Promoting selective electrochemical CO₂ reduction under unconventionally acidic conditions through secondary coordination sphere positioning

Abstract

While several studies have investigated the effects of protic secondary coordination sphere (SCS) groups on the kinetics of iron tetraphenylporphyrin (FeTPP) catalysed CO₂ reduction, few have examined how a protic SCS might alter reaction selectivity. Under mildly acidic conditions, FeTPP-based catalysts are selective towards the 2e⁻/2H⁺ reduction of CO₂ to CO; however, in the presence of more acidic proton donors, indiscriminate proton transfers often result in parasitic H₂ evolution. This report investigates how SCS amide positioning alters CO versus H₂ selectivity during CO₂ reduction with a series of four FeTPP isomers bearing SCS amides at varying positions around the porphyrin core: NH donors are placed at either the meta or ortho position of the meso aryl porphyrin ring, as well as proximal (closer) or distal (farther away) to the porphyrin plane. In the presence of a conventional, weakly acidic proton source (phenol; pK_a = 29.2 in MeCN), all isomers display the expected high faradaic efficiency (FE) towards CO (FE_CO = 67–85%) along with minimal H₂ evolution (FE_H2 = 3–13%). With a significantly stronger acid (3,5-bis(trifluoromethyl)phenol; pK_a = 23.8 in MeCN), H₂ becomes the major product when using the ortho-distal or both meta isomers (FE_H2 = 45–65%) as well as unfunctionalized FeTPP (FE_H2 = 78%). Importantly, the ortho-proximal isomer shows dramatically rescued CO selectivity under these unconventionally acidic conditions (FE_CO = 83 ± 4%). These results show how proper SCS placement impacts reaction selectivity during CO₂ reduction, particularly with respect to minimizing indiscriminate proton transfers that lead to undesirable reactivity.

---

1. Introduction

Electrochemical CO₂ reduction offers a sustainable strategy to convert this inert molecule into value-added products, but significant kinetic barriers hinder efficient and selective CO₂ conversion.^1,2 While many transition metal electrocatalysts show promising reactivity,^3,4 efforts to improve their activity through primary coordination sphere modifications often follow electronic scaling relationships, wherein more negative reduction potentials are required to achieve faster catalytic turnover.^5,6 Taking inspiration from enzymes,^7–9 these scaling relationships may be ‘broken’ through modification of the secondary coordination sphere (SCS). The SCS consists of functional groups in the catalyst periphery that can modulate the kinetics and thermodynamics of individual reaction steps, or change the reaction mechanism altogether.^6,10–14

During electrocatalytic CO₂ reduction, protic functional groups in the SCS may alter reaction barriers by stabilizing high-energy transition states and/or by relaying protons from an exogenous acid to bound substrate. Important contributions were made by Savéant and co-workers, who showed that modification of iron tetraphenylporphyrin (FeTPP) to include phenolic SCS donors substantially increases the kinetics of CO₂ reduction while maintaining high selectivity for CO.^15,16 Many subsequent investigations—including some from our laboratory—sought to understand how the kinetics of FeTPP-catalysed CO₂ reduction depend on protic SCS donor positioning,¹⁷ proton transfer driving forces,^18–22 reactant templating,²³ and altering the number of protic groups in the SCS.^24–26

Although the kinetic advantages of protic SCS groups have been frequently explored, very few studies have focused on how the SCS influences product selectivity during CO₂ reduction, especially with respect to SCS donor positioning. Typically, the selectivity of FeTPP catalysts is constrained to either the 2e⁻/2H⁺ reduction of CO₂ to CO or—under sufficiently acidic conditions—the background reduction of protons to form H₂. However, recent reports have suggested that protic SCS groups may beneficially alter the selectivity of this platform. One notable example from Robert and co-workers describes the challenging 8e⁻/8H⁺ reduction of CO₂ into CH₄ with a catalyst bearing pendent phenols, although this transformation was driven photocatalytically.²⁷ The Dey group reported that iron porphyrinoid catalysts bearing protic SCS groups alter selectivity between CO and HCOO⁻,²⁸ or enhance selectivity towards CO and against parasitic H₂ evolution.²⁹ Nevertheless, a full understanding of the numerous roles of the SCS is still emerging, and it remains unclear how an SCS group might be designed to achieve a desired selectivity change during FeTPP-catalysed CO₂ reduction.

When FeTPP catalysts are paired with highly acidic proton sources in solution, indiscriminate proton transfers often result in background H₂ evolution or catalyst degradation, rather than productive CO evolution.^30–32 For example, Costentin and Savéant reported that CO₂ reduction with [FeTPP]Cl using acetic acid (pK_a(DMSO) = 12.6,³³ pK_a(MeCN) = 23.5)³⁴ results in a FE_CO of only ∼30%, along with formation of a green catalyst byproduct hypothesized to be a hydrogenated porphyrinoid.³⁰ To avoid low CO faradaic efficiencies and/or catalyst decomposition, FeTPP-catalysed CO₂ reduction typically employs mild Brønsted acids, such as phenol (pK_a(DMSO) = 18.0,³⁵ pK_a(MeCN) = 29.2);³⁴ however, this limits catalytic activity as FeTPP systems generally exhibit faster kinetics with stronger acids.^18,30 In order to maintain high CO selectivity under highly acidic conditions, it is crucial for protons to be directed to the bound CO₂ substrate rather than to the metal or porphyrin ligand. Whether and how a protic SCS regulates the desired proton transfers during CO₂ reduction remains poorly understood.

Against this backdrop, we sought to establish design principles for a protic SCS that can improve selectivity towards CO and against parasitic H₂ evolution. We hypothesized that indiscriminate proton transfers—and their deleterious effects on selectivity—could be minimized with an SCS donor that is suitably positioned near the catalytic active site. Accordingly, we designed a series of FeTPP complexes bearing pendent amides at varying positions around the catalytic active site to investigate the effects of SCS donor positioning on the selectivity and activity of FeTPP-catalysed CO₂ reduction (Scheme 1). We report the synthesis and electrochemical investigation of four positional isomers having NH donors at either the meta or ortho position of the meso aryl porphyrin ring, as well as either proximal (closer) or distal (farther away) to the porphyrin plane. The CO₂ reduction selectivity and activity of each isomer is reported in MeCN electrolyte in the presence of two different exogenous acids: phenol as a more conventional Brønsted acid (pK_a = 29.2 in MeCN), and 3,5-bis(trifluoromethyl) phenol as a stronger proton donor (pK_a = 23.8 in MeCN).³⁴ Using phenol as the exogenous acid, proper SCS donor positioning increases the CO₂ reduction rate constant by 10-fold while only modestly affecting CO vs. H₂ selectivity. However, in the presence of the more acidic proton source, we observe marked differences in reaction selectivity between each isomer: when the NH of the amide is positioned at the ortho-proximal position, CO is produced as the major product at a high faradaic efficiency (FE) of 83 ± 4% and only minor quantities of H₂ are formed (FE_H2 = 7 ± 1%), while each of the remaining isomers display a low FE for CO (FE_CO ≤ 27%) and significant background H₂ evolution (FE_H2 ≥ 45%). Together, these results provide insight into how suitable SCS donor positioning can regulate proton transfer to the active site and effect beneficial changes in reaction selectivity.

Scheme 1 Ortho distal versus proximal positioning of the SCS donor alters selectivity between CO and H₂ in the presence of a strong exogenous acid.

2. Results and discussion

2.1 Design and synthesis of four amide functionalized porphyrin isomers

Building on our previous work with FeTPP catalysts bearing tuneable SCS amides,^17,18 we envisioned a series of four structurally analogous positional isomers bearing amide pendants at varying positions around the iron porphyrin core: Fe-ortho-proximal features the amide pendant at the ortho position of the meso aryl ring with the NH proximal (closer) to the porphyrin plane, while Fe-ortho-distal places the NH distal (farther away) from the active site. The effects of the latter isomer on activity and selectivity are of particular interest since previous reports on protic SCS groups have largely focused on donors positioned at the ortho-proximal position for the sake of synthetic practicality.^{15,18,24,36–39} We additionally prepared the two corresponding meta isomers (Fe-meta-distal and Fe-meta-proximal) to serve as control complexes. The meta isomers were selected as suitable controls over the para-substituted congeners because iron porphyrins functionalized with amide groups at the para-position of the meso porphyrin ring are known to display non-idealized behaviour under electrochemical conditions.^17,23

The synthetic details of each iron porphyrin isomer are outlined in Scheme 2. Fe-ortho-proximal was prepared by the method we have previously reported:¹⁸ first, 5-(ortho-nitrophenyl)-10,15,20(triphenyl) porphyrin, 1-ortho, is reduced by SnCl₂ in concentrated HCl to afford 5-(ortho-aminophenyl)-10,15,20(triphenyl) porphyrin, 2-ortho. Subsequent reaction of 2-ortho with triethylamine and in situ generated 4-trifluormethyl benzoyl chloride provides the ortho-proximal amide porphyrin, 3-ortho, which is metalated with FeCl₂·4H₂O to yield Fe-ortho-proximal. The corresponding meta isomer, Fe-meta-proximal, was synthesized according to the same procedure by using the meta-nitrophenyl porphyrin, 1-meta, as starting material. To prepare the distal isomer, Fe-ortho-distal, 5-(methyl ortho-benzoate)-10,15,20(triphenyl) porphyrin, 4-ortho, is first hydrolysed in refluxing 1,4-dioxane and aqueous KOH to afford 5-(ortho-carboxyphenyl)-10,15,20(triphenyl) porphyrin, 5-ortho. This intermediate is treated with SOCl₂ and triethylamine before the addition of 4-trifluoromethyl aniline to yield the ortho-distal amide functionalized porphyrin, 6-ortho. Metalation of freebase 6-ortho using FeCl₂·4H₂O provides Fe-ortho-distal. The corresponding meta isomer, Fe-meta-distal, was prepared identically by starting from the methyl meta-benzoate porphyrin, 4-meta. All porphyrin ligands were fully characterized by ¹H, ¹⁹F, and ¹³C NMR spectroscopies. Each freebase and metalloporphyrin was additionally characterized by UV-vis spectroscopy and electrospray ionization mass spectrometry. Detailed synthetic procedures and characterization of all novel compounds can be found in the SI.

Scheme 2 Synthesis of each amide-functionalized porphyrin isomer. Top route is used to prepare Fe-ortho-proximal and Fe-meta-proximal, while bottom route is used to prepare Fe-ortho-distal and Fe-meta-distal.

2.2 Computational and spectroscopic analysis of pendent amide conformation

First, computational analysis was performed on each isomer to predict the lowest energy conformation of each amide pendant. To streamline these calculations, the analogous closed-shell zinc porphyrins were analysed in favour of the open-shell iron complexes. We note that the conformation of the porphyrin scaffold is nearly identical when comparing iron and zinc porphyrins,^17,18,23 and metal identity is thus expected to have a minimal effect on pendent amide orientation. Conformational screening of each amide pendant was first performed using Macromodel and the lowest energy conformers were subsequently optimized using Density Functional Theory (DFT) (see Fig. S1 and the SI for details). The computed structure of each zinc complex is depicted in Fig. 1. Consistent with our previous reports,^17,18 the amide NH in Zn-ortho-proximal is positioned to facilitate intramolecular proton transfer and/or hydrogen bonding with a metal-bound CO₂ adduct. By contrast, in Zn-ortho-distal there is a notable interaction between the amide aryl ring and the porphyrin π system that leads the amide NH to point slightly away from the metal active site in the predicted lowest energy conformation. The pendent amides in the two meta isomers are positioned far from the active site and are therefore expected to exhibit minimal interactions with bound substrate, consistent with their role as positional controls.

Fig. 1 Computed DFT structures of each Zn complex displaying the predicted lowest energy conformation of the amide pendant (see the SI for computational details).

We turned to NMR spectroscopy for experimental corroboration of these computational results. ¹H NMR spectra of the freebase porphyrins show that the chemical shift of the amide proton varies significantly with its positioning relative to the porphyrin ring: this peak appears at 9.82 ppm for the ortho-proximal isomer and 11.04 ppm for the ortho-distal isomer (Fig. S25 and S29). Ring currents through the porphyrin macrocycle result in shielding of protons located near or above the porphyrin centre and deshielding of those in the porphyrin periphery.^40–44 Accordingly, we conclude that the amide proton of the ortho-proximal isomer is oriented more towards the porphyrin centre, while the amide proton of the ortho-distal isomer is directed away from the centre. The amide protons in the two meta isomers have similar chemical shifts (10.84 ppm and 10.79 ppm, Fig. S36 and S39), supporting the idea that the chemical shift difference between the two ortho isomers truly reflects positional changes rather than different through-bond effects. Furthermore, Nuclear Overhauser Effect Spectroscopy (NOESY) experiments show a cross peak indicating a through-space interaction between the amide proton and one of the peripheral β-pyrrolic protons for the ortho-distal isomer (Fig. S30). This finding confirms a predominant conformation wherein the amide proton is oriented away from the porphyrin active site. No such cross peak interactions are seen with the ortho-proximal isomer (Fig. S26), indicating that the amide pendant is more rigidly positioned above the porphyrin centre. In sum, the ¹H NMR spectra and NOESY data both support the structural conclusions from computational analysis.

2.3 Electrochemical and pK_a measurements under an inert atmosphere

To evaluate the electrochemical behaviour of each isomer in the absence of substrate, cyclic voltammetry (CV) experiments were performed under 1 atm Argon (Ar) in MeCN with 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF₆) as supporting electrolyte. Each complex displays three redox events that are typical of iron porphyrins (Fig. S2): a broad Fe^III/II couple, which is associated with axial chloride and solvent binding equilibria, and two further reductions that are formally termed Fe^II/I and Fe^I/0, although they are more accurately described as ligand-centred reductions that give the Fe(II) porphyrin radical anion and dianion, respectively.^45,46 The formal Fe^I/0 couple is most relevant for CO₂ reduction, since the Fe⁰ species is traditionally catalytically active.^47–49 At a scan rate of 0.1 V s⁻¹, Fe-ortho-proximal, Fe-meta-proximal, and Fe-meta-distal each display a chemically reversible Fe^I/0 couple; however, Fe-ortho-distal only displays a reversible couple at higher scan rates (≥10 V s⁻¹) (Fig. S3). The reduction potentials of the Fe^I/0 couple (E_1/2(Fe^I/0)) for the two ortho isomers are ∼65 mV more positive than the E_1/2(Fe^I/0) values for the corresponding meta isomers (Table 1), and each catalyst is freely diffusing under the specified electrochemical conditions (Fig. S4).

Table 1 Summary of thermodynamic parameters for each amide functionalized Fe-porphyrin

Catalyst	E1/2(Fe^I/0) (V vs. Fc/Fc^+)^a	SCS pKa (MeCN)^b
a Uncertainty is ±5 mV.b Values are an average of four separate titration points.
Fe-ortho-proximal	−2.000	24.7 ± 0.2
Fe-ortho-distal	−1.995	24.1 ± 0.1
Fe-meta-proximal	−2.068	—
Fe-meta-distal	−2.056	—

---

The pK_as of both ortho isomers were measured by UV-Vis spectrophotometry in MeCN to determine how distal versus proximal positioning of the NH donor might affect amide acidity. These experiments involve equilibration between a deprotonated colorimetric indicator of known pK_a (phenol red) and an amide-functionalized iron porphyrin of unknown pK_a (Fig. S5). As shown in Table 1, Fe-ortho-proximal (pK_a = 24.7 ± 0.2) and Fe-ortho-distal (pK_a = 24.1 ± 0.1) have similar pK_as, suggesting that NH donor placement has a relatively minor effect on SCS acidity.

2.4 Electrocatalytic CO₂ reduction in the presence of a mild Brønsted acid

To first assess catalytic activity in the presence of a mild Brønsted acid, we evaluated the electrochemical behaviour of each isomer by CV under 1 atm CO₂ with 50 mM exogenous phenol (pK_a = 29.2 in MeCN).³⁴ Catalyst concentrations are kept low (0.25 mM) to minimize the effects of catalyst aggregation.⁵⁰ Peak-shaped current responses with large i/i⁰_p values (catalytic current normalized to the current in absence of substrate) are observed at potentials near the Fe^I/0 couple in all cases (Fig. 2 and S6). Notably, Fe-ortho-distal exhibits a peak i/i⁰_p that is ∼5 times lower than the values observed for the other three isomers, suggesting that this catalyst is less active under the specified conditions (vide infra). Qualitatively, these data indicate that each isomer is active towards electrocatalytic CO₂ reduction and that catalytic activity depends on positioning of the amide pendant.

Fig. 2 Catalytic cyclic voltammograms for each iron porphyrin isomer using phenol as the proton donor. Colours and symbols correspond to abbreviated structures shown on the left. Conditions: 0.25 mM catalyst, 50 mM phenol, 1 atm CO₂ (0.28 M), and 0.1 M TBAPF₆ in MeCN.

Controlled potential electrolysis (CPE) and headspace gas chromatography analysis were performed to evaluate product selectivity across the isomer series. Accordingly, each catalyst was electrolyzed at a potential near the Fe^I/0 couple under 1 atm CO₂ in the presence of 250 mM phenol (Fig. S8–S11 and Table S1). Under these conditions, three of the four isomers—Fe-ortho-proximal, Fe-meta-distal, and Fe-meta-proximal—display selectivity that is typical for FeTPP catalysts in the presence of a mild Brønsted acid:^{6,13,17,19,47,51,52} FE_CO ≥ 80% and FE_H2 ≤ 5% (Table 2). By comparison, Fe-ortho-distal shows somewhat poorer selectivity towards CO (FE_CO = 67 ± 2%) and increased H₂ evolution (FE_H2 = 13 ± 1%), confirming that placement of the NH donor at the ortho-distal position adversely alters reaction selectivity. In the presence of a mild exogenous acid, SCS donor positioning therefore appears to have a modest effect on product selectivity.

Table 2 CO₂ reduction selectivity and activity data for each amide positional isomer using phenol as the exogenous acid

Catalyst	FECO^a (%)	FEH2^a (%)	log(kobs/s^−1) (FOWA)^b	log(kobs/s^−1) (plateau)^c
a Faradaic efficiencies are averages of two independent experiments with [phenol] = 250 mM.b Values are averages of 3 independent measurements; uncertainties represent one standard deviation.c Uncertainties are estimated at approximately ±15%.^12 All kinetic data were obtained with [phenol] = 50 mM.
Fe-ortho-proximal	85 ± 3	4 ± 1	4.85 ± 0.06	5.1
Fe-ortho-distal	67 ± 2	13 ± 1	3.99 ± 0.08	4.0
Fe-meta-proximal	82 ± 1	3 ± 1	3.86 ± 0.07	—
Fe-meta-distal	80 ± 1	5 ± 1	4.13 ± 0.08	4.1

---

After evaluating reaction selectivity, we determined how the kinetics of CO₂ reduction depend on NH donor positioning. Kinetics may be extracted from CV data via catalytic plateau currents—when side phenomena like substrate consumption or catalyst deactivation are minimal—or by the more general foot-of-the-wave analysis (FOWA), which has been extensively described by Savéant and co-workers.^53,54 At slower scan rates (0.1 V s⁻¹) with 50 mM phenol and 1 atm CO₂, plateau currents are not observed (Fig. 2). FOWA was therefore used to extract pseudo-first order rate constants (k_obs, s⁻¹) for each catalyst (Table 2 and Fig. S15). At fast scan rates (50 to 250 V s⁻¹), catalytic plateau currents are obtained for three of the four isomers (Fe-ortho-proximal, Fe-ortho-distal, and Fe-meta-distal) (Fig. S16–S18). The associated k_obs values from plateau currents are in good agreement with the kinetics obtained from FOWA, which corroborates the FOWA approach. Fe-ortho-proximal is the most active catalyst in the series, displaying a k_obs that is roughly an order of magnitude larger than the values associated with each of the remaining three isomers. While this comparison does not account for differences in catalyst electronics—wherein a more negative E_1/2(Fe^I/0) often correlates with faster turnover^5,6,17—Fe-ortho-proximal displays the largest k_obs despite its relatively positive E_1/2(Fe^I/0), which suggests that the SCS in this catalyst engages in significant through-space interactions. It is worth noting that Fe-ortho-distal displays comparable rate constants to the meta isomers despite its significantly lower catalytic peak current (i/i⁰_p). This observation may indicate that catalyst inhibition limits the peak current for Fe-ortho-distal during successive turnovers, as detailed by Savéant.^53,54 We additionally measure a kinetic isotope effect (KIE; H/D) of 1.5–1.7 across the series, confirming that proton transfer contributes to the rate-limiting step of catalysis (Fig. S19 and Table S4). These KIE values are also consistent with previously reported values for FeTPP catalysed CO₂ reduction.^17,30 Together, these results confirm that—under mildly acidic conditions—SCS donor positioning can significantly alter kinetic barriers but leads to only modest changes in reaction selectivity during FeTPP-catalysed CO₂ reduction.

2.5 Electrocatalytic activity and selectivity with a strongly acidic exogenous proton donor

Building on the results with the mild Brønsted acid phenol, we next sought to investigate reaction selectivity and activity with the significantly more acidic 3,5-bis(trifluoromethyl) phenol (pK_a = 23.8 in MeCN).³⁴ With increasing concentrations of acid under 1 atm CO₂, the voltammograms of each catalyst display increasingly large current responses near the Fe^I/0 couple, consistent with a proton-dependent electrocatalytic response (Fig. 3). It is notable that at lower acid concentrations, Fe-ortho-proximal exhibits two multi-electron waves—one with peak i/i⁰_p at ca. −2.0 V and the second at ca. −2.2 V—that become indistinguishable at higher acid concentrations (≥50 mM). This suggests that two separate potential-dependent pathways may be operative with this combination of acid and catalyst. In contrast, the remaining three isomers show only one apparent catalytic wave, even at lower acid concentrations. Compared to their responses with the less acidic phenol, each isomer generally shows a more positive current onset with the stronger acid, which is consistent with faster catalysis and/or involvement of new reaction mechanisms.

Fig. 3 Catalytic cyclic voltammograms for (A) Fe-ortho-proximal, (B) Fe-ortho-distal, (C) Fe-meta-proximal, and (D) Fe-meta-distal with increasing amounts of 3,5-bis(trifluoromethyl) phenol. Conditions: 0.25 mM catalyst, 0, 5, 10, 20, 30, and 50 mM 3,5-bis(trifluoromethyl) phenol, 1 atm CO₂ (0.28 M), and 0.1 M TBAPF₆ in MeCN.

To evaluate whether and how product selectivity changes under these more acidic conditions, CPE experiments were performed with each catalyst under 1 atm CO₂ using 50 mM 3,5-bis(trifluoromethyl) phenol (Fig. S12 and Table S2). When Fe-ortho-distal is electrolysed at potentials near E_1/2(Fe^I/0), minimal quantities of CO are produced (FE_CO = 11 ± 1%) and significant background proton reduction is observed (FE_H2 = 65 ± 2%) (Fig. 4 and Table 3). Both Fe-meta-distal and Fe-meta-proximal display similarly unfavourable reaction selectivity. Nevertheless, these three catalysts still show more favourable selectivity than iron tetraphenylporphyrin triflate, [FeTPP]OTf (a catalyst lacking amide pendants), which almost exclusively catalyses background proton reduction (FE_H2 = 78 ± 5%) (Fig. S7 and S13, Table 3). This finding suggests that the amide pendants in the three aforementioned isomers still confer a marginal selectivity advantage under the specified acidic conditions. In stark contrast, Fe-ortho-proximal displays a remarkably high FE_CO = 83 ± 4% and minimal H₂ evolution (FE_H2 = 7 ± 1%), irrespective of whether the electrolysis is performed near the more positive (ca. −2.0 V) or negative (ca. −2.2 V) voltammetric feature. These results indicate that the two catalytic waves are associated with two distinct pathways for CO evolution. It is noteworthy that the ratio of CO : H₂ formed with Fe-ortho-proximal is nearly identical to that obtained under far less acidic conditions (Table 2). Thus, suitably positioned SCS donors prevent indiscriminate proton transfers that otherwise lead to high FE_H2 and/or to a lowered total FE for gaseous products (Scheme 3).

Fig. 4 Faradaic efficiencies (FE) for CO (solid colours) and H₂ (hashed colours) obtained from CPE experiments with each iron porphyrin catalyst in the presence of 3,5-bis(trifluoromethyl) phenol. The amide pendants associated with each iron porphyrin catalyst are depicted beneath the plot, alongside [FeTPP]OTf as a control lacking a protic SCS. Conditions: 0.1 mM catalyst, 50 mM 3,5-bis(trifluoromethyl) phenol, 1 atm CO₂, and 0.1 M TBAPF₆ in MeCN.

Table 3 CO₂ reduction selectivity and activity data for each iron porphyrin catalyst with 50 mM 3,5-bis(trifluoromethyl) phenol

Catalyst	FECO^a (%)	FEH2^a (%)	log(kobs/s^−1) (plateau)^b
a Faradaic efficiencies for CO (left) and H2 (right) that are averages from three independent experiments.b Uncertainty is estimated at approximately ±15%.
Fe-ortho-proximal	83 ± 4	7 ± 1	4.7
Fe-ortho-distal	11 ± 1	65 ± 2	—
Fe-meta-proximal	27 ± 5	45 ± 6	—
Fe-meta-distal	14 ± 3	55 ± 3	—
[FeTPP]OTf	4 ± 2	78 ± 5	—

---

Scheme 3 Simplified reaction scheme showing how suitably positioned SCS groups promote selective CO₂ reduction (top), while improperly positioned SCS donors enable indiscriminate proton transfers that lead to H₂ and an overall decrease in total FE for gaseous products (bottom).

The favorable selectivity displayed by Fe-ortho-proximal with 3,5-bis(trifluoromethyl) phenol prompted us to establish a practical limit to exogenous acid acidity. Accordingly, we evaluated the CO₂ reduction selectivity of Fe-ortho-proximal in the presence of 50 mM pentafluorophenol (pK_a = 20.1)³⁴ (Fig. S14). When this catalyst is electrolyzed at potentials near E_1/2(Fe^I/0) under 1 atm of CO₂, minimal quantities of CO are observed (FE_CO = 7 ± 1%), and H₂ is obtained as the major reduction product (FE_H2 = 66 ± 2%) (Table S3). Thus, the pK_a of 3,5-bis(trifluoromethyl) phenol likely represents the maximum acidity limit wherein Fe-ortho-proximal remains selective towards CO, and further decreasing the exogenous acid pK_a by ∼4 units results in poor reaction selectivity.

After evaluating the selectivity of each catalyst, we next sought to measure catalytic rate constants under these strongly acidic conditions and compare them to the kinetic parameters obtained in the presence of phenol. This analysis was restricted to Fe-ortho-proximal with 3,5-bis(trifluoromethyl) phenol, as it is the only isomer/acid pairing with high selectivity for CO and thus the only catalyst for which the current response is mostly attributable to a single reaction product. The presence of two overlapping catalytic features when using 3,5-bis(trifluoromethyl) phenol precludes reliable application of FOWA to determine k_obs. Instead, we report a k_obs that is derived from catalytic plateau currents at high scan rates (100 V s⁻¹) and note that this value likely represents an amalgamation of the rate constants associated with the two operating catalytic pathways (Fig. S20). The log(k_obs) of 4.7 measured with 3,5-bis(trifluoromethyl) phenol (Table 3) is similar to the log(k_obs) of 5.1 obtained with phenol (Table 2).

This similarity in k_obs is initially surprising, as it contrasts with previous work from our group demonstrating that more acidic exogenous phenols can significantly increase CO₂ reduction rate constants.¹⁸ Savéant and co-workers have also reported a similar Brønsted relationship with unfunctionalized FeTPP and a range of structurally diverse exogenous acids.³⁰ Departure from the expected Brønsted relationship could arise if k_obs has contributions from multiple catalytic pathways or if there is a substantial change in the character of the rate-limiting transition state under highly acidic conditions. Another possible explanation is that the similar k_obs values obtained with different exogenous acids is due to the SCS amide's role as a proton relay in the rate-limiting step of catalysis. Once the exogenous acid is sufficiently acidic and present in high enough concentration to favour fast SCS re-protonation, k_obs should depend only on the intramolecular proton transfer step and therefore should become invariant with exogenous acid concentration.^55,56 Indeed, we find that k_obs becomes independent of 3,5-bis(trifluoromethyl) phenol concentration at acid concentrations >50 mM (Fig. S21).

Since water can play a key role in SCS proton transfers,^{21,23,24,29,57} we investigated the potential role of adventitious water on catalytic activity. We thereby measured kinetic parameters with dry MeCN electrolyte and 3,5-bis(trifluoromethyl) phenol at a concentration where catalytic rates still depend on exogenous acid concentration (20 mM) in the presence of an equimolar concentration of water (Fig. S22). Under these conditions, the value of log(k_obs) is 4.1 and does not significantly change in the presence or absence of water, confirming that adventitious water plays a minimal role during rate-limiting proton transfer.

The electrocatalytic behaviour of Fe-ortho-proximal with 3,5-bis(trifluoromethyl) phenol as an exogenous acid is clearly more complex than what is typical for FeTPP catalysts.^30,47,58 Based on computational results and NMR spectroscopy, it appears that the NH donor in Fe-ortho-proximal is ideally oriented to direct proton transfers from the exogenous acid to bound CO₂ at the active site. In contrast, the lowest energy conformers of the other three catalysts indicate that the NH donors are either too far away (in the case of Fe-meta-proximal and Fe-meta-distal) or misdirected (as for Fe-ortho-distal). We propose that this difference in alignment of the proton transfer coordinate is responsible for the selectivity enhancement observed with Fe-ortho-proximal. It is also possible that this change in donor positioning results in stronger or weaker hydrogen bonding interactions between the SCS and CO₂, and that more suitably positioned NH donors (i.e., Fe-ortho-proximal) might better stabilize the catalyst-CO₂ adduct; however, in our previous report describing a series of ortho-proximal isomers bearing amides of variable pK_a, we found that these SCS amides have a minimal influence on the CO₂ binding equilibrium,¹⁸ suggesting that the differences in product selectivity described here are more likely the result of an SCS proton relay effect. Furthermore, we speculate that the optimal alignment of an SCS proton donor lowers transition state energies for multiple CO₂ reduction pathways that differ in the sequence of electron and proton transfers. In a related fashion, Aukauloo and co-workers reported an FeTPP catalyst bearing SCS urea groups that activates CO₂ via the unconventional formal Fe^I oxidation state.²⁶ The Dempsey group also observed a similar phenomenon during proton reduction, wherein changes to solution and SCS acidity alter the sequence of proton and electron transfer events leading to metal hydride formation.^59–61 Continued exploration of the interplay between SCS donor positioning and solution acidity may thus present opportunities to investigate novel changes in both the reaction mechanism and selectivity of FeTPP-catalysed CO₂ reduction.

3. Conclusions

Herein, we have evaluated the product selectivity of four iron porphyrin isomers bearing a pendent amide at varying positions around the catalytic active site. The NH donor of each amide was installed at the ortho- or meta-positions of the meso porphyrin ring, and either proximal or distal to the iron porphyrin core. In the presence of phenol—a conventionally mild Brønsted acid—each isomer shows favourable selectivity for CO, with minimal H₂ evolution. Under these conditions, the Fe-ortho-proximal isomer displays the most favourable kinetics, confirming that suitable positioning of SCS groups can lower the barrier to CO₂ reduction. Remarkably, in the presence of a significantly more acidic exogenous acid—3,5-bis(trifluoromethyl) phenol—it is found that SCS donor positioning has a significant effect on reaction selectivity: Fe-ortho-proximal remains highly selective for CO, while the remaining three isomers mainly promote undesired H₂ evolution. Together, these results demonstrate that proper SCS donor positioning is an important design consideration that can beneficially alter reaction selectivity, particularly with regard to avoiding indiscriminate proton transfers. This work motivates future investigations into the role of SCS donor positioning on changing reaction outcomes, and how such effects might be relevant to promoting the ever-challenging reduction of CO₂ beyond 2e⁻.

Author contributions

KT and EMN conceived of the project and wrote the manuscript, while JPR performed all computational analysis. KT conducted all other experimental work and data analysis.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the SI. Supplementary information: experimental procedures, characterization data, and computational details. See DOI: https://doi.org/10.1039/d5sc04700e.

Acknowledgements

This work was supported by the University of British Columbia, the Natural Sciences and Engineering Research Council of Canada (RGPIN-2021-03691 and DGECR-2021-00427), and the Research Corporation for Science Advancement (27752). KT acknowledges support from a British Columbia Graduate Fellowship as well as UBC Chemistry for a Head's departmental scholarship. EMN is grateful to CIFAR for support as an Azrieli Global Scholar. Computational resources were provided from the Digital Research Alliance of Canada and the Advanced Research Computing (ARC) centre at the University of British Columbia.

Notes and references

1. A. M. Appel, J. E. Bercaw, A. B. Bocarsly, H. Dobbek, D. L. DuBois, M. Dupuis, J. G. Ferry, E. Fujita, R. Hille, P. J. A. Kenis, C. A. Kerfeld, R. H. Morris, C. H. F. Peden, A. R. Portis, S. W. Ragsdale, T. B. Rauchfuss, J. N. H. Reek, L. C. Seefeldt, R. K. Thauer and G. L. Waldrop, Chem. Rev., 2013, 113, 6621–6658.
2. O. S. Bushuyev, P. De Luna, C. T. Dinh, L. Tao, G. Saur, J. van de Lagemaat, S. O. Kelley and E. H. Sargent, Joule, 2018, 2, 825–832.
3. A. J. Morris, G. J. Meyer and E. Fujita, Acc. Chem. Res., 2009, 42, 1983–1994.
4. E. E. Benson, C. P. Kubiak, A. J. Sathrum and J. M. Smieja, Chem. Soc. Rev., 2009, 38, 89–99.
5. I. Azcarate, C. Costentin, M. Robert and J.-M. Savéant, J. Phys. Chem. C, 2016, 120, 28951–28960.
6. I. Azcarate, C. Costentin, M. Robert and J.-M. Savéant, J. Am. Chem. Soc., 2016, 138, 16639–16644.
7. C. L. Drennan, J. Heo, M. D. Sintchak, E. Schreiter and P. W. Ludden, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 11973–11978.
8. W. W. Cleland, T. J. Andrews, S. Gutteridge, F. C. Hartman and G. H. Lorimer, Chem. Rev., 1998, 98, 549–562.
9. M. Zhao, H.-B. Wang, L.-N. Ji and Z.-W. Mao, Chem. Soc. Rev., 2013, 42, 8360–8375.
10. M. W. Drover, Chem. Soc. Rev., 2022, 51, 1861–1880.
11. J. Trouvé and R. Gramage-Doria, Chem. Soc. Rev., 2021, 50, 3565–3584.
12. D. J. Martin and J. M. Mayer, J. Am. Chem. Soc., 2021, 143, 11423–11434.
13. A. Chapovetsky, M. Welborn, J. M. Luna, R. Haiges, T. F. Miller III and S. C. Marinescu, ACS Cent. Sci., 2018, 4, 397–404.
14. A. W. Nichols and C. W. Machan, Front. Chem., 2019, 7, 1–19.
15. C. Costentin, S. Drouet, M. Robert and J.-M. Savéant, Science, 2012, 338, 90–94.
16. C. Costentin, G. Passard, M. Robert and J.-M. Savéant, J. Am. Chem. Soc., 2014, 136, 11821–11829.
17. E. M. Nichols, J. S. Derrick, S. K. Nistanaki, P. T. Smith and C. J. Chang, Chem. Sci., 2018, 9, 2952–2960.
18. K. Teindl, B. O. Patrick and E. M. Nichols, J. Am. Chem. Soc., 2023, 145, 17176–17186.
19. C. G. Margarit, C. Schnedermann, N. G. Asimow and D. G. Nocera, Organometallics, 2019, 38, 1219–1223.
20. C. G. Margarit, N. G. Asimow, M. I. Gonzalez and D. G. Nocera, J. Phys. Chem. Lett., 2020, 11, 1890–1895.
21. P. Sen, B. Mondal, D. Saha, A. Rana and A. Dey, Dalton Trans., 2019, 48, 5965–5977.
22. A. Sonea, N. R. Crudo and J. J. Warren, J. Am. Chem. Soc., 2024, 146, 3721–3731.
23. J. S. Derrick, M. Loipersberger, S. K. Nistanaki, A. V. Rothweiler, M. Head-Gordon, E. M. Nichols and C. J. Chang, J. Am. Chem. Soc., 2022, 144, 11656–11663.
24. P. Gotico, B. Boitrel, R. Guillot, M. Sircoglou, A. Quaranta, Z. Halime, W. Leibl and A. Aukauloo, Angew. Chem., Int. Ed., 2019, 58, 4504–4509.
25. A. Sonea, K. L. Branch and J. J. Warren, ACS Catal., 2023, 13, 3902–3912.
26. S. Amanullah, P. Gotico, M. Sircoglou, W. Leibl, M. J. Llansola-Portoles, T. Tibiletti, A. Quaranta, Z. Halime and A. Aukauloo, Angew. Chem., Int. Ed., 2024, 63, e202314439.
27. H. Rao, L. C. Schmidt, J. Bonin and M. Robert, Nature, 2017, 548, 74–77.
28. S. Amanullah, P. Saha and A. Dey, J. Am. Chem. Soc., 2021, 143, 13579–13592.
29. S. Ghosh, S. Patra, A. Bisoi, S. Ghosh, N. Maurya, A. Das, P. C. Singh and A. Dey, ACS Catal., 2025, 15, 3595–3610.
30. C. Costentin, S. Drouet, G. Passard, M. Robert and J.-M. Savéant, J. Am. Chem. Soc., 2013, 135, 9023–9031.
31. C. G. Margarit, N. G. Asimow, C. Costentin and D. G. Nocera, ACS Energy Lett., 2020, 5, 72–78.
32. I. Bhugun, D. Lexa and J.-M. Savéant, J. Am. Chem. Soc., 1996, 118, 3982–3983.
33. F. Bordwell and D. Algrim, J. Org. Chem., 1976, 41, 2507–2508.
34. A. Kütt, S. Tshepelevitsh, J. Saame, M. Lõkov, I. Kaljurand, S. Selberg and I. Leito, Eur. J. Org Chem., 2021, 2021, 1407–1419.
35. F. G. Bordwell and J. Cheng, J. Am. Chem. Soc., 1991, 113, 1736–1743.
36. S. S. Roy, K. Talukdar and J. W. Jurss, ChemSusChem, 2021, 14, 662–670.
37. M. L. Pegis, B. A. McKeown, N. Kumar, K. Lang, D. J. Wasylenko, X. P. Zhang, S. Raugei and J. M. Mayer, ACS Cent. Sci., 2016, 2, 850–856.
38. S. Sinha and J. J. Warren, Inorg. Chem., 2018, 57, 12650–12656.
39. K. T. Ngo, M. McKinnon, B. Mahanti, R. Narayanan, D. C. Grills, M. Z. Ertem and J. Rochford, J. Am. Chem. Soc., 2017, 139, 2604–2618.
40. U. Simonis, F. A. Walker, P. L. Lee, B. J. Hanquet, D. J. Meyerhoff and W. R. Scheidt, J. Am. Chem. Soc., 1987, 109, 2659–2668.
41. J. P. Collman, X. Zhang, K. Wong and J. I. Brauman, J. Am. Chem. Soc., 1994, 116, 6245–6251.
42. D. Reddy and T. K. Chandrashekar, J. Chem. Soc., Dalton Trans., 1992, 619–625.
43. J. P. Collman, A. O. Chong, G. B. Jameson, R. T. Oakley, E. Rose, E. R. Schmittou and J. A. Ibers, J. Am. Chem. Soc., 1981, 103, 516–533.
44. C. J. Medforth, in Fundamentals of Porphyrin Chemistry, 2022, pp. 611–629.
45. C. Römelt, J. Song, M. Tarrago, J. A. Rees, M. van Gastel, T. Weyhermüller, S. DeBeer, E. Bill, F. Neese and S. Ye, Inorg. Chem., 2017, 56, 4745–4750.
46. C. Römelt, S. Ye, E. Bill, T. Weyhermüller, M. van Gastel and F. Neese, Inorg. Chem., 2018, 57, 2141–2148.
47. I. Bhugun, D. Lexa and J.-M. Savéant, J. Am. Chem. Soc., 1996, 118, 1769–1776.
48. I. Bhugun, D. Lexa and J.-M. Savéant, J. Phys. Chem., 1996, 100, 19981–19985.
49. M. Hammouche, D. Lexa, J. M. Savéant and M. Momenteau, J. Electroanal. Chem. Interfacial Electrochem., 1988, 249, 347–351.
50. K. L. Branch, E. R. Johnson and E. M. Nichols, ACS Cent. Sci., 2024, 10, 1251–1261.
51. K. Kosugi, M. Kondo and S. Masaoka, Angew. Chem., Int. Ed., 2021, 60, 22070–22074.
52. M. R. Narouz, P. De La Torre, L. An and C. J. Chang, Angew. Chem., Int. Ed., 2022, 61, e202207666.
53. C. Costentin, S. Drouet, M. Robert and J.-M. Savéant, J. Am. Chem. Soc., 2012, 134, 11235–11242.
54. C. Costentin and J.-M. Savéant, ChemElectroChem, 2014, 1, 1226–1236.
55. M. Haake, B. Reuillard, M. Chavarot-Kerlidou, C. Costentin and V. Artero, Angew. Chem., Int. Ed., 2024, 63, e202413910.
56. J.-M. Savéant, Angew. Chem., Int. Ed., 2019, 58, 2125–2128.
57. C.-Y. Yeh, C. J. Chang and D. G. Nocera, J. Am. Chem. Soc., 2001, 123, 1513–1514.
58. C. Costentin, M. Robert and J.-M. Savéant, Chem. Soc. Rev., 2013, 42, 2423–2436.
59. C. L. Montgomery, M. Z. Ertem, L. Chevalier and J. L. Dempsey, J. Am. Chem. Soc., 2024, 146, 30020–30032.
60. C. L. Montgomery, M. Z. Ertem, Z. H. Claytor and J. L. Dempsey, Inorg. Chem., 2025, 64, 10139–10149.
61. C. L. Montgomery, J. Amtawong, A. M. Jordan, D. A. Kurtz and J. L. Dempsey, Chem. Soc. Rev., 2023, 52, 7137–7169.

---