anodic wave corresponding to bonded CO oxidation.21b Interestingly, in our experiments, the intensity of the surface UPD wave, which is correlated with the anodic wave, doubles in the presence of CO2 (Figure 3b). This observation is in line with a two-electron reduction of CO2 to CO as opposed to the one-electron formation of adsorbed hydrogen. In situ Fourier transform infrared (FTIR) spectra measurements during an anodic sweep after a 30 min electrolysis of a 10 mM acetic acid/acetate solution in the presence of CO2 at pH 5.05 on Pt electrodes were made to further prove formation of 13924

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928

Article

Journal of the American Chemical Society

Figure 4. (a) FTIR spectra at various potentials at a Pt electrode (1 mm diameter) during the anodic scan (1 mV s−1) after a 30 min electrolysis at −0.33 V vs SHE in a 10 mM acetic acid/acetate solution at pH 5.05 with 0.4 M K2SO4 under CO2. From top to bottom, spectra correspond to potentials changing by 50 mV intervals from the starting potential at −0.33 V vs SHE. (b) Current recorded during the anodic scan corresponding to part a. Black arrow indicates initial potential and scan direction. (c) “IR spectra” at various potentials at a Pt electrode during the anodic scan (1 mV s−1) after a 30 min electrolysis at −0.14 V vs SHE in a 10 mM acetic acid/acetate solution at pH 5.05 with 0.4 M K2SO4 under CO2. From top to bottom, spectra correspond to potentials changing by 50 mV intervals from the starting potential at −0.14 V vs SHE. (d) Current recorded during the anodic scan corresponding to part c. The black arrow indicates initial potential and scan direction.

According to a previous study of CO2 reduction on a single crystal Pt electrode in acidic media, observation of COL would imply that our polycrystalline electrode surface contains predominantly Pt(110) vs Pt(100) planes.22 This is actually in line with the observation of two UPD waves, the most positive of which can be attributed to Pt(100) planes and the most negative one being attributed to Pt(110) planes by analogy to CVs recorded on single crystal electrodes reported in ref 23. We also note that formation of COL does not inhibit H2 evolution during cyclic voltammetry unless repetitive scans are recorded. This may be due to the presence of Pt(111) planes known to have little activity toward CO2 reduction in the UPD region. In addition, the maximum surface coverage of adsorbed CO is less on (111) than on (100) faces, which might explain that a large amount of (111) faces will lead to a lesser or at least slower blocking of the electrode surface.24 Accumulation of CO is required to fully block the electrode surface. Such an accumulation may result from lateral diffusion of adsorbed CO from Pt(110) planes to others to finally block the entire surface. Cyclic voltammograms recorded under CO after a 10 min electrolysis at −0.21 or −0.36 V vs SHE show a complete disappearance of the hydrogen evolution wave in the

potential range where a reversible HER wave is expected (see Figure S2). In the case of pyridinium, no clear UPD wave is observed prior to the hydrogen evolution wave under argon (or nitrogen) atmosphere (see comparison of Figure 3a and 3b), and it is noticed that the capacitive current of the Pt electrode is smaller than in the presence of acetic acid. These observations are in line with the known adsorption of pyridine on Pt electrodes.19 The cyclic voltammogram of a pyridine/ pyridinium solution recorded under CO after a 10 min electrolysis at −0.36 V vs SHE shows a new (as compared to the CV under argon) small anodic feature but much smaller than the CO stripping wave observed in the presence of acetic acid (Figure S3 in comparison with Figure S2). This indicates that CO adsorption sites are much less abundant due to pyridine adsorption on Pt. Nevertheless, a competition between Pyads and COads is observed when pure CO is adsorbed at −0.18 V vs SHE in the presence of pyridinium as shown in the in situ FTIR spectra recorded during the CO stripping experiment (Figure 5a), but a large amount of CO is required to displaced Pyads so as to be able to detect it on the surface. The COL absorption band at 2050 cm−1 is associated with the 13925

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928

Article

Journal of the American Chemical Society

Figure 5. (a) FTIR spectra at various potentials at a Pt electrode recorded during anodic scan (at 1 mV s−1) after CO adsorption at −0.18 V vs SHE in a 10 mM pyridinium/pyridine solution at pH 5.35 with 0.4 M K2SO4. From top to bottom, spectra correspond to potentials changing by 50 mV intervals from the starting potential at −0.18 V vs SHE. (b) FTIR spectra at various potentials at a Pt electrode during anodic scan (1 mV s−1) after a 20 min electrolysis at −0.18 V vs SHE in a 10 mM pyridinium/pyridine solution at pH 5.35 with 0.4 M K2SO4 under CO2. From top to bottom, spectra correspond to potentials changing by 50 mV intervals from the starting potential at −0.18 V vs SHE. (c) FTIR spectra at various times at a Pt electrode recorded during electrolysis at −0.39 V vs SHE in a 10 mM pyridinium/pyridine solution at pH 5.35 with 0.4 M K2SO4 under CO2. (d) FTIR spectra at various potentials at a Pt electrode during anodic scan (1 mV s−1) after a 30 min electrolysis at −0.39 V vs SHE in a 10 mM pyridinium/pyridine solution at pH 5.35 with 0.4 M K2SO4 under CO2. From top to bottom, spectra correspond to potentials changing by 50 mV intervals from the starting potential at −0.29 V vs SHE.

downward CO2 absorption band at 2343 cm−1 that presents increasing intensity once the potential is high enough for CO oxidation to occur. Modifications on the ν19b ring vibration mode of pyridine are also clearly highlighted. Initially adsorbed on Pt surface, pyridine desorbs from the surface as soon as the potential value is high enough (upward band at 1444 cm−1), giving rise to PyrH+ (downward band at 1491 cm−1), which coordinates to PtO surface at higher potential values (downward band at 1456 cm−1). Here again the pH value varies during the anodic scan, as indicated by the appearance/ disappearance of absorption bands in the 1050−1250 cm−1 spectral range, corresponding to SO42−/HSO4− species. Besides, it is seen in Figure S3 that, under CO atmosphere,

acid reduction is also inhibited. In the presence of CO2, the hydrogen evolution wave increases due to additional reduction of H2CO3,16 leading to the consumption of CO2 (upward directed band at 2343 cm−1, Figure 5c) as well as the formation of HCO3− (appearance of two downward bands at 1361 and 1303 cm−1 during the electrolysis at −0.39 V vs SHE, Figure 5c) and H2, but in contrast to the behavior described above with acetic acid, no additional anodic wave is observed (Figure 6). Note also that pyridine adsorption also largely prevents Pt oxide formation (comparison of Figures 6 and 4a). Comparison of anodic scans after short electrolysis on the hydrogen evolution wave with acetic acid or pyridinium under CO2 clearly shows a CO stripping wave with acetic acid 13926

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928

Article

Journal of the American Chemical Society

present investigation shows that this picture is not quite accurate. Our study confirms the lack of methanol formation upon bulk electrolysis of PyH+ solutions at a Pt electrode and provides a detailed account of the Faradaic yield for H2 production as a function of the electrode potential. However, our main finding consists of the observation that CO2 reduction in the presence of PyH+ as well as acetic acid is accompanied by a strong inhibition of the electrode process. Close combination of cyclic voltammetry and in situ infrared spectroscopy and constant comparison between PyH+ and AcOH (of almost equal pKa) were necessary to decipher the course of the reaction owing to the fact that the IR responses are perturbed by PyH+ adsorption. It finally appears that inhibition is caused by the reduction of CO2 to CO, whose high affinity with platinum triggers the formation of a Pt−CO film that prevents the reaction process. A paradoxical situation is thus faced in which the high affinity of Pt for CO helps to decrease the overpotential for the reduction of CO2 and therefore blocks the electrode, preventing the reaction process.

Figure 6. Cyclic voltammograms on a Pt disk electrode (7 mm diameter) of a 10 mM pyridinium/pyridine solution at pH 5.05 in the presence of 0.4 M K2SO4 at 0.1 V s−1. Red: under N2. Blue: under CO2.

whereas no feature is observed with pyridinium (Figure S4). The absence of CO detection is confirmed by in situ IR spectra recorded upon anodic scan following electrolysis at −0.18 or −0.39 V vs SHE under CO2 showing no band at 2050 cm−1 (Figure 5). The same measurements undertaken in 0.4 M K2SO4 in the absence of pyridinium/pyridine are given in Figure S5 for comparison. At first sight, these experiments would seem to indicate that there is no significant CO2 reduction on Pt electrode in the presence of pyridinium/ pyridine. However, the absence of CO detection with pyridinium by our in situ IR setup does not necessarily mean that CO is not formed. Indeed, as already mentioned, adsorption of pyridine strongly reduces the CO adsorption signal, as it also reduces the capacitive current and the observation of a PtO wave and prevents UPD wave observation although it does not completely suppress hydrogen evolution. Besides, more sensitive experiments employing in situ surfaceenhanced IR spectroscopy have shown adsorbed CO2 reduced species in the presence of pyridine/pyridinium.18b Hydrogenation of pyridinium has also been claimed to be observed recently and to be responsible for CO2 reduction via hydride transfer.25 One may wonder if such species could be responsible for electrode inhibition. If this were to be the case, inhibition should be specific to experiments in the presence of pyridinium, which is not the case. Besides, no sixelectron surface wave corresponding to pyridinium hydrogenation is detected in cyclic voltammetry. It seems more likely that a small amount of CO, neither clearly visible in our IR experiments nor as an anodic feature in CV owing to pyridine adsorption (which also prevents UPD and PtO wave observation), is formed and rapidly inhibits the electrode surface by blocking the few active sites still available at the initial stage of the process.



ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b08028. Additional figures and experimental section (PDF)



AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Hachem Dridi: 0000-0002-6191-6181 Kouakou Boniface Kokoh: 0000-0002-5379-7792 Cyrille Costentin: 0000-0002-7098-3132 Jean-Michel Savéant: 0000-0003-1651-3153 Notes

The authors declare no competing financial interest.



REFERENCES

(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15729. (2) Gray, H. B. Eng. Sci. 2008, 2, 26. (3) (a) Aresta, M.; Dibenedetto, A.; Angelini, A. J. CO2 Util. 2013, 3−4, 65. (b) Aresta, M.; Dibenedetto, A.; Angelini, A. Chem. Rev. 2014, 114, 1709. (4) Lamy, E.; Nadjo, L.; Savéant, J.-M. J. Electroanal. Chem. Interfacial Electrochem. 1977, 78, 403. (5) Fisher, B. J.; Eisenberg, R. J. Am. Chem. Soc. 1980, 102, 7361. (6) Hammouche, M.; Lexa, D.; Savéant, J.-M.; Momenteau, M. J. Electroanal. Chem. Interfacial Electrochem. 1988, 249, 347. (7) Windle, C. D.; Perutz, R. N. Coord. Chem. Rev. 2012, 256, 2562. (8) Schneider, J.; Jia, H.; Muckerman, J. T.; Fujita, E. Chem. Soc. Rev. 2012, 41, 2036. (9) (a) Costentin, C.; Robert, M.; Savéant, J.-M. Chem. Soc. Rev. 2013, 42, 2423. (b) Costentin, C.; Robert, M.; Savéant, J.-M. Acc. Chem. Res. 2015, 48, 2996. (c) Azcarate, I.; Costentin, C.; Robert, M.; Savéant, J.-M. J. Phys. Chem. C 2016, 120, 28951. (d) Azcarate, I.; Costentin, C.; Robert, M.; Savéant, J.-M. J. Am. Chem. Soc. 2016, 138, 16639. (10) Berardi, S.; Drouet, S.; Francas, L.; Gimbert-Surinach, C.; Guttentag, M.; Richmond, C.; Stoll, T.; Llobet, A. Chem. Soc. Rev. 2014, 43, 7501.



CONCLUSIONS After the initial excitement triggered by the report that a molecule as simple as protonated pyridine (PyH+) could catalyze the formation of methanol from the reduction of CO2 on a platinum electrode, further studies concluded, on preparative-scale and cyclic voltammetry bases, that in fact no methanol was formed and that the apparent catalytic currents actually reflected H2 evolution involving proton reduction from PyH+ and H2CO3 (generated by aquation of CO2). In other words, CO2 would not really be reduced but would merely indirectly participate in H2 evolution via proton reduction. The 13927

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928

Article

Journal of the American Chemical Society (11) Gottle, A. J.; Koper, M. T. M. Chem. Sci. 2017, 8, 458. (12) Ngo, K. T.; McKinnon, M.; Mahanti, B.; Narayanan, R.; Grills, D. C.; Ertem, M. Z.; Rochford, J. J. Am. Chem. Soc. 2017, 139, 2604. (13) Roy, S.; Sharma, B.; Pécaut, J.; Simon, P.; Fontecave, M.; Tran, P. D.; Derat, E.; Artero, V. J. Am. Chem. Soc. 2017, 139, 3685. (14) (a) Seshadri, G.; Lin, C.; Bocarsly, A. B. J. Electroanal. Chem. 1994, 372, 145−150. (b) Cole, E. B.; Lakkaraju, P. S.; Rampulla, D. M.; Morris, A. J.; Abelev, E.; Bocarsly, A. B. J. Am. Chem. Soc. 2010, 132, 11539. (15) (a) Stalder, C. J.; Chao, S.; Summers, D. P.; Wrighton, M. S. J. Am. Chem. Soc. 1983, 105, 6318. (b) Bookbinder, D. C.; Lewis, N. S.; Wrighton, M. S. J. Am. Chem. Soc. 1981, 103, 7656. (16) Costentin, C.; Canales, J. C.; Haddou, B.; Savéant, J.-M. J. Am. Chem. Soc. 2013, 135, 17671−17674. (17) (a) A more recent study,17b carried out at high CO2 pressure, reports the formation of a small amount of methanol but only at the very early stage of electrolysis. The Faradaic yield peaks at 6% and then decreases to almost zero as electrolysis continues. (b) Rybchenko, S. I.; Touhami, D.; Wadhawan, J. D.; Haywood, S. K. ChemSusChem 2016, 9, 1660. (18) (a) Hydrogen has also been reported as the major product in a recent study,18b but low overall Faradaic yields have been reported at high overpotential depending on the operating conditions.. (b) Dunwell, M.; Yan, Y.; Xu, B. ACS Catal. 2017, 7, 5410. (19) Cai, W.-B.; She, C.-X.; Ren, B.; Yao, J.-L.; Tian, Z.-W.; Tian, Z.Q. J. Chem. Soc., Faraday Trans. 1998, 94, 3127. (20) Conway, B. E.; Tilak, B. V. Electrochim. Acta 2002, 47, 3571. (21) (a) Giner, J. Electrochim. Acta 1963, 8, 857. (b) Marcos, M. L.; Gonzalez Velasco, J.; Hahn, F.; Beden, B.; Lamy, C.; Arvia, A. J. J. Electroanal. Chem. 1997, 436, 161. (22) Nikolic, B. Z.; Huang, H.; Gervasio, D.; Lin, A.; Fierro, C.; Adzic, R. R.; Yeager, E. B. J. Electroanal. Chem. Interfacial Electrochem. 1990, 295, 415. (23) Clavilier, J. In Electrochemical Surface Science: Molecular Phenomena at Electrode Surfaces; Soriago, M., Ed., ACS Publications: Washington, D.C., 1988; p 205. (24) Le Valant, A.; Comminges, C.; Can, F.; Thomas, K.; Houalla, M.; Epron, F. J. Phys. Chem. C 2016, 120, 26374. (25) Kronawitter, C. X.; Chen, Z.; Zhao, P.; Yang, X.; Koel, B. E. Catal. Sci. Technol. 2017, 7, 831.

13928

DOI: 10.1021/jacs.7b08028 J. Am. Chem. Soc. 2017, 139, 13922−13928