
==== Front
ACS Catal
ACS Catal
cs
accacs
ACS Catalysis
2155-5435
American Chemical Society

10.1021/acscatal.4c03521
Research Article
Time-Resolved Infrared Spectroscopic Evidence for Interfacial pH-Dependent Kinetics of Formate Evolution on Cu Electrodes
https://orcid.org/0000-0001-7254-4072
Katsoukis Georgios *
Heida Hilbert
Gutgesell Merlin
https://orcid.org/0000-0001-5898-6384
Mul Guido
Department of Chemical Engineering, MESA+ Institute for Nanotechnology, University of Twente Faculty of Science and Technology, Drienerlolaan 5, Enschede 7522 NB, The Netherlands
* E-mail: g.katsoukis@utwente.nl.
03 09 2024
20 09 2024
14 18 1386713876
14 06 2024
19 08 2024
19 08 2024
© 2024 The Authors. Published by American Chemical Society
2024
The Authors
https://creativecommons.org/licenses/by/4.0/ Permits the broadest form of re-use including for commercial purposes, provided that author attribution and integrity are maintained (https://creativecommons.org/licenses/by/4.0/).

By deployment of rapid-scan (second time scale) electrochemical FT-IR reflection–absorption spectroscopy, we studied the reduction of CO2 in 0.1 M Na2SO4 in deuterated water at a pD of 3.7. We report on the impact of dynamic changes in the bicarbonate equilibrium concentration in the vicinity of a polycrystalline Cu electrode, induced by step changes in applied electrode potential. We correlate these changes in interfacial composition and concentrations of dissolved species to the formation rate of formate, and provide evidence for the following conclusions: (i) the kinetics for the conversion of dissolved CO2 to formate (formic acid) are fast, (ii) bicarbonate is also converted to formate, but with less favorable kinetics, and (iii) carbonate does not yield any formate. These results reveal that formate formation requires (mildly) acidic conditions at the interface for CO2 to undergo a proton-coupled conversion step, and we postulate that bicarbonate reduction to formate is driven by catalytic hydrogenation via in situ formed H2. Interestingly CO was not observed, suggesting that the kinetics of the CO2 to CO reaction are significantly less favorable than formate formation under the experimental conditions (pH and applied potential). We also analyzed the feasibility of pulsed electrolysis to enhance the (average) rate of formation of formate. While a short positive potential pulse enhances the CO2 concentration, this also leads to the formation of basic copper carbonates, resulting in electrode deactivation. These observations demonstrate the potential of rapid-scan EC-IRRAS to elucidate the mechanisms and kinetics of electrochemical reactions, offering valuable insights for optimizing catalyst and electrolyte performance and advancing CO2 reduction technologies.

electrochemical CO2 reduction
Cu electrodes
formate selectivity
FT-IR reflection−absorption spectroscopy
interfacial pH
Nederlandse Organisatie voor Wetenschappelijk Onderzoek 10.13039/501100003246 ECCM.TT.ECCM.004 Erasmus+ 10.13039/501100010790 NA document-id-old-9cs4c03521
document-id-new-14cs4c03521
ccc-price
==== Body
pmcIntroduction

Many research groups around the world focus on the development of CO2 storage and/or utilization solutions and technologies to mitigate climate change.1 Fortunately, the rapid growth in the generation of electricity by wind turbines or solar cells significantly reduces emissions of CO2. Moreover, CO2 can now be considered a feedstock for hydrocarbons, e.g., by utilization of electrochemistry (the CO2 reduction reaction, CO2RR).2

The CO2RR has been widely explored for a large variety of electrode materials in the past six decades, revealing trends and patterns, indicating that CO2 activation and C–C bond formation are not only intrinsically influenced by the composition, morphology, and interfacial structure of the electrode at both the nano- and mesoscales3 but also by a complex interplay between intrinsic electrode properties and extrinsic effects such as the dynamic composition of the interface and electrical double layer near the electrode surface.4−6 The electrical double layer is strongly influenced by electrolyte parameters such as pH, buffer strength, nature of cations/anions, and mass transport limitations.7−9

Independent of the electrode material, the thermodynamics and kinetics of both the CO2RR and the hydrogen evolution reaction (HER) change with (interfacial) pH. For the HER, the proton reduction reaction (2H+ + 2e– ↔ H2) requires a lower overpotential than the water reduction reaction (2H2O + 2e– ↔ H2 + 2OH–), which poses an additional barrier for an O–H bond of water to dissociate.10 Therefore, the HER can usually be suppressed when operating at high pH. However, the CO2RR usually requires a series of proton-coupled electron transfer steps, while the proton coadsorption energy on the electrode is predicted to determine the pathway toward specific products such as formic acid, CO, or more deeply reduced species.11 In addition, the CO2RR is strongly influenced by a sequence of (interfacial) pH-dependent homogeneous reactions occurring at the interface12: 1) Hydration of CO2 to bicarbonate (HCO3– + H+) is slow and takes statistically around 27 s per CO2 molecule (khydration = 0.0371 s–1) at room temperature.

2) The reverse dehydration of HCO3– to CO2 is also slow at pH 7 (ca. 16 s per CO2), but fast at pH 4 (ca. 16 ms per CO2) (kdehydration = 2.67 × 107 g·mol–1·s–1).

3) Hydroxylation of CO2 to HCO3– (CO2 + OH– ↔ HCO3–) is effectively an irreversible second-order reaction and is relatively slow at pH 7 (ca. 200 s per CO2) and fast at pH 11 (ca. 20 ms per CO2) (khydroxylation = 2.23 × 106 g·mol–1·s–1; kdehydroxylation = 9.71 × 10–5 s–1, > 1 day per CO2 molecule). The reaction of HCO3– to CO2 will therefore proceed via dehydration.

4) Protolysis of HCO3– to carbonate (HCO3– ↔ CO32–+ H+) and, inversely, protonation occurs in the nano- to microsecond regime (kprotolysis = 3.06 × 105 s–1; kprotonation = 6 × 1012 g·mol–1·s–1).

These reactions primarily determine the time for the bicarbonate equilibrium to adjust to its local environment when it is out of equilibrium. Consequently, the time-scale and magnitude of the equilibrium shift will be determined not only by the rate of the HER (depending on applied potential) but also by the buffer capacity, ion effects, and mass transport – in particular advection – from the bulk electrolyte to the electrode interface. In addition, the CO2RR itself produces one hydroxide ion per electron transferred. The hydroxide ion produced at the interface will rapidly react with CO2 according to (3).13 A significant amount of interfacial CO2 will be depleted just by the buffering reactions, and increasing mass transport of CO2 to the electrode surface does not necessarily increase the CO2 reduction rate because of reaction (3).14

Understanding the complex interplay between spatiotemporally dependent acid–base bicarbonate equilibria and the intrinsic properties of electrode materials that undergo surface reconstruction is complicated. ATR-FTIR, Raman, and rotating ring disc electrode measurements have proven to be powerful techniques to obtain insights into the interfacial pH vs current densities (or applied potential) under a variety of CO2RR conditions at the electrode–electrolyte interface.15−18 Contrary to most studies advocating a high pH is necessary for effective conversion of CO2,3 in a systematic study that excludes the use of porous electrodes, a high faradaic efficiency of the CO2RR to CO on Au in a bicarbonate buffer was correlated to an interfacial pH needing to be relatively low, which is 6.8 in a CO2-saturated bicarbonate buffer.19 A low interfacial pH was able to be maintained using suitable cations (Cs+ > K+ > Na+ > Li+), high mass transport, and a high buffer capacity.17,19−21 These studies have focused mainly on steady state conditions, while transients in interfacial composition, in particular on short time scales (seconds), have received considerably less attention, since these are spectroscopically more difficult to resolve. We now elucidate how preconditioning of the Cu electrode, applied potential, and, most importantly, the dynamic changes of local concentrations impact the formation rate of products, in particular formate. We use a polycrystalline Cu electrode pretreated with multiple oxidative and reductive CV scans to obtain a steady electrochemically active Cu surface that does not undergo drastic changes upon sweeping or stepping of the potential. We apply rapid-scan electrochemical FT-IR reflection–absorption spectroscopy (EC-IRRAS) in a CO2-purged aqueous Na2SO4 electrolyte at pD 3.7 under stagnant conditions. This allows the observation of changes within the electrical double and diffusion layer. Dissolved CO2 is now resolved by IRRAS to be the primary reactant to form formate, while mass transport limitations induce a concentration change in dissolved CO2 to (bi)carbonate and eventually carbonate, which have a significantly lower contribution in the evolution rate of formate.

Materials and Methods

Materials

A Cu disk insert (5.0 mm OD, 4.0 mm thick) for ChangeDisk RDE setups was purchased from Pine Research, and 1 and 0.3 mm alumina polishing suspensions, deagglomerated, were purchased from Allied High-Tech Products. A standard Ag/AgCl reference electrode (3 M) was purchased from redox.me. Na2SO4 (>99%) and D2O (99.95%) were purchased from Aldrich. CO2 (>99.99%) was purchased from Linde.

Electrochemical FT-IR Setup

A Bruker V80v was used equipped with an LN2 cooled medium-band (12000–600 cm–1) MCT detector, an MIR polarizer (KRS-5) inserted into an automatic polarizer rotational unit, which was set to p-polarization, an A530/V reflection unit for electrochemical cells, and a ZnSe transparent IR hemispherical crystal. Gold mirrors below the ZnSe crystal were set to an incident angle of 30 degrees, which results into another refraction at the ZnSe–water interface of around 70–80 degrees (depending on the wavenumber determined by the refractive index of water). This enables surface electric field enhancement through the “grazing angle”. A VersaSTAT 3 potentiostat was connected through BNC via a trigger connector box (Bruker, E525/Z) to allow triggering the potentiostat within the OPUS 3D software for synchronized IR vs I–V data accumulation. The resolution was set to 8 cm–1, and the aperture was set to 1.5 mm.

Electrochemical FT-IRRAS Measurements (EC-IRRAS)

Prior to any measurement, the Cu disk and ZnSe IR crystal were polished and cleaned in an ultrasonic bath with distilled water. Before inserting all tools and chemicals into the IR spectrometer they were thoroughly dried to minimize any H2O contamination, since all experiments were performed in D2O. The electrochemical cell compartment was filled with 4 mL of electrolyte (see below for more details on the electrolyte), and the working electrode was inserted into a PTFE change disk to prevent any interaction of the side-walls of the electrode and placed upside-down (Otto configuration) onto the ZnSe crystal and aligned until a healthy signal could be obtained through external reflection. This was usually expressed by having a “thin enough” electrolyte layer of around 1.5–2 μm in between the electrode and the ZnSe crystal. The reference and counter electrodes were inserted into the electrolyte on opposite sides of the working electrode, respectively. Before purging the electrolyte with CO2, we performed 50 cyclic voltammograms at 0.2 V/s between +0.3 and −0.5 V vs RHE to oxidize, and desorb any contaminants from the Cu electrode, and condition the surface to obtain a stable electrochemical response. Then, the working electrode was lifted and reinserted into the electrolyte for mixing purposes and to prevent an IR response from contaminants formed near the Cu electrode (very intense signals were observed that could be assigned to C–H in D2O). Two types of experiments were performed: staircase linear sweep voltammetry (SCLSV) and chronoamperometry, using potential steps. The ohmic resistance was determined via electrochemical impedance spectroscopy 15 times, and the average of ca. 1000 Ohm was taken for correction in all cases. The resulting spectra were processed via OriginPro and baseline corrected using asymmetric least-squares smoothing with an asymmetric factor of 0.01 for positive peaks or 0.99 for negative peaks, a threshold of 0.0001, a smoothing factor of 9, and 10 iterations.

Electrolyte Considerations

We used a 0.1 M Na2SO4 electrolyte that was set at pD 4 using concentrated H2SO4 (the amount of H+ in the solution is over 6 orders of magnitude smaller than D+). The electrolyte was purged for 30 min with CO2 until saturation, which resulted in a pD of around 3.7. We chose 0.1 M Na2SO4 at pD 3.7 saturated with CO2 over more conventional CO2-saturated 0.1 M KDCO3 (pD 6.8), so the evolution rate of the D-formate can be tracked over a wider pD range. The CO2 saturation concentration in a 0.1 M Na2SO4 solution was determined at pD 3.7 after continuous bubbling at 293 K and 1 atm (salinity = 14.204 g·kg–1), and amounts to 36.422 mM. This results into a bulk concentration of DCO3– of 0.037 mM, and negligible amounts of CO32–.22 The zero-point energy of D+ compared to H+ leads to elevated activation barriers and slower rates in the DER than in the HER. As a consequence, the quantitative results from our experiments might be different for experiments done under nondeuterated conditions. However, qualitatively, the observations will uphold as the concentration “trends” are similar in the HER and DER, both being promoted at lower pH(D).

Definition of the Term “Interfacial pD”

From Figure S2, we estimate that the distance between the Cu electrode and the ZnSe crystal is approximately 1.5 to 2 μm. Around 10% of the vibrational intensities originate from within 6 nm of the electrode and over 40% from within 50 nm of the electrode as approximated by calculations performed in Note S2. From this, it can be concluded that the interfacial pD is determined at an average distance of around 100 nm.

Results and Discussion

Staircase Linear Sweep Voltammetry (SCLSV)

To establish the methodology for determining the local pD before applying it to the rapid-scan measurements, we performed staircase linear sweep voltammetry for a CO2-saturated solution in 0.1 M Na2SO4 in D2O at an applied potential from +0.4 V to −1.1 V vs RHE using 50 mV steps, which led to an actual potential from +0.4 V to −0.6 V (Figure 1, right) after correcting for the pD change (−0.059 V pD) near the electrode interface.

Figure 1 Contour plot of the ΔIR spectra–log(Rt/R0) from +0.4 V to −0.6 V vs RHE (from bottom to top) in 50 mV steps using +0.5 V as a baseline. The step time was 128.1 s and 20 spectra (6.4 s/spectrum) were taken per step. The gray contours are for guidance. The scale bars are in ΔmOD and the intensity changes indicated by the change in color according to the legend – note the scale is different for various spectral ranges. The vibrational assignments (text) are color-coded per compound: dark blue is dissolved CO2, orange is dissolved D-bicarbonate, red is dissolved carbonate, green is D-formate, light blue is precipitated basic Cux(CO3)y(OD)z, and black are deuterated water and sulfate. The time vs applied (black) and corrected (red) potentials including error bars are shown on the right.

Figure 1 shows a contour plot of the in situ IRRA spectra taken over time. A summary of the observed frequencies is shown in Table 1. The current–voltage profile can be found in (Figure S1. The CO2 asymmetric stretch is located at 2342 cm–1. The aqueous DCO3– related bands are located at 841, 1031, 1365, 1628, and 2023 cm–1. The aqueous CO32– bands are at 880, (1050 cm–1 obscured) 1410, and 1769 cm–1. We also observed a broad band at around 1450 cm–1 evolving in the low overpotential regime from +0.0 V to −0.4 V vs RHE, which we assigned to the asymmetric carbonate stretch of basic Cux(CO3)y(OD)z – see below for further discussion. It is noteworthy that we did not observe any appreciable amount of CO2 reduction toward CO, despite of a noise level below 50 μ OD that allows us to time-trace the very weak C–OD stretch vibration of bicarbonate at 2023 cm–1 – discussion follows below. We can observe a weak shoulder appearing at 1560–1570 cm–1 at −0.6 V vs RHE, which we assign to dissolved deuterated formate (D-formate) – debate will follow further below.

Table 1 Assignment of Observed IR Bands

frequency (cm–1)	assignment	
841	δoop(CO3; DCO3–) aq.25	
880	δoop(CO3; CO32–) aq.	
979	νs(SO42–) aq.26	
1103, 1194	νas(SO42–) aq.26	
1217	δ(D2O) bend	
1365	νs(DCO3–) aq.25	
1404	νs(COO) D-formate monodentate tilted	
1410	νas(CO32–) aq.25	
1450	νas(CO32–) basic CuCO3 (Supporting Information)	
1560	νas(COO) D-formate monodentate tilted	
1572	νas(COO) D-formate aq.27	
1628	νas(DCO3–) aq.25	
1770	νcomb(CO32–) aq.25	
2023	ν(OD; DCO3–) aq.25	

The major spectral changes upon cathodic stepping can be traced back to(1) removal of SO42– (i.e., asymmetric stretch doublet at 1106 and 1196 cm–1) from the beam path due to negative charging of the electrode,

(2) dynamic changes in the absorption region of D2O (i.e., bending mode at 1208 cm–1) due to molecular realignment along the electric field, which changes the interactions with p-polarized light, paired with a strong perturbation of the water network,23 convoluted within dispersion effects due to the large thickness of the “water film” that would require a Kramers–Kronig analysis to determine the actual absorption – we therefore omit the analysis of spectral features near (deuterated) water absorption,24 and

(3) shifting of the CO2/DCO3–/CO32– equilibrium from CO2 over DCO3– (see strong positive intensities at 1628, 1365, and 841 maximizing at around 55 min) to CO32– (at 880 and 1410 cm–1, strong and still rising at 65 min) that originates from the reduction of protons (D+).

The results in changing bicarbonate and carbonate concentrations are largely in agreement with data reported based on the calculations9 and ATR-IR spectra28 of former studies by our group.

The carbonate/bicarbonate system has been successfully used as a pH indicator for infrared spectroscopy.29 By tracing the intensity of the asymmetric stretch of DCO3– at 1628 cm–1 and the out-of-plane bend of CO32– at 880 cm–1 (see Figure 2a), we can determine the molar fractions of CO2, DCO3–, and CO32– shown in Figure 2b. This, in turn, enables us to determine the local pD using the Henderson–Hasselbalch (HH) equation, see Figure 2c. See Note S1 for more details on the calculations. The error bars become large at pD values of above 10. Knowing the pDt as a function of time allows us to correct the overpotential via ERHE,corrected = ERHE,applied – 0.059 pDt – iR (R ≈ 1kΩ; due to the low currents, the ohmic resistance correction is almost negligible (<20 mV) even at a high overpotential). The HH equation can be applied because the equilibrium adjusts faster than the time-resolution of the experiment (see the Introduction section for the time constants). The local pD corresponds approximately to the average pD between the ZnSe crystal and the electrode, which has an estimated thickness of around 1.5 to 2.0 μm (see Figure S2), although we can expect some degree of signal enhancement of vibrational features closer to the surface, which shifts the midpoint closer to the electrode. Noteworthily, a change in local pD from around 4 to 5 occurs in the first 10 min in the potential regime from +0.4 V until +0.2 V vs RHE. We noticed that this initial pD change is sensitive to the electrode pretreatment conditions, attributed to the reduction of Cu2O to Cu (Cu2O + 2e– + 2D+ → Cu + D2O). The pD remains stable between 5 and 6 until the onset potential for the DER at 40 min is at around −0.4 V vs RHE. Then, the depletion of CO2 and formation of bicarbonate take place buffering the pD until the capacity is used up at 55 min, leading to a swift increase in pD. The pD then starts to be buffered through the bicarbonate/carbonate equilibrium at around a pKa2 of 10.97.29Figure 2e shows the pD vs applied potential and corrected potential, which matches well the order of magnitude of the pH determined near the surface of an electrode using a variety of complementary techniques such as nonfaradaic probe reactions.30 Interestingly, only when the buffer capacity of the CO2/bicarbonate equilibrium becomes negligible does the corrected potential start to deviate strongly from the applied potential. This explains why the current is temporarily plateauing in I vs E curves (appears as a small peak/shoulder at around −0.4 to −0.7 V vs RHE in the CV) in dependence of the buffer strength, i.e. bicarbonate concentration.31

Figure 2 a) Intensity of dissolved νas(CO2) at 2343 cm–1, νas(DCO3–) at 1628 cm–1, and δoop(CO3; CO32–) at 880 cm–1 over time. Note that the Lambert–Beer law does not apply to the CO2 band intensity due to the OD stretch of D2O between 2250 and 2750 cm–1, blocking almost quantitatively the IR light. b) Calculated molar fraction of CO2 (blue), DCO3– (orange), and CO32– (red) over time. c) Calculated near electrode pD over time and d) potential profile vs time (same as Figure 1, right). e) pD-corrected potential showing the typical Hendersson–Hasselbalch-alike sigmoidal pD change of a buffer solution at −0.5 V vs RHE.

Formate Evolution Vs Interfacial pD

Now that the spectral intensities have been correlated to potential-dependent local pD changes and concentrations of CO2, D-bicarbonate, and carbonate, respectively, we will discuss the time-dependent formation of formate in second-time resolution and correlate this to the main reacting species. We traced the evolution of D-formate over time (2.8 s per spectrum) using potential steps. Each experiment serves a purpose: in experiment (a), we step from +0.5 V to −0.9 V and probe the D-formate evolution starting from an oxidized Cu electrode; in (b), we step closely before the onset of the DER from −0.5 V to −0.9 V; in (c), we step from +0.5 V to −1.1 V toward a more negative potential; in (d), we step from +0.6 V to −0.6 V oxidative potential to the onset potential of D-formate; in (e), we apply a constant current profile. We applied different “pre”potentials to investigate the potential role of surface–adsorbate concentration on the initial kinetics once stepping toward a reductive potential. Figure 3 shows the rapid-scan infrared spectra for each experiment and the corresponding time traces at 840 cm–1 (bicarbonate), 880 cm–1 (carbonate), 1560 cm–1 (D-formate), and 2342 cm–1 (CO2).

Figure 3 Time-resolved EC-IRRA spectra from 0 to 100 s showing the depletion of aqueous CO2 (blue), and the formation of aqueous bicarbonate (orange), aqueous carbonate (red), and D-formate (green): a) +0.5 V to −0.9 V, b) −0.5 V to −0.9 V, c) −0.5 V to −1.1 V, d) +0.6 V to −0.6 V, and e) chronoamperometry at −0.02 mA·cm–2 showing also the applied potential vs time profile. The lines indicate the standard electrode potentials of Cu2+/Cu, H+/H2, and the onset potential for the formation of D-formate (i.e., +0.8 V). Note that the “absolute” CO2 band intensity is reduced (here, ca. 0.005) due to the D2O OD stretch blocking most of the IR light. The formation rate of D-formate is fast at high concentrations of CO2, slow at high concentrations of D-bicarbonate, and absent at high concentrations of carbonate. The I vs E curves are provided in Figure S3.

Interestingly, in all chronoamperometry experiments (a) to (d), the formation of D-formate initially shows a rapid increase (see the green curve in Figure 3), before slowing down significantly at around 7 s and continuing at a slow, fixed rate. This transient in the kinetic regime coincides with the decreasing trend in the concentration of CO2, mirrored by the formation of D-bicarbonate. We can observe three noteworthy regimes, in which the formation kinetics of D-formate are different:

(1) When CO2 is abundantly available at the initial reductive potential applied, the formation of formate is maximized, as seen in all kinetic traces (Figure 3a–d) of the 1560 cm–1 band. This implies the presence of a CO2 pathway toward formate.

(2) When CO2 is depleted and bicarbonate is abundantly available, the formation of formate continues at a slower rate, implying that CO2 is not mandatory for the formation of formate but can proceed via a bicarbonate pathway.

(3) When bicarbonate is depleted, the formation of formate comes to a halt.

Regime (1) is located between 0 and 7 s in Figure 3a–d. Regime (3) is only visible in Figure 3c,e, where the 841 cm–1 band approaches zero at around 80 s, causing the growth of the 1560 cm–1 band to stop. Regime (2) is the intermediate phase, where the bicarbonate concentration plateaus.

The formation rate of D-formate is fastest before the onset of DCO3–, which points to the fact that CO2 is directly involved in the formate evolution mechanism and goes hand in hand with a previously proposed proton-coupled electron-transfer step to (adsorbed) CO2. It is debated if the formation of formate can also proceed via a direct conversion of bicarbonate.32 We observe that the formate evolution is not hindered upon depletion of CO2 at the interface and still evolves in the presence of bicarbonate but at a much slower rate. This suggests that both the CO2 and bicarbonate pathways are plausible, with the CO2 pathway being kinetically, significantly more favorable. At the onset of the formation of carbonate, D-formate evolution comes to a complete halt, indicating that carbonate is not a likely precursor for the formation of formate at the potentials applied here. This fact remains valid despite measuring an average pD at a distance of 100 nm from the electrode surface. Since deviations in the actual surface pD by 1–3 orders of magnitude might be possible, we cannot quantify the concentration overpotentials and determine accurate kinetic parameters for the reactions. However, this does not alter the fundamental observation that the formate formation rate is initially higher when CO2 is abundantly available and continues to evolve even after CO2 is depleted. It is worth noting that the initial surface–adsorbate concentration, which we attempted to tune through applying different “pre”potentials, does not have an effect on the initial formation kinetics of formate.

To obtain a better understanding of the D-formate evolution kinetics, we need to understand quantitatively what the pD-corrected overpotential is near the interface to assess whether the faster formation kinetics are truly because of the CO2 or bicarbonate concentration at the interface, or can be traced back to a reduced thermodynamic driving force with increasing local pD. We plotted the formation rate of formate vs the pD and iR-corrected potential as shown in Figure 4. It can be seen that the deviation of the pD and iR-corrected potential is around 0.2 to 0.4 V less negative. We also observe that the overpotential vs RHE (red curve) is decreasing over time, reducing reaction rates. However, it is clearly noticeable that the D-formate formation rate (black curve) does not follow the same trend but shows, in all cases, a higher rate in the first 7 to 10 s. The initial fast rate, hence, is due to the abundant availability of near-surface CO2 in the initial stages of applying the reductive potential. The “prepotentials” applied, namely, +0.6 V (Figure 3d), +0.5 V (Figure 3a), and −0.5 V (Figure 3b,c), play an important role in ensuring that the initial concentration of CO2 is indeed maximized at the beginning. Figure 3e, for example, shows that formate evolution takes place when the HER is already ongoing, preventing the evolution of formate via the CO2 pathway, as it is already depleted by the shift in equilibrium. As a consequence, the equilibrium shift results into an increase in the Nernstian (concentration) overpotential for the reduction reaction. This concentration overpotential is indirectly connected to the interfacial pH, which governs the relative concentrations of CO2, bicarbonate, and carbonate. This supports the observations of the three kinetic regimes mentioned above, excluding an explanation via reduced overpotentials due to local pD shifts.33

Figure 4 pD-corrected applied potential (red) vs the D-formate formation rate extracted from the growth of the band over time (1st derivative) at 1560 cm–1 (black). a–d) Inset shows the applied potential step. The red horizontal line shows the applied potential. The deviation of the pD and iR-corrected potential is in the range of 0.2 to 0.4 V.

These results imply that the kinetics of the electrochemical conversion of CO2 to formate are favorable under acidic conditions (pH ∼ 4), under which the competing HER is extensive for the applied Cu surface. It is obvious that all CO2 to formate conversion electrocatalysts reported in the literature are sluggish HER catalysts, because they consequently can maintain an acidic local pD.11 In addition, the faster kinetics, when preconditioning the electrode oxidatively (>+0.2 V vs RHE), reveals a possible explanation why pulsed and alternating current profiles can increase the production rates, since this way a more or less acidic interfacial composition (and associated relatively high CO2 concentration) can be maintained.33

The underlying reason why CO2 predominantly converts to formate at acidic pH can be explained by taking a closer look at the band position of the asymmetric and symmetric stretch of the D-formate vibrations and the adsorption orientation. We assign the 1560 cm–1 band to the asymmetric stretch of the carboxy group of D-formate νas(COO) (see Figure 5a,b).27 The asymmetric stretch for bidentate D-formate is invisible in p-polarized FT-IRRAS due to a dipole moment parallel to the surface (see Figure 5c). The asymmetric stretch for monodentate adsorbed D-formate in turn is usually located between 1600 and 1650 cm–1, which is too high to assign this vibration to the 1560 cm–1 band.34 Tilted adsorbed35 and aqueous27 D-formate match the position well and is usually observed at multilayer coverages and heterogeneous surfaces. For the tilted-adsorbed case, the νs(COO) of D-formate is usually located at ca.35 1360–1400 cm–1 but due to part of the dipole moment being in parallel to the electrode surface, it is very weak (Figure 5d). Therefore, we cannot deconvolute the signal from the bicarbonate symmetric stretch (1368 cm–1). We did not observe the postulated (O,O)-bidentate-formate within the time resolution of our experiment. We identified the carbonate asymmetric stretch (aq) at 1404 cm–1, which originates from the higher surface pD within the Helmholtz layer than the near-surface pD we determined in our experiment (100 nm). Note that the influence of the C–H(D) on both formate-related COO stretching vibrations are small upon isotope labeling, which cannot be captured by the resolution of our experiment (8 cm–1).27 The band at 1310 cm–1 in H2O can be assigned to the COH bend of bicarbonate.

Figure 5 Deconvoluted EC-IRRA spectra after 1.7 s in H2O (a) and 6.7 s D2O (b) using +0.5 V vs RHE as a baseline. Orange are bicarbonate-related bands, red is carbonate, blue is CuCO3, and green is D-formate.27 c) Normalized dipole moment vector perpendicular to the surface vs tilt angle of formate, visualizing the allowed observable vibrational modes via IRRAS. d) Allowed vibrational modes (green), weekly allowed (yellow), and not allowed (red). Only the tilted adsorbed formate matches the observations. Note that for H2O, −0.7 V was applied to prevent the water bending mode from dominating the spectrum.

It has been suggested via computational work by the group of Goddard36 – supported by theoretical works by the groups of Rossmeisl and Strasser11 – that the electrochemical reaction pathway for CO2 toward HCOO– proceeds via physisorbed CO2 that reacts directly with a surface H* through a nucleophilic attack, which bypasses the initial activation of CO2 by the Cu surface. It was also speculated that this may be achievable by controlling the pH. The formation of a tilted-adsorbed formate suggests that CO2 hydrogenation indeed occurs through hydrogen adsorption on the electrode surface. The role of bicarbonate in the electrochemical CO2 reduction toward formate on Cu remains unclear. The direct catalytic hydrogenation of carbonate/bicarbonate in liquid phase via H2 is thermodynamically favored over the direct CO2 reduction,37 which is also favorable when using Pd catalysts.38 We speculate that the high H2 concentration in the vicinity of the electrode due to the HER may promote the hydrogenation of bicarbonate/carbonate. The slow kinetics of (bi)carbonate conversion on Cu may originate additionally from electrostatic repulsion from the electrode preventing an appreciable adsorption of bicarbonate and carbonate. More experimental work, similar to that applied for Pd electrocatalysts adding H2 or D2 to the CO2 feed gas, is needed to elucidate the bicarbonate reduction pathway on Cu.

CO Evolution

It is worth dedicating a discussion to the absence of product CO under the experimental conditions applied. The formation of surface adsorbed CO in its different configurations can usually be observed at 2020–2070 cm–1.28 Our signal-to-noise is in the 0.01 mOD range, which means CO can be observed in the submonolayer regime.39 We can also exclude any preadsorption of CO on the Cu electrode at oxidative preconditioning potentials. At first glance, this is a surprising result considering the clear observation of CO in ATR-FTIR and Raman experiments.28,40−43 We observe two main differences in our experimental setup compared to what is commonly used in the literature: 1) in ATR-FTIR and Raman, a surface-roughened Cu electrode is used to be able to observe any bands through surface plasmon enhancement; 2) these methodologies are operated in the Kretschmann configuration which allows for significant mass transport toward the electrode by continuous bubbling of CO2 and forced advection. We exclude the surface roughening being responsible for the observation of CO compared to a polycrystalline Cu electrode, since it does not justify the absence of any CO in the 0.01 mOD regime. Instead, the swift pH change at the electrode interface in our configuration upon applying potentials more negative than −0.4 V vs RHE limits the availability of CO2 at the interface. From the experimental results, it is apparent that formate formation is very rapid, likely much more rapid than formation of CO, which does not instantly evolve, and likely requires constant supply of CO2 toward the interface. This coincides with the observation that CO poisoning of the HER is an important mechanism by which Cu maintains a high Faradaic efficiency for the CO2RR in aqueous electrolytes.44 As reported in the literature, at electrolyte flow rates of 2 mL/min CO appears after 7 s upon applying a potential of −0.7 V vs RHE, which is around the time CO2 is depleted in our configuration.40 Operating EC-IRRAS under a flow configuration is a challenging task that requires methodology development. We are currently testing the feasibility of microfluidic flow and radial flow cells.45

Fast Formation of Basic Cu Carbonate Degrades the Cu Electrode in Pulsed Electrolysis

We performed electrolysis using a square wave potential (3 loops) of 130 s at +0.5 V and −0.7 V vs RHE, with the idea to acidify the interface temporarily and eventually increase the yield of D-formate (see Figure 6). During the reductive applied potential, we observe the same features as in our previous chronoamperometric measurements. Upon reapplying +0.5 V, we can also observe that we are reacidifying our interface due to the recovery of the CO2 signal at 2342 cm–1. Unfortunately, we also reoxidize D-formate whenever we apply a positive potential of +0.5 V, which can be explained by the long residence time of D-formate in the vicinity of the Cu electrode in the stagnant electrolyte. In addition, the yield decreased with each loop. The only band that steadily increases is the band at 1457 cm–1, which we can assign to the asymmetric stretch (νas(CO3, CuCO3)) of an amorphous malachite analogue that resembles the vibrational features of the mineral georgeite Cu2(CO3)(OH)2·6H2O.46 Through mixing of an aqueous CuSO4 solution with KHCO3 in water, we observe the formation of a Cu carbonate precipitate with a vibrational feature at 1465 cm–1 (see Figure S4). Basic Cu carbonate species at the interface have previously been observed as malachite- and azurite-type basic Cu carbonate. This also precipitates after long-term (multiple hours to days) or pulsed electrolysis experiments, due to Cu corrosion.47 In fact, we observed the evolution of this broad band at 1457 cm–1 within a few seconds. The pKa of the reaction Cu2+ + HCO3– → CuCO3 + H+ is 3.5, which means that Cu carbonates form also in the presence of bicarbonate at pH 3.5 or higher following the reaction Cu2+ + HCO3– + 2H2O → Cu2(CO3)(OH)2(·xH2O) + 3H+.48 Alternating current electrolysis demonstrates, on the one hand, a fast degradation pathway of Cu due to the dissolution of Cu ions and reaction with bicarbonate toward basic Cu carbonate. On the other hand, we do not observe associated formation of formate. It is important to mention that formation of formate by conversion of basic Cu carbonate has been demonstrated recently by Lopez et al.,49 allbeit with a low Faradaic Efficiency of only 1%. Therefore, under our experimental conditions, the formation of formate will be insignificant.

Figure 6 a) Applied square wave potential over time: 130 s −0.7 V vs RHE and 130 s + 0.5 V vs RHE (3 loops). b) EC-IRRAS spectra from 0 to 30 s taking point zero as the beginning of −0.7 V for each loop. c) Time traces for the evolution of the basic Cu carbonate (light blue, 1457 cm–1), carbonate (red, 880 cm–1), bicarbonate (orange, 841 cm–1), CO2 (dark blue, 2342 cm–1), and D-formate (green, 1560 cm–1).

Conclusion

In conclusion, we successfully probed the electrochemical formation kinetics of D2 and D-formate with respect to the near Cu-electrode pD in a CO2 saturated aqueous (D2O) electrolyte via time-resolved electrochemical FT-IR reflection–absorption spectroscopy using p-polarized light. We used the bicarbonate and carbonate-related IR bands as an indicator to determine the local pD. Three interfacial regimes can be distinguished in which the formation kinetics of D-formate vary. The formation rate of D-formate is fastest at high abundance of CO2 at the interface, which points to the fact that CO2 is directly involved in the formate evolution mechanism. We observe that the formate evolution is not hindered upon depletion of CO2 at the interface and still evolves in the presence of bicarbonate, although at a much slower rate. This suggests that both the CO2 and bicarbonate pathway are plausible, with the CO2 pathway being kinetically more favorable whereas carbonate does not yield any D-formate at the applied potentials. The formation of a tilted-adsorbed formate suggests that the CO2 hydrogenation occurs through hydrogen adsorbed on the electrode surface. The conversion of CO2 to formate is therefore most favorable when the interfacial pH is acidic, likely around pH 4, to maximize CO2 concentration and minimize H+ concentration (see Figure S7 for more details on the hypothesis). The slow rate of bicarbonate conversion to formate suggests that hydrogenation of bicarbonate occurs catalytically (after formation of H2), but this requires further study.

The absence of any observation of CO can be traced back to the likely much slower kinetics in forming CO than that of formate under the applied reaction conditions. The formation of Cu carbonate has been demonstrated and is a possible degradation mechanism for Cu electrodes applied under experimental conditions for pulsed electrolysis.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscatal.4c03521.Method description for local pH determination (Note S1); estimation of the distance from the electrode (Note S2); scheme of the spectroelectrochemical cell (Scheme S1); iR correction (Figure S1); quantification of D2O bending mode (Figure S2); I–V curves (Figure S3); FTIR spectrum of CuSO4 (Figure S4); quantification of the bicarbonate 1627 cm–1 peak (Figure S5); signal amplification vs distance from the electrode (Figure S6); hypothesis that a pH of around 4 may be ideal (Figure S7) (PDF)

Supplementary Material

cs4c03521_si_001.pdf

Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

This work was funded by the Dutch Research Council (NWO) (ECCM.TT.ECCM.004) and the European Commission (Erasmus+).

The authors declare no competing financial interest.

Acknowledgments

We thank Prof. Jennifer Strunk for enabling an exchange via the Erasmus+ program.
==== Refs
References

Friedlingstein P. ; O’Sullivan M. ; Jones M. W. ; Andrew R. M. ; Bakker D. C. E. ; Hauck J. ; Landschützer P. ; Le Quéré C. ; Luijkx I. T. ; Peters G. P. ; et al. Global Carbon Budget 2023. Earth Syst. Sci. Data 2023, 15 (12 ), 5301–5369. 10.5194/ESSD-15-5301-2023.
Daaboul J. ; Moriarty P. ; Honnery D. Net Green Energy Potential of Solar Photovoltaic and Wind Energy Generation Systems. J. Cleaner Prod. 2023, 415 , 137806 10.1016/J.JCLEPRO.2023.137806.
Nitopi S. ; Bertheussen E. ; Scott S. B. ; Liu X. ; Engstfeld A. K. ; Horch S. ; Seger B. ; Stephens I. E. L. ; Chan K. ; Hahn C. ; et al. Progress and Perspectives of Electrochemical CO2 Reduction on Copper in Aqueous Electrolyte. Chem. Rev. 2019, 119 (12 ), 7610–7672. 10.1021/ACS.CHEMREV.8B00705.31117420
Marcandalli G. ; Monteiro M. C. O. ; Goyal A. ; Koper M. T. M. Electrolyte Effects on CO2 Electrochemical Reduction to CO. Acc. Chem. Res. 2022, 55 (14 ), 1900–1911. 10.1021/acs.accounts.2c00080.35772054
de Salles Pupo M. M. ; Kortlever R. Electrolyte Effects on the Electrochemical Reduction of CO2. ChemPhyschem 2019, 20 (22 ), 2926–2935. 10.1002/cphc.201900680.31600018
Seh Z. W. ; Kibsgaard J. ; Dickens C. F. ; Chorkendorff I. ; No̷rskov J. K. ; Jaramillo T. F. Combining Theory and Experiment in Electrocatalysis: Insights into Materials Design. Science 2017, 355 (6321 ), eaad4998 10.1126/science.aad4998.28082532
Varela A. S. ; Kroschel M. ; Reier T. ; Strasser P. Controlling the Selectivity of CO2 Electroreduction on Copper: The Effect of the Electrolyte Concentration and the Importance of the Local PH. Catal. Today 2016, 260 , 8–13. 10.1016/j.cattod.2015.06.009.
Marcandalli G. ; Goyal A. ; Koper M. T. M. Electrolyte Effects on the Faradaic Efficiency of CO2 Reduction to CO on a Gold Electrode. ACS Catal. 2021, 11 (9 ), 4936–4945. 10.1021/acscatal.1c00272.34055454
Kas R. ; Kortlever R. ; Yılmaz H. M. ; Koper M. T. ; Mul G. Manipulating the Hydrocarbon Selectivity of Copper Nanoparticles in CO2 Electroreduction by Process Conditions. ChemElectrochem 2015, 2 (3 ), 354–358. 10.1002/celc.201402373.
Zheng Y. ; Jiao Y. ; Vasileff A. ; Qiao S. Z. The Hydrogen Evolution Reaction in Alkaline Solution: From Theory, Single Crystal Models, to Practical Electrocatalysts. Angew. Chem., Int. Ed. 2018, 57 (26 ), 7568–7579. 10.1002/anie.201710556.
Bagger A. ; Ju W. ; Varela A. S. ; Strasser P. ; Rossmeisl J. Electrochemical CO2 Reduction: A Classification Problem. ChemPhyschem 2017, 18 (22 ), 3266–3273. 10.1002/cphc.201700736.28872756
Schulz K. G. ; Riebesell U. ; Rost B. ; Thoms S. ; Zeebe R. E. Determination of the Rate Constants for the Carbon Dioxide to Bicarbonate Inter-Conversion in pH-Buffered Seawater Systems. Mar. Chem. 2006, 100 (1–2 ), 53–65. 10.1016/j.marchem.2005.11.001.
Singh M. R. ; Clark E. L. ; Bell A. T. Effects of Electrolyte, Catalyst, and Membrane Composition and Operating Conditions on the Performance of Solar-Driven Electrochemical Reduction of Carbon Dioxide. Phys. Chem. Chem. Phys. 2015, 17 (29 ), 18924–18936. 10.1039/C5CP03283K.26103939
Goyal A. ; Marcandalli G. ; Mints V. A. ; Koper M. T. M. Competition between CO2 Reduction and Hydrogen Evolution on a Gold Electrode under Well-Defined Mass Transport Conditions. J. Am. Chem. Soc. 2020, 142 (9 ), 4154–4161. 10.1021/jacs.9b10061.32041410
Henckel D. A. ; Counihan M. J. ; Holmes H. E. ; Chen X. ; Nwabara U. O. ; Verma S. ; Rodríguez-López J. ; Kenis P. J. A. ; Gewirth A. A. Potential Dependence of the Local pH in a CO2 Reduction Electrolyzer. ACS Catal. 2021, 11 (1 ), 255–263. 10.1021/acscatal.0c04297.
Ayemoba O. ; Cuesta A. Spectroscopic Evidence of Size-Dependent Buffering of Interfacial pH by Cation Hydrolysis during CO2 Electroreduction. ACS Appl. Mater. Interfaces 2017, 9 (33 ), 27377–27382. 10.1021/acsami.7b07351.28796478
Singh M. R. ; Kwon Y. ; Lum Y. ; Ager J. W. ; Bell A. T. Hydrolysis of Electrolyte Cations Enhances the Electrochemical Reduction of CO2 over Ag and Cu. J. Am. Chem. Soc. 2016, 138 (39 ), 13006–13012. 10.1021/jacs.6b07612.27626299
Kas R. ; Ayemoba O. ; Firet N. J. ; Middelkoop J. ; Smith W. A. ; Cuesta A. In-Situ Infrared Spectroscopy Applied to the Study of the Electrocatalytic Reduction of CO2: Theory, Practice and Challenges. ChemPhyschem 2019, 20 (22 ), 2904–2925. 10.1002/cphc.201900533.31441195
Liu X. ; Monteiro M. C. O. ; Koper M. T. M. Interfacial pH Measurements during CO2 Reduction on Gold Using a Rotating Ring-Disk Electrode. Phys. Chem. Chem. Phys. 2023, 25 (4 ), 2897–2906. 10.1039/D2CP05515E.36633182
Monteiro M. C. O. ; Dattila F. ; Hagedoorn B. ; García-Muelas R. ; López N. ; Koper M. T. M. Absence of CO2 Electroreduction on Copper, Gold and Silver Electrodes without Metal Cations in Solution. Nat. Catal. 2021, 4 (8 ), 654–662. 10.1038/s41929-021-00655-5.
Kyriacou G. Z. ; Anagnostopoulos A. K. Influence CO2 Partial Pressure and the Supporting Electrolyte Cation on the Product Distribution in CO2 Electroreduction. J. Appl. Electrochem. 1993, 23 (5 ), 483–486. 10.1007/BF00707626.
CO2 in Seawater: Equilibrium, Kinetics, Isotopes, Zeebe R. E. ; Wolf-Gladrow D. , eds.; Elsevier, 2001. DOI: 10.1016/S0422-9894(01)80007-6.
Cassone G. ; Sponer J. ; Trusso S. ; Saija F. Ab Initio Spectroscopy of Water under Electric Fields. Phys. Chem. Chem. Phys. 2019, 21 (38 ), 21205–21212. 10.1039/C9CP03101D.31368466
Yamamoto K. ; Ishida H. Optical Theory Applied to Infrared Spectroscopy. Vib. Spectrosc. 1994, 8 (1 ), 1–36. 10.1016/0924-2031(94)00022-9.
Davis A. R. ; Oliver B. G. A Vibrational-Spectroscopic Study of the Species Present in the CO2-H2O System. J. Solution Chem. 1972, 1 (4 ), 329–339. 10.1007/BF00715991.
Mabrouk K. B. ; Kauffmann T. H. ; Aroui H. ; Fontana M. D. Raman Study of Cation Effect on Sulfate Vibration Modes in Solid State and in Aqueous Solutions. J. Raman Spectrosc. 2013, 44 (11 ), 1603–1608. 10.1002/JRS.4374.
Rotzinger F. P. ; Kesselman-Truttmann J. M. ; Hug S. J. ; Shklover V. ; Grätzel M. Structure and Vibrational Spectrum of Formate and Acetate Adsorbed from Aqueous Solution onto the TiO2 Rutile (110) Surface. J. Phys. Chem. B 2004, 108 (16 ), 5004–5017. 10.1021/jp0360974.
Moradzaman M. ; Mul G. Infrared Analysis of Interfacial Phenomena during Electrochemical Reduction of CO2 over Polycrystalline Copper Electrodes. ACS Catal. 2020, 10 (15 ), 8049–8057. 10.1021/acscatal.0c02130.
Baldassarre M. ; Barth A. The Carbonate/Bicarbonate System as a pH Indicator for Infrared Spectroscopy. Analyst. 2014, 139 (9 ), 2167–2176. 10.1039/c3an02331a.24622696
Ryu J. ; Wuttig A. ; Surendranath Y. Quantification of Interfacial pH Variation at Molecular Length Scales Using a Concurrent Non-Faradaic Reaction. Angew. Chem., Int. Ed. 2018, 57 (30 ), 9300–9304. 10.1002/anie.201802756.
Marcandalli G. ; Villalba M. ; Koper M. T. M. The Importance of Acid-Base Equilibria in Bicarbonate Electrolytes for CO2 Electrochemical Reduction and CO Reoxidation Studied on Au(Hkl) Electrodes. Langmuir 2021, 37 (18 ), 5707–5716. 10.1021/acs.langmuir.1c00703.33913319
Lee C. W. ; Cho N. H. ; Im S. W. ; Jee M. S. ; Hwang Y. J. ; Min B. K. ; Nam K. T. New Challenges of Electrokinetic Studies in Investigating the Reaction Mechanism of Electrochemical CO2 Reduction. J. Mater. Chem. A 2018, 6 (29 ), 14043–14057. 10.1039/C8TA03480J.
Obasanjo C. A. ; Gao G. ; Crane J. ; Golovanova V. ; García de Arquer F. P. ; Dinh C. T. High-Rate and Selective Conversion of CO2 from Aqueous Solutions to Hydrocarbons. Nat. Commun. 2023, 14 (1 ), 1–12. 10.1038/s41467-023-38963-y.36596776
Shiozawa Y. ; Koitaya T. ; Mukai K. ; Yoshimoto S. ; Yoshinobu J. Quantitative Analysis of Desorption and Decomposition Kinetics of Formic Acid on Cu(111): The Importance of Hydrogen Bonding between Adsorbed Species. J. Chem. Phys. 2015, 143 (23 ), 234707 10.1063/1.4937414.26696070
Sim W. S. ; Gardner P. ; King D. A. Multiple Bonding Configurations of Adsorbed Formate on Ag{111}. J. Phys. Chem. 1996, 100 (30 ), 12509–12516. 10.1021/jp9600726.
Cheng T. ; Xiao H. ; Goddard W. A. Reaction Mechanisms for the Electrochemical Reduction of CO2 to CO and Formate on the Cu(100) Surface at 298 K from Quantum Mechanics Free Energy Calculations with Explicit Water. J. Am. Chem. Soc. 2016, 138 (42 ), 13802–13805. 10.1021/jacs.6b08534.27726392
Moret S. ; Dyson P. J. ; Laurenczy G. Direct Synthesis of Formic Acid from Carbon Dioxide by Hydrogenation in Acidic Media. Nat. Commun. 2014, 5 (1 ), 1–7. 10.1038/ncomms5017.
Blom M. J. W. ; van Swaaij W. P. M. ; Mul G. ; Kersten S. R. A. Mechanism and Micro Kinetic Model for Electroreduction of CO2 on Pd/C: The Role of Different Palladium Hydride Phases. ACS Catal. 2021, 11 (12 ), 6883–6891. 10.1021/acscatal.1c01325.
Zhang D. ; Jansen C. ; Berg O. T. ; Bakker J. M. ; Meyer J. ; Kleyn A. W. ; Juurlink L. B. F. RAIRS Characterization of CO and O Coadsorption on Cu(111). J. Phys. Chem. C 2022, 126 (31 ), 13114–13121. 10.1021/acs.jpcc.2c02541.
An H. ; Wu L. ; Mandemaker L. D. B. ; Yang S. ; de Ruiter J. ; Wijten J. H. J. ; Janssens J. C. L. ; Hartman T. ; van der Stam W. ; Weckhuysen B. M. Sub-Second Time-Resolved Surface-Enhanced Raman Spectroscopy Reveals Dynamic CO Intermediates during Electrochemical CO2 Reduction on Copper. Angew. Chem., Int. Ed. 2021, 60 (30 ), 16576–16584. 10.1002/anie.202104114.
Xu H. ; Fan Z. ; Zhu S. ; Shao M. A Minireview on Selected Applications of in Situ Infrared Spectroscopy in Studying CO2 Electrochemical Reduction Reaction. Curr. Opin. Electrochem. 2023, 41 , 101363 10.1016/j.coelec.2023.101363.
Zhu S. ; Li T. ; Cai W.-B. ; Shao M. CO2 Electrochemical Reduction As Probed through Infrared Spectroscopy. ACS Energy Lett. 2019, 4 (3 ), 682–689. 10.1021/ACSENERGYLETT.8B02525.
Dunwell M. ; Lu Q. ; Heyes J. M. ; Rosen J. ; Chen J. G. ; Yan Y. ; Jiao F. ; Xu B. The Central Role of Bicarbonate in the Electrochemical Reduction of Carbon Dioxide on Gold. J. Am. Chem. Soc. 2017, 139 (10 ), 3774–3783. 10.1021/jacs.6b13287.28211683
Ooka H. ; Figueiredo M. C. ; Koper M. T. M. Competition between Hydrogen Evolution and Carbon Dioxide Reduction on Copper Electrodes in Mildly Acidic Media. Langmuir 2017, 33 (37 ), 9307–9313. 10.1021/acs.langmuir.7b00696.28453940
Cychy S. ; Hiltrop D. ; Andronescu C. ; Muhler M. ; Schuhmann W. Operando Thin-Layer ATR-FTIR Spectroelectrochemical Radial Flow Cell with Tilt Correction and Borehole Electrode. Anal. Chem. 2019, 91 (22 ), 14323–14331. 10.1021/acs.analchem.9b02734.31609106
Kondrat S. A. ; Smith P. J. ; Wells P. P. ; Chater P. A. ; Carter J. H. ; Morgan D. J. ; Fiordaliso E. M. ; Wagner J. B. ; Davies T. E. ; Lu L. ; Bartley J. K. ; Taylor S. H. ; Spencer M. S. ; Kiely C. J. ; Kelly G. J. ; Park C. W. ; Rosseinsky M. J. ; Hutchings G. J. Stable Amorphous Georgeite as a Precursor to a High-Activity Catalyst. Nature 2016, 531 (7592 ), 83–87. 10.1038/nature16935.26878237
Jiang S. ; D’Amario L. ; Dau H. Copper Carbonate Hydroxide as Precursor of Interfacial CO in CO2 Electroreduction. ChemSuschem 2022, 15 (8 ), e202102506 10.1002/cssc.202102506.35289108
Stiff M. J. Copper/Bicarbonate Equilibria in Solutions of Bicarbonate Ion at Concentrations Similar to Those Found in Natural Water. Water Res. 1971, 5 (5 ), 171–176. 10.1016/0043-1354(71)90049-2.
Ma H. ; Ibáñez-Alé E. ; Ganganahalli R. ; Pérez-Ramírez J. ; López N. ; Yeo B. S. Direct Electroreduction of Carbonate to Formate. J. Am. Chem. Soc. 2023, 145 , 24707–24716. 10.1021/jacs.3c08079.37924283
