Direct observation of intermediates formed during steady-state electrocatalytic O₂ reduction by iron porphyrins

Abstract

Heme/porphyrin-based electrocatalysts (both synthetic and natural) have been known to catalyze electrochemical O₂, H⁺, and CO₂ reduction for more than five decades. So far, no direct spectroscopic investigations of intermediates formed on the electrodes during these processes have been reported; and this has limited detailed understanding of the mechanism of these catalysts, which is key to their development. Rotating disk electrochemistry coupled to resonance Raman spectroscopy is reported for iron porphyrin electrocatalysts that reduce O₂ in buffered aqueous solutions. Unlike conventional single-turnover intermediate trapping experiments, these experiments probe the system while it is under steady state. A combination of oxidation and spin-state marker bands and metal ligand vibrations (identified using isotopically enriched substrates) allow in situ identification of O₂-derived intermediates formed on the electrode surface. This approach, combining dynamic electrochemistry with resonance Raman spectroscopy, may be routinely used to investigate a plethora of metalloporphyrin complexes and heme enzymes used as electrocatalysts for small-molecule activation.

Electrocatalysis has taken center stage in contemporary research because of its indisputable relevance in the thriving area of renewable energy. Electrocatalytic O₂ reduction, O₂ evolution, H₂ evolution, and H₂ oxidation thus are areas of great interest and have attracted focused effort across several disciplines of science (1–11). Both natural systems [metalloenzymes such as cytochrome c oxidase (CcO), hydrogenase, and glucose oxidase] and biomimetic systems (i.e., synthetic/biochemical mimics of the natural enzyme) are known to function as electrocatalysts (12–15). In particular, it has been known for five decades that heme/porphyrin-based electrocatalysts (both natural and synthetic) catalyze electrochemical O₂, H⁺, and CO₂ reduction (14, 16–19). Several excellent metalloporphyrin-based O₂ reduction electrocatalysts have been reported, some mimicking natural active sites and some inspired by their design (14, 20). Of these catalysts, the Fe-based catalysts generally have provided valuable insights into the O₂ reduction mechanism and a deeper understanding of the structure–function correlations of key enzymes such as CcO, the terminal enzyme in the mitochondrial electron transport chain (1, 21). Similarly, several naturally occurring enzymes have been used as electrocatalysts for O₂ reduction (e.g., CcO, microperoxidase) and substrate oxidation using O₂ (e.g., cytochrome P450) (22–24). Apart from porphyrins, transition metal complexes of corroles, a related macrocycle that lacks one of the four mesocarbons of porphyrins, have been used extensively as electrocatalysts for O₂ reduction and H₂O oxidation (5, 6). These catalytic systems are efficient in catalyzing selective O₂ activation/reduction reactions when the electrons (i.e., the reducing component) are provided from the electrode and O₂ (the oxidizing component) is obtained from the bulk solution. An inherent advantage of this construct is its heterogeneity, which allows the use of water-insoluble catalysts to catalyze the reaction in an environmentally friendly solvent, such as water.

Elaborate mechanistic investigations have been conducted on these electrocatalysts; however, these investigations were limited to chemical and electrochemical perturbations (4). So far, there are no direct spectroscopic data on the intermediates formed during electrocatalysis. This is in sharp contrast to the investigation of O₂ activation by iron porphyrin catalysts and heme enzymes in homogeneous solutions. In these cases, a broad spectrum of reactive intermediate species, formed during single-turnover reactions, is reported (25–29). The problem lies mainly in investigating a thin layer of catalyst (a monolayer in many cases) on an electrode and getting a good signal-to-noise ratio.

Surface enhanced resonance Raman spectroscopy (SERRS) is particularly suited for investigating metalloporphyrins and metalloenzymes attached to surfaces. The surface enhancement in addition to conventional resonance Raman (rR) effect (by exciting the intense absorption bands of these complexes) provides excellent enhancement of signals (30, 31). Additionally, metalloporphyrin systems are particularly well suited for Raman investigations. Several symmetric ligand vibrations (intense in Raman spectroscopy) have been identified that reflect the oxidation and spin states of the metals chelated by these macrocycles with remarkable fidelity (32, 33). In particular, the combination of the ν₄ (∼1,330–1,375 cm⁻¹) and the ν₂ (∼1,540–1,575 cm⁻¹) ligand vibrations are used routinely to diagnose the ground-state electronic structure of iron porphyrin complexes and enzyme active sites. SERRS has been used extensively to investigate various physical aspects of cytochrome c functionalized electrodes, some of which have been under different applied potentials on the electrode (34).

Studying the reaction mechanism of a monolayer of catalyst/protein requires accumulation of a good signal-to-noise ratio on a small time scale. For example, the catalytic cycle of CcO, which goes through several intermediates, takes less than 1 ms at room temperature (35, 36). Collecting a good signal in such short time scales, without cryogenically trapping the intermediates (cannot be achieved during electrocatalysis), presents a formidable challenge. Alternatively, reactions may be monitored under steady-state conditions. In such a case, only the species that are rate determining will accumulate under steady-state conditions. In principle, any spectroscopic data obtained will contain contributions from all the species present and may be hard to resolve. However, in the case of metalloporphyrins, the individual intermediates may be identified by using a combination of the oxidation and spin-state marker bands ν₄ and ν₂ in the resonance Raman spectrum (which are unique and resolvable, Table 1) and metal ligand vibrations (with appropriate isotopic substitutions). In a related approach, the intermediates formed during the electrocatalytic H₂O₂/O₂ reduction by Bi/Pd-modified Au surfaces on a static electrode have been identified (37, 38).

Table 1.

Oxidation- and spin-state marker bands of iron porphyrins during steady-state conditions at room temperature (29, 32)

Samples	Species	Ligation	Spin	ν4, cm−1	ν2, cm−1
Oxidized
FeEs4	Fe(III) (32)	5	5/2	1,364	1,555
FeTPP	Fe(III) (32)	5	5/2	1,360	1,552
Reduced
FeEs4	Fe(II) (32)	5	2	1,349	1,547
FeTPP	Fe(II) (32)	5	2	1,344	1,539
During steady state
FeEs4	Fe(III)	6	1/2	1,369	1,565
	Fe(II)	5	2	1,352	1,540
	Fe(IV) = O (29)	6	—	1,371	1,571
FeTPP	Fe(III)	6	1/2	1,366	1,565
	Fe(II)	5	2	1,347	1,535
	Fe(IV) = O	6	—	1,371	1,570

In this study, we report an SERRS setup modified to accommodate a conventional rotating disk electrochemistry (RDE) or rotating ring disk electrochemistry (RRDE) setup (39) bearing iron porphyrin catalysts and engaged in steady-state electrocatalytic O₂ reduction. The data provide direct first-hand information about the nature of intermediates involved in steady-state O₂ reduction and helps define the mechanism of these catalysts.

Results and Discussion

The experimental station is described in Fig. 1. The working disk electrode is mounted at the tip of a conventional RDE shaft b. An air-tight, water-jacketed, three-electrode electrochemical cell c with a Pyrex optical window g is built such that both the incident laser and the Raman scattering are collected through this optical window. The scattered light is collected with a lens h and reflected on mirror j mounted on holder k toward another lens, which focuses it on the entrance slit of the spectrograph. Two iron porphyrin catalysts are investigated: iron tetraphenylporphyrin (FeTPP) (40) and iron α₄-tetra-2-(4-carboxymethyl-1,2,3-triazolyl)-phenylporphyrin (FeEs₄) (3) (Fig. 2). Both are known to electrocatalytically reduce O₂, and the latter recently was shown to have a hydrogen-bonding distal superstructure that can stabilize an Fe-O₂ adduct (28).

Fig. 1.

Experimental setup for SERRS-RDE.

Fig. 2.

Iron porphyrin complexes α₄-FeEs₄ (Left) and FeTPP (Right).

A cyclic voltammetry (CV) scan of the FeEs₄ catalysts (Fig. 3A, black line) immobilized on the octanethiol (C₈SH) self-assembled monolayer (SAM)-covered Au disk electrode shows the half-wave potential (E_1/2) of the Fe^III/II process at −100 mV vs. Ag/AgCl in a deaerated buffer solution. The integrated charge under this wave indicates a surface coverage of 8 × 10¹¹ molecules per square centimeter. Such low surface density is consistent with the formation of a monolayer and not a multilayer (typically 10¹⁴ molecules per square centimeter) of catalyst on the SAM (41, 42). This also is suggested by the atomic force microscopy (AFM) image of the electrode, which shows a smooth surface (with a height of 1.70–1.76 nm) without any multilayer (Fig. 3B) (43). SERRS-RDE data obtained by applying positive (0 V) and negative (−0.5 V) potentials (i.e., at potentials above and below the E_1/2) at 200 rpm (Pine Research Instrumentation; AFE6M rotor) rotation rate show the ν₄ and ν₂ vibrations (Table 1) indicative of oxidized (Fig. 4, bold black line) and reduced (Fig. 4, dashed black line) iron porphyrin complexes, respectively. Alternatively, a linear sweep voltammetry (LSV) scan performed in aerated buffer solutions using an RDE shows that as the potential of the working electrode is lowered gradually below the E_1/2, a catalytic current gradually rises as the result of electrocatalytic reduction of O₂ (Fig. 3A, gray line) by the Fe^II porphyrin produced on the electrodes at these potentials (i.e., below E_1/2). This catalytic current becomes potential independent at lower potentials and becomes mass transfer limited. The current is determined by the flow of the substrate, O₂, to the electrode, i.e., on the angular rotation rate of the electrode (38). This also is indicated by a plot of 1/i_cat (at −0.5 V) vs. 1/ω^1/2, which is linear (Fig. 3C).

Fig. 3.

(A) CV and LSV data of FeEs₄ physisorbed on C₈SH SAM in deoxygenated and air-saturated pH 7 buffers, respectively, at a scan rate of 50 mV/s using Ag/AgCl as a reference and Pt wire as counter electrodes. The dotted lines represent the variable potentials applied during the variable-potential SERRS-RDE experiments. (B) AFM image of the C₈SH SAM-covered electrode-bearing catalyst. (C) Plot of 1/i_cat vs. 1/ω^1/2 obtained at −0.5 V.

Fig. 4.

SERRS-RDE data of FeEs₄ (A) and FeTPP (B) in the low-frequency (Left) and high-frequency (Right) regions, respectively, where the oxidized spectra are indicated by bold black lines, the reduced spectra under anaerobic conditions by dashed black lines, and the reduced spectra under aerobic conditions by gray lines.

SERRS-RDE data collected on the electrode held at −0.5 V (i.e., during a bulk electrolysis experiment using RDE) show a spectrum (Fig. 4, gray line) different from the one obtained in the absence of O₂ (Fig. 4, dashed black line) (38). Several ν₄ and ν₂ vibrations are observed in which a particular set of ν₄ and ν₂ vibrations uniquely represents an iron porphyrin complex having a particular oxidation, spin, and ligation state (Table 1). For the α₄-FeEs₄ complex, there is a clear increase in intensity at 1,369 cm⁻¹ and 1,565 cm⁻¹, suggesting the formation of a low-spin Fe^III species (Fig. 4A, gray line) during turnover (32). Additionally, peaks are observed at 1,352 cm⁻¹ and 1,540 cm⁻¹ (Fig. 4A, gray line), which are characteristic of the reduced Fe^II species. There also is an increase in intensity at 1,371 cm⁻¹ and 1,571 cm⁻¹, which clearly is observed in the difference spectrum (Fig. 5B) and suggests formation of high-valent Fe^IV = O species (29). Thus, in addition to the resting high-spin Fe^III species, a high-spin Fe^II species, a low-spin Fe^III species, and an Fe^IV = O species are observed during steady-state O₂ reduction by the FeEs₄ catalyst. In the low-energy region of the Raman spectrum, there are increases in intensities at 779 cm⁻¹, 631 cm⁻¹, and 570 cm⁻¹ (Fig. 4A, gray line).

Fig. 5.

Difference spectra of the SERRS-RDE data at different potentials −0.15 V (red), −0.25 V (yellow), −0.35 V (blue), and −0.5 V (green) from 0 V of FeEs₄ (A) and FeTPP (C) in the low-frequency (Left) and high-frequency (Right) regions, respectively. The ν₄ and ν₂ regions of the spectrum obtained from the difference of the spectrum at −0.5 V from the spectrum at 0 V of FeEs₄ and FeTPP, along with their fits, are shown in B and D, respectively. (E) The difference spectra obtained at different rotation rates of FeEs₄.

For the FeTPP complex, very similar increases in intensities are observed in the high-frequency region of the spectrum, suggesting the formation of high-spin Fe^II, low-spin Fe^III, and Fe^IV = O species during turnover (Fig. 4B, gray line). In the low-frequency region, peaks at 634 cm⁻¹, 677 cm⁻¹, 741 cm⁻¹, 778 cm⁻¹, and 830 cm⁻¹ gain intensity (Fig.4B, gray line).

The O₂ reduction current is potential dependent; hence, the SERRS-RDE data were collected by applying different potentials at the working electrode (rotation rate, 200 rpm). In the case of FeEs₄, the difference spectrum (data at a particular potential – resting oxidized) in the low-energy region shows that as the potential is lowered, peaks at 830 cm⁻¹, 782 cm⁻¹, 631 cm⁻¹, and 570 cm⁻¹ gain intensity (Fig. 5A), whereas the peak at the 580-cm⁻¹ region loses intensity as the electrocatalytic O₂ reduction current increases and approaches a potential-independent mass transfer-controlled limit. In the higher-energy region, the ν₄ and ν₂ vibrations at 1,352 cm⁻¹ and 1,540 cm⁻¹ (high-spin Fe^II), 1,369 cm⁻¹ and 1,565 cm⁻¹ (low-spin Fe^III), and 1,371 cm⁻¹ and 1,571 cm⁻¹ (Fe^IV = O) gain intensity (Fig. 5 A and C). Alternatively, the ν₄ and ν₂ vibrations at 1,364 cm⁻¹ and 1,555 cm⁻¹ (high-spin Fe^III) lose intensity (Fig. 5 A and C). Thus, as the catalytic current transitions from the potential-dependent region to the mass transfer-limited region, the high-spin Fe^II, the low-spin Fe^III, and an Fe^IV species gradually build up at the electrode at the expense of the high-spin Fe^III species (Fig. 5B). In the case of FeTPP, the analysis is complicated slightly by decay of the catalyst during the variable potential SERRS-RDE experiments (14). However, a clear increase in the intensities of the ν₄ and ν₂ vibrations at 1,347 cm⁻¹ and 1,535 cm⁻¹, 1,366 cm⁻¹ and 1,565 cm⁻¹, and 1,371 cm⁻¹ and 1,570 cm⁻¹ indicates a buildup of high-spin Fe^II, low-spin Fe^III, and Fe^IV = O species, respectively, on the electrode as the potential is lowered and the catalytic O₂ reduction current increases to attain a mass transfer-limited value (Fig. 5D). The data obtained at multiple rotation rates (Fig. 5E) show that at rotation rates >100 rpm, the distribution of the species observed remains the same, indicating that the reaction is in a steady state and there is a buildup of different intermediate species involved in this multistep catalytic reaction. In the low-energy region, peaks at 596 cm⁻¹, 638 cm⁻¹, and 779 cm⁻¹ gain intensity as the potential is lowered gradually to negative values, i.e., where the Fe^III is reduced to Fe^II (Fig. 5C). [Note that the peaks at 677 cm⁻¹, 742 cm⁻¹, and 832 cm⁻¹ gain intensity only at a very negative potential (Fig. 5C, green).]

Experiments were performed in ¹⁸O₂-saturated pH 7 buffers with FeEs₄ to gain insight into the nature of these intermediates. The variable potential SERRS-RDE data (Fig. 6 A and B), as well as the difference data, show significant differences between ¹⁶O₂ and ¹⁸O₂ data (Fig. 6 C and D). In particular, the 830-cm⁻¹ band, which gains intensity in ¹⁶O₂ buffer (Fig. 6 A and C), loses intensity in the ¹⁸O₂-saturated buffer (Fig. 6 B and D). Similarly, the 631-cm⁻¹ band loses intensity in the ¹⁸O₂ buffer while it gains intensity in ¹⁶O₂-saturated buffer, whereas the 580-cm⁻¹ band loses intensity in ¹⁶O₂ buffer but not in ¹⁸O₂ buffer (Fig. 6). There is a larger increase in intensity in the 770–780-cm⁻¹ region in the case of ¹⁸O₂ buffer relative to the increase observed in ¹⁶O₂ buffer. In particular, this peak, centered at 780.5 cm⁻¹ in ¹⁸O₂ buffer, is asymmetric because of the presence of a shoulder at higher energy that is not observed in ¹⁶O₂ buffer. This indicates the growth of a new peak in ¹⁸O₂ buffer. Furthermore, the intensity of the band at 746 cm⁻¹ increases more in ¹⁸O₂ buffer than it does in ¹⁶O₂ buffer. The differences in the intensities of these vibrations in the low-energy region of the spectrum between ¹⁶O₂- and ¹⁸O₂-containing buffers imply the presence of Fe-O_x species derived from O₂ during reduction. The ¹⁶O₂/¹⁸O₂ difference spectrum (obtained at different potentials) reveals a clear shift from 830 cm⁻¹ to 783 cm⁻¹ (Fig. 6 E and F). There is an additional shift from 631 cm⁻¹ to 594 cm⁻¹. Note that these differences arise only at potentials in which electrocatalytic O₂ reduction is observed, and their intensities increase as the potential is lowered and hence arise as the result of incorporation of ¹⁸O₂ into intermediates formed during electrocatalytic O₂ reduction.

Fig. 6.

SERRS-RDE data of FeEs₄ obtained at various potentials under ¹⁶O₂-containing (A) and ¹⁸O₂-containing (B) pH 7 buffers. The spectra obtained from the difference of various potentials from 0 V for ¹⁶O₂- and ¹⁸O₂-containing buffers are shown in C and D, respectively. (E) Difference spectra of ¹⁸O₂ from ¹⁶O₂ obtained at respective potentials. (F) The shifts obtained, along with their fits.

The peak at 782 cm⁻¹ that gradually gained intensity in the variable-potential SERRS-RDE experiment in both ¹⁶O₂ and ¹⁸O₂ (Fig. 6) possibly arises from an intermediate formed during O₂ reduction as well. However, a clear isotopic shift was not observed for these species. Note that a weak peak that grows in at 752 cm⁻¹ in the ¹⁸O₂ difference spectrum might be the result of an isotope shift of this species, i.e., 782 cm⁻¹ in ¹⁶O₂ to 752 cm⁻¹ in ¹⁸O₂; possibly, this arises from an Fe^IV = O species. Ferryl species are known to undergo fast exchange of the ferryl oxygen with water (44, 45). Thus, any Fe^IV = ¹⁸O formed will quickly exchange in H₂¹⁶O in the medium to form Fe^IV = ¹⁶O, leading to limited population of Fe^IV = ¹⁸O species. (An experiment in H₂O¹⁸ might help this assignment, but a minimum of 30 mL of this buffer would be required, which makes this experiment commercially unviable.) The 830-cm⁻¹ and 631-cm⁻¹ bands represent the O-O and Fe-O vibrations of a low-spin Fe^III-OOH species, respectively, and the 782-cm⁻¹ band represents the Fe-O stretch of an Fe^IV = O species. These assignments are consistent with the observation of ν₄ and ν₂ bands characteristic of low-spin Fe^III species and Fe^IV species in the variable-potential SERRS-RDE data (32). The ^16/18O₂ isotope analysis of FeTPP is complicated by its decay during electrocatalytic O₂ reduction.

In addition to the SERRS-RDE data, the formation of the high-valent Fe^IV = O species also was evaluated using this species (E⁰ ∼ 1 V) produced during O₂ reduction on the electrode to chemically oxidize [Fe(CN)₆]⁴⁻ in situ. The redox potential (E^o) for [Fe(CN)₆]^3−/4− is 250 mV vs. Ag/AgCl at pH 7. The Fe^III porphyrin species used here, which have an E^o of –100 mV, are incapable of oxidizing [Fe(CN)₆]⁴⁻ to [Fe(CN)₆]³⁻. In an RRDE setup, any [Fe(CN)₆]³⁻ produced on the working electrode is diffused radially to the Pt ring (as a result of the hydrodynamic current produced by the rotation of the electrode-bearing cylindrical shaft) and may be detected by reducing it back to [Fe(CN)₆]⁴⁻ by applying a potential of 0 V on the Pt ring (the Pt ring is independently addressable in an RRDE experiment using a bipotentiostat). In the presence of 10 mM K₄[Fe(CN)₆] in the buffer solution, a clear reduction current is observed in the ring and the onset of this current coincides with the onset of the O₂ reduction current. This indicates that the [Fe(CN)₆]⁴⁻ present in the solution is getting oxidized to [Fe(CN)₆]³⁻ by the Fe^IV = O species produced on the electrode during O₂ reduction (42), further supporting the formation of high-valent Fe^IV = O species on the electrode during O₂ reduction. Note that Fe^III-OOH species, also produced on the electrode, is proposed to oxidize C-H bonds (46). However, its oxidation proceeds by C-H bond activation and not via a classical outer/inner sphere mechanism generally operative in [Fe(CN)₆]⁴⁻ (47, 48). These processes are driven by free energy of electron transfer. The E^o of an Fe^III-OOH species likely is close to that of the resting Fe^III-OH species (E^o = −100 mV), which is much lower than the E^o of [Fe(CN)₆]⁴⁻ (E^o = 250 mV). Thus, electron transfer from [Fe(CN)₆]⁴⁻ to an Fe^III-OOH species is a significantly uphill process, i.e., oxidation of [Fe(CN)₆]⁴⁻ by an Fe^III-OOH species is not likely. Note that no current is observed in the Pt ring if there is no K₄[Fe(CN)₆] present in the solution or no catalyst is present on the electrode. The oxidation of [Fe(CN)₆]⁴⁻ by small amounts of H₂O₂ produced in the electrode as the result of incomplete reduction of O₂ may be precluded, as it is a very slow process under these conditions (49). It is important to note that the observation of an Fe^IV = O species on the electrode implies that its reduction to Fe^III-OH is slow despite a large driving force for electron transfer (ET) from the electrode (held at −0.5 V) to the Fe^IV = O species (E^o = ∼1 V). This is the case in the active site of CcO, where the ET from heme a to heme a₃ (with marginal driving force) is >10⁵ s⁻¹ (50), whereas the ET from heme a to P_R (the Fe^IV = O species produced during O₂ reduction) is ∼10³ s⁻¹ despite having a much larger driving force (51).

Based on the intermediates identified, a mechanistic proposal of O₂ reduction by these Fe porphyrins is shown in Fig. 7. The SERRS-RDE data indicate that Fe^II species, Fe^III high-spin species, low-spin Fe^III-OOH species, and Fe^IV = O species are present during steady-state electrocatalytic O₂ reduction. Furthermore, the variable-potential SERRS-RDE data indicate that as the potential is lowered, there is an increase in the population of the high-spin Fe^II species, low-spin Fe^III-OOH species, and Fe^IV = O species and a decrease in population of the resting high-spin Fe^III species. Thus, at potentials in which the current is potential dependent, the population of the O₂-derived species is less, as the reduction of the high-spin Fe^III to the catalytically active Fe^II state (i → ii in Fig. 7) is the rate-determining step leading to a higher population of the resting high-spin Fe^III species on the electrode. However, as steady-state O₂ reduction current is obtained at low potentials, the low-spin Fe^III-OOH species is observed along with a high-spin Fe^II species and an Fe^IV = O species. This implies that the decay of these species—i.e., reductive cleavage of the O-O bond of the Fe^III-OOH species (iv → v in Fig. 7), O₂ binding to Fe^II (ii → iii in Fig. 7), and reduction of Fe^IV = O (v → i in Fig. 7)—comprises rate-determining steps during O₂ reduction. At this stage, we could not find evidence of an Fe-O₂ adduct in these data. Because resonance Raman is not quantitative, the relative population of these species, i.e., the relative rates of these steps, cannot be determined using these data. However, it establishes, with confidence, the species accumulated under steady state and provides a qualitative description of the rate-determining steps in O₂ reduction by these electrocatalysts.

Fig. 7.

The oxygen-reducing mechanism by an iron porphyrin, showing the various possible intermediates. The marker bands of the individual species are given in boxes, with the ν(Fe-O) values (¹⁶O₂/¹⁸O₂) corresponding to those observed for FeEs₄.

Conclusion

A SERRS-RDE experimental setup is developed and the mechanism of O₂ reduction by a set of iron porphyrin complexes is investigated using in situ resonance Raman spectroscopy on the system under steady state. The oxidation and spin-state marker bands help identify species having the Fe in different oxidation states. Experiments using isotopically labeled O₂ enable identification of O₂-derived intermediates accumulating under steady-state conditions. This approach, which combines dynamic electrochemistry with in situ Raman spectroscopy, allows direct identification of intermediates involved in O₂ reduction by an iron porphyrin-based electrocatalyst. This technique may be used to probe electrochemical systems using synthetic as well as biological catalysts routinely used for small-molecule activation and to gain first-hand information on the reaction mechanism of these electrocatalysts. The ability to control the applied potential on the working electrode enables investigation of reaction mechanisms when the catalysis is electron transport controlled, as well as when it is mass transport controlled.

Materials and Methods

All reagents were of the highest grade commercially available and were used without further purification. Octanethiol (C₈SH), potassium hexafluorophosphate (KPF₆), and all buffers were purchased from Sigma–Aldrich. Disodium hydrogen phosphate dihydrate (Na₂HPO₄⋅2H₂O) and potassium chloride (KCl) were purchased from Merck. All solvents used were of spectroscopic grade and were purchased from Merck. Au disks for the RDE and RRDE experiments and Ag disks for SERRS experiments were purchased from Pine Instruments. ¹⁸O₂ (99%) was purchased from ICON Isotopes. All electrochemical experiments were performed using a CH Instruments model CHI710D electrochemical analyzer. Bipotentiostat reference electrodes were purchased from CH Instruments. The RRDE setup from Pine Research Instrumentation (E6 Series ChangeDisk tips with AFE6M rotor) was used to obtain the RRDE data. SERRS-RDE data were collected using a Trivista 555 spectrograph (Princeton Instruments) and 413.1-nm excitation from a Kr⁺ laser (Coherent Innova Sabre SBRC-DBW-K), rotating the electrode at 200 rpm unless noted otherwise.

Freshly cleaned Au and roughened Ag disks were immersed in an ethanolic solution of C₈SH (0.4 mM) for 24 h. A dilute solution of the catalyst (1 mM) in chloroform was dropped on these electrodes and kept for 30 min. The unabsorbed or loosely absorbed catalyst was washed off using CHCl₃. The electrochemical experiments were done in Au electrodes, and the SERRS-RDE experiments were performed on roughened Ag disks, because SERRS cannot be performed on Au surfaces at <600 nm and the rR experiments on porphyrin complexes need excitation near 400 nm. All these surfaces show similar coverage and O₂ reduction currents. It is assumed that the spectroscopic data obtained on SAM-covered Ag surfaces are valid for SAM-covered Au surfaces as well.