Direct Quantum Yield Measurement of a Ligand-to-Metal Charge Transfer Photocatalyst via Cyclic Voltammetry Authors: Nya E. Black, Dara Cheng, Brandon P. Grasty, Dylan G. Boucher Published: 2026 ABSTRACT Measurements of molecular quantum yields are crucial for understanding and directing photocatalytic activation of inert substrates. However, present methods to study photocatalysts are limited by high costs, disparate measurement and photocatalytic reaction time scales, or poor luminescent behavior. Electroanalytical methods such as cyclic voltammetry (CV) are a promising orthogonal measurement to address these limitations. Here, CV was used to directly measure the molecular quantum yield of a model iron chloride ligand-to-metal charge transfer (LMCT) photocatalyst. Cyclic voltammograms of the LMCT photocatalyst exhibited catalytic currents that were dependent on light intensity. Correlation between light intensity and catalytic current was used to derive a light-dependent plateau equation which enables the direct measurement of quantum yield from CV. Finite element simulations further supported the quantum yield measurements. Control experiments and simulations incorporating photothermal convection indicated minimal thermal interference on the measured quantum yield. We used our quantum yield measurement to extract quantitative information about the follow up C–H activation step via the transient chlorine radical. Overall, these measurements establish cyclic voltammetry as a quantitative method for determining photocatalytic quantum yields. This work enables a rapid, simple, and orthogonal approach to study molecular photocatalysts. Introduction Molecular photocatalytic charge transfer utilizes energy from light to drive challenging redox reactions. Through the photogeneration of highly reducing or oxidizing intermediates, the field of photoredox catalysis has enabled the activation of a wide array of inert substrates. − In particular, light-driven bond homolysis from ligand-to-metal charge transfer (LMCT) has become a synthetically valuable strategy where reactive radical intermediates are generated in proximity to catalytic metal centers allowing cooperative bond-forming reactions. − While a broad understanding of LMCT photocatalysis has been known for many decades, directly relating excited state dynamics to synthetic outcomes remains a significant challenge , (Figure a). Phenomena such as radical complexation, , fast internal relaxation, or ligand-centered excited state quenching dramatically alter the kinetic and thermodynamic landscape of LMCT photocatalysis. Further, many first-row transition metals utilized in LMCT photocatalysis feature either picosecond rates of bond homolysis or poorly luminescent excited states, thus preventing simple characterization of photodynamics from Stern–Volmer emission quenching. , These characteristics of LMCT photocatalysts complicate the understanding of their kinetics and downstream synthetic utility. (a) Ligand-to-metal charge transfer (LMCT) photocatalytic reactions have been predominantly studied using ultrafast spectroscopy and synthetic product analysis methods. (b) An overview of the work performed here: cyclic voltammetry (CV) was used to directly measure the quantum yield of a LMCT photocatalyst. Conventional methods to study photocatalytic reactions generally involve spectroscopic methods and product analysis. Ultrafast spectroscopic techniques, such as transient absorption, are used to monitor catalyst optical properties on a femtosecond to nanosecond time scale providing information about excited-state lifetimes, quenching rates, and quantum yield of LMCT bond homolysis (ΦHom). , However, this technique is unable to capture slower catalytic processes and requires expensive equipment and highly specialized training, which limits widespread adoption. Alternatively, synthetic product analysis is used to construct photocatalytic mechanisms from chemical intuition and product distributions. While useful for optimizing synthetic conditions, this approach does not capture the dynamic nature of the photocatalytic cycle and relies on extrapolating known mechanisms to novel chemistry. Additionally, the relatively low quantum yield of common LMCT photocatalysts leads to synthetic processes that take hours to days, thereby increasing time and material costs. An alternative method for studying LMCT photocatalysts that is rapid, accessible, and provides a more comprehensive understanding of the entire photocatalytic process would advance the development of more efficient and selective catalysts. Emerging interest in electro-photocatalytic (e-PC) synthetic methods ,− suggests electrochemistry could be used as an orthogonal mechanistic tool for studying photocatalysis. Electroanalytical methods, such as cyclic voltammetry (CV), provide quantitative information about molecular catalytic mechanisms. CV could provide complementary mechanistic information about photocatalysts to support existing measurements through real-time analyses, interrogation of specific kinetic steps, simplified measurements of nonluminescent catalysts, and rapid optimization and screening for downstream applications. However, establishing methods to perform direct electrochemical measurements of photocatalyst reaction rates via CV have been hampered by the lack of an ideally behaving model system. The first reported use of CV to analyze photocatalysis was in the 1980s where Lund and Carlsson measured the photoreduction of aryl chlorides via pyrene. Ultimately a precise quantitative description of photocatalytic behavior was obfuscated by photothermal interference of the catalytic current. In 2020, a quantitative framework for e-PC was proposed by Costentin et al. This theoretical work provided the conceptual basis to determine photocatalytic reaction kinetics from CV while also differentiating catalytic current from photothermal effects. Relatedly, Cossairt et al., and later Dempsey et al., estimated the rate of photoinduced charge transfer from quantum dots to a redox mediator via CV but were unable to directly quantify excited state photocatalyst concentrations due to the complexity of quantum dot photoexcitation. Recently, Fantin et al. explored CV measurements of organocatalytic photoreductants and provided a useful approach for characterizing and quantifying photochemical electron transfer rates via CV. However, the short-lived excited states of the catalysts precluded a direct relationship between observed current increases, light intensity, and photocatalytic rates, thus necessitating the use of numerical simulations to fit voltammetric behavior. Presently, a simple and direct measurement of photocatalyst kinetics with CV has not been achieved due to the lack of an appropriate model system. For this purpose, we chose a simple iron chloride (hereafter Fe­(II/III)­Cl x ) model system to measure LMCT e-PC behavior via CV (Figure b). Fe­(III)­Cl3 ([Fe­(III)­Cl4]− in chloride-containing solution) is one of the most prevalent photocatalysts for visible light-induced homolysis as it undergoes first order homolytic photoelimination to generate a chlorine radical. The chlorine radical is subsequently capable of activating a wide array of C–H bonds. ,,, In this work, CVs of Fe­(II/III)­Cl x exhibited ideal catalytic behavior and enabled a straightforward analysis of the catalytic current. CV measurements of Fe­(II/III)­Cl x under varying light intensities demonstrated nearly ideal sigmoidal catalytic CV behavior, allowing us to directly quantify photocatalyst kinetics. From these CVs, we derived a light-dependent current plateau expression which we used to directly calculate the quantum yield (ΦCV) of the Fe­(II/III)­Cl x photochemical pathway from Fe­(II) turnover at the electrode. Finite element simulations and control experiments quantitatively support the measurement. We also utilized the measurement to extract kinetic information about the transient chlorine radical. Finally, we extended our measurement to a cerium chloride (Ce­(III/IV)­Cl x ) LMCT photocatalyst to demonstrate the generality of our approach. Overall, these measurements establish CV as a simple, practical technique to study the photocatalytic mechanisms of LMCT photocatalysts. Experimental Materials and Methods Detailed experimental procedures, simulation parameters, and analyses are available in the Supporting Information. Electro-Photochemistry A three-electrode electrochemical cell with a glassy carbon working electrode (CH Instruments), Ag wire pseudoreference electrode (CH Instruments), and platinum mesh counter electrode (Strem Chemicals) was used for CV experiments. The electrochemical cell was positioned so the base was not touching the glass of a Pyrex dish filled with 240 mL of water (Figure S1). This water bath helped to minimize bulk temperature changes of the solution. A 390 nm light emitting diode (LED) (Kessil PR160L) was used for illumination conditions, and a lab stand and clamp were used to hold the LED. The electrochemical cell was positioned so that the working electrode was 1.905 cm (0.75 in) above the center of the LED. Electrolyte Solutions The Fe­(II/III)­Cl x redox couple was studied using 1 mM iron dichloride (FeCl2, Strem Chemicals) in 0.1 M tetrabutylammonium chloride (TBACl, Sigma-Aldrich) in acetonitrile (MeCN, Sigma-Aldrich). TBACl was used as the supporting electrolyte because it provided the necessary chloride anions for the electron transfer at the electrode to occur. Additionally, all solutions were prepared and stored in a glovebox. CV measurements were performed outside the glovebox after the solution was sparged with N2. N2 was kept continuously flowing throughout the headspace during measurements. The control solution was prepared with 1 mM FeCl2 in 0.1 M tetraethylammonium chloride (TEACl, Sigma-Aldrich) in MeCN, and a general ferrocene (Fc) photothermal control was prepared with 1 mM sublimed Fc (Sigma-Aldrich) in 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6, Sigma-Aldrich) in MeCN. The Ce­(III/IV)­Cl x redox couple was studied with analogous conditions to the Fe­(II/III)­Cl x except 1 mM cerium trichloride (CeCl3, Thermo Fisher Scientific) was used for cerium-containing solutions. UV–vis Spectroscopy and Photoexcitation Rate Constant Spectrophotometric measurements were conducted using a Cary 60 UV–vis spectrophotometer (Agilent). Samples were placed in 1 cm path length quartz cuvette. The blank contained 0.1 M TBACl in MeCN, and UV–vis was used to calculate the molar absorptivity of [FeCl4]− at 390 nm. Spectra were taken of 100 μM, 50 μM, 25 μM, 12.5 μM, and 6.25 μM FeCl3 in 0.1 M TBACl in MeCN. Beer’s law was used to calculate the molar absorptivity (ε) from the slope of the absorbance at 390 nm vs the concentration of [FeCl4]−. The photoexcitation rate constant σI (units of s–1) was calculated by converting the molar absorptivity at 390 nm (ε) to absorption cross section (σ) multiplied by the photon flux (I) which was measured by a calibrated photodiode set to measure power (mW) at 390 nm. Irradiance (optical power per unit area) was obtained by dividing the power values by the 0.79 cm2 detector area, and irradiance was converted from mW cm–2 to photon flux (cm–2 s–1). Simulations All cyclic voltammetry simulations were conducted with COMSOL Multiphysics 6.2. COMSOL uses the finite element method to compute numerical solutions for systems of partial differential equations. A time-dependent solver was utilized to compile and solve the equation system. The transport of diluted species module was used to model mass transport, homogeneous reactions, and electrode kinetics. The python library MPh (1.1.5) was used to interface COMSOL with Python, allowing for iterative simulations of high volumes of variable combinations. Results and Discussion Cyclic Voltammetry Measurement of Photocatalysis in a Fe­(II/III)­Cl x Model System To establish a CV measurement of photocatalysis, the e-PC cycle of the Fe­(II/III)­Cl x redox couple was examined under illuminated and nonilluminated conditions. Under dark conditions, CVs demonstrated a reversible one-electron oxidation of [Fe­(II)­Cl3]− to [Fe­(III)­Cl4]− at −0.403 V vs Fc+/0 (Figure S2). Since the LMCT transition of [Fe­(III)­Cl4]− is centered at 360 nm and ranges from 340 nm – 400 nm (Figure S3), a 390 nm LED was chosen for photoexcitation. Upon illumination with 390 nm light, the anodic current increased as a function of light intensity yielding a sigmoidal CV trace with a diminished return wave (Figure a). Peak heights of CVs under illumination demonstrated sensitivity to the scan rate. However, unlike the dark conditions, the forward anodic current decayed to a consistent plateau even at higher scan rates (Figure S4). Taken together, these results indicate photocatalytic conditions analogous to a homogeneous electrocatalytic reaction under kinetic control. Under these conditions higher light intensity, rather than substrate concentration, increases the catalytic current. (a) Cyclic voltammetry depicting the catalytic current dependence on irradiance of 1 mM FeCl2 in 0.1 M TBACl in MeCN under 390 nm LED illumination at a 25 mV/s scan rate. The irradiances at 39.9 mW cm–2, 92.3 mW cm–2, 141 mW cm–2, and 186 mW cm–2 correspond to photon fluxes of 7.83E16, 1.81E17, 2.76E17, and 3.65E17 cm–2 s–1. (b) Plot of ic/ip against υ –1/2 where individual data points are an average of triplicate measurements with the associated standard error. The slope of each line is used to determine k obs at different irradiances. (c) A schematic of the e-PC reaction mechanism for the Fe­(II/III)­Cl x redox couple at the electrode. In the zone diagram formalism developed by Savéant et al., mechanisms of homogeneous redox catalysis are represented by CV shapes classified into separate regions (zones) that represent different limiting features (diffusion, catalysis kinetics, substrate consumption) during a CV experiment. The axes indicate how a system transitions from one CV shape to another by changing variables such as concentration or scan rate, and it is useful for characterizing reaction mechanisms and determining the kinetics of the system. For the Fe­(II/III)­Cl x system, the changes in CV shape with increasing light intensity are consistent with a transition from no catalysis (zone D) through no substrate consumption (zone KD) to pure kinetic conditions, no substrate consumption (zone KS) with increasing light intensity. The current increases can therefore be described by the following plateau expression for a homogeneous electrocatalytic reaction: 1 icip=10.446RTnFvkobs where i c is the current of the catalytic plateau, i p is the current of the corresponding redox peak under nonilluminated conditions, R is the gas constant, T is temperature, n is the number of electrons transferred at the electrode, F is Faraday’s constant, v is the scan rate, and k obs is the observed first-order rate constant. The observed rate, k obs , provides an experimental handle to determine the kinetic order and overall rate dependence on reaction conditions to diagnose reaction mechanisms. To use the ic/ip analysis, CVs must be S-shaped with no peak as defined by CVs in the KS region of the zone diagram, or they must exhibit scan rate-independent limiting currents which can be achieved in the KD region. Since the CVs of Fe­(II/III)­Cl x met the criteria required to use the ic/ip analysis, the k obs dependence on light intensity was determined by running CVs at 25 mV/s, 50 mV/s, 100 mV/s, 250 mV/s, and 500 mV/s with 39.9 mW cm–2, 92.3 mW cm–2, 141 mW cm–2, and 186 mW cm–2 irradiance. The irradiance measurements corresponded to photon fluxes of 7.83E16, 1.81E17, 2.76E17, and 3.65E17 cm–2 s–1 which were measured using a photodiode, and a more detailed explanation of photodiode measurements is provided in the SI. Triplicate measurements were obtained from different solutions to account for day-to-day variability. Eq was used to determine k obs for each light intensity by plotting ic/ip vs v –1/2. k obs values from 39.9 mW cm–2 to 186 mW cm–2 irradiance were 0.06 ± 0.02 s–1, 0.317 ± 0.006 s–1, 0.607 ± 0.007 s–1, and 0.824 ± 0.006 s–1, respectively (Figure b). Plotting the k obs values against the photon flux demonstrated a linear dependence of k obs on photon flux for the range of intensities used in the paper (Figure S5). In a similar fashion, the current increased linearly with [Fe­(II)­Cl3]− concentration, indicating an overall first order dependence (Figure S6). The normalized current provided a k obs that was independent of [Fe­(II)­Cl3]− and TBACl concentrations (Figures S7, S8), indicating a zero-order k obs dependence on the catalyst or substrate as expected from eq . The CVs of the Fe­(II/III)­Cl x redox couple are indicative of the e-PC cycle depicted in Figure c, which is consistent with synthetic studies on similar systems. The reversibility of the dark CVs at all scan rates indicated that the chloride transfer upon oxidation of [Fe­(II)­Cl3]− was fast relative to the scan rates used (up to 500 mV/s). The catalytic current plateaus and loss of the return wave suggested that the 390 nm light generated an excited state [Fe­(III)­Cl4]− that underwent homolytic photoelimination of a chloride via LMCT to generate a chlorine radical and regenerate [Fe­(II)­Cl3]− . The catalytic current was dependent on light intensity which implied that the radical generating LMCT process was involved in the rate-limiting step, consistent with photolysis studies on other chlorine radical photocatalysts. We hypothesized that generated chlorine radicals were being scavenged by hydrogen atom transfer (HAT) from either the TBACl supporting electrolyte or the MeCN solvent. Thus, control experiments for photocatalysis were employed using either 0.1 M tetraethylammonium chloride (TEACl) or 0.1 M tetraphenylphosphonium chloride (PPh4Cl) in MeCN. Since the transient chlorine radical generated after LMCT can undergo H atom abstraction and impact the overall rate of turnover, the control experiments were designed to contain electrolytes with shorter alkyl groups (TEACl) than TBACl or no sp 3 C–H bonds (PPh4Cl) to suppress H atom abstraction and limit turnover by forcing recombination of the chlorine radical with Fe­(II)­Cl x . Both supporting electrolytes exhibited the same CV behavior as TBACl under dark conditions, but upon illumination, CVs of Fe­(II)­Cl x in TEACl and PPh4Cl exhibited minimal current increases indicating that the e-PC cycle could not turnover under these conditions (Figure S9). These results, taken alongside the relative independence of k obs on TBACl concentration, suggested that 1) chlorine radicals were being efficiently scavenged by TBACl and 2) HAT from MeCN does not occur appreciably on the time scale of the CV. Thus, the light-dependent LMCT step can be taken to be rate-limiting in these conditions. To verify that photocatalysis occurs under 390 nm illumination, CVs illuminated by different wavelengths of light were performed. CVs with 427, 456, or 525 nm light demonstrated little to no increases in current (Figure S10). In contrast, 370 nm, which is closer to the maximum absorbance value for [Fe­(III)­Cl4]−, produced a greater current than those illuminated at 390 nm (Figure S11). However, shorter wavelengths of light are more likely to initiate side reactions which could convolute the catalytic contribution to the current. Consequently, 390 nm light was used for subsequent analyses to produce the excited state while minimizing side reactions. Overall, these studies supported the [Fe­(III)­Cl4]− photocatalytic mechanism where specific wavelengths of light in its absorbance spectrum must be used to produce the excited state. The e-PC mechanism for the Fe­(II/III)­Cl x model system is governed by the following rates: heterogeneous electron transfer (k 0 ), photoexcitation (σI), excited-state relaxation (k b ), and radical generating turnover (k Hom ). As mentioned previously, the nonilluminated CVs indicated that the chloride addition is fast and not rate limiting, and thus it is appropriate to treat the heterogeneous electron transfer and chloride addition as a single step. The second step describes the photoexcitation of [Fe­(III)­Cl4]− to its excited state. Similar to the theoretical work performed by Costentin et al., we can incorporate the light intensity into a first order rate constant. Our treatment of the first order rate constant differs from expressions used in previous analyses as we convert the decadic molar absorptivity (ε) to absorbance cross section (σ) to obtain the first order rate constant, σI. σ is the absorbance cross section of [Fe­(III)­Cl4]− (units of cm2) which represents the probability of catalyst light absorption and is more appropriate than the decadic ε to account for a per molecule photon absorption event. I is the photon flux (units of cm–2 s–1) which describes the rate of photons incident to the electrode surface. We note that the [Fe­(III)­Cl4]− concentration is localized to the electrode surface, therefore we assume the local light intensity throughout the diffusion layer is effectively equal to the incident intensity. Taken together, σI is the rate of molecular photoexcitation with units of s–1. This definition of the photoexcitation rate constant provides a method to experimentally quantify light, as σ was obtained using the ε of [Fe­(III)­Cl4]− measured from UV–vis (Figure S3) and I was measured via a calibrated photodiode (Table S1). As Fe­(III) photocatalysts have reported excited states ranging from the picosecond to nanosecond time scale, which is significantly shorter than the time scale of the CV scan, the excitation step must include a reverse reaction. k b is the inverse lifetime of the photoexcited state and accounts for the relaxation of the photoexcited species and thus has units of s–1. The sigmoidal shape of the CVs during illumination indicates that the light excitation step is followed by a turnover step, with rate constant k Hom that is unimolecular in this mechanism and thus has units of s–1. The Fe­(II/III)­Cl x model system can be described by this simplified kinetic model because k obs exhibits a zero-order dependence on TBACl and [Fe­(II)­Cl3]− concentration, strongly supporting the LMCT step as rate-limiting. A more comprehensive treatment of the kinetics, including chlorine radical recombination and hydrogen atom transfer is included in section . The rate constants in the proposed mechanism (Scheme ) can be mathematically related to the plateau current of the CV. To do so we follow the general framework established by Savéant as well as a similar derivation for bimolecular systems from Costentin et al. For conciseness, we present a simplified algebraic derivation based on steady-state approximation. A full derivation using dimensionless parameters and diffusion competition is available in the SI. When the current reaches its limiting value, the system is assumed to be at steady state and independent of heterogeneous electron transfer kinetics. In other words, at potentials sufficient to oxidize [Fe­(II)­Cl3]−, the concentrations of all electroactive species are constant at the electrode surface. Furthermore, the high relaxation rate indicates that the concentration of the excited-state species near the electrode is constant and small. Under these assumptions, the rate of concentration changes of the excited species, [Fe­(III)]*, at the electrode is zero. Thus, we rearrange the rate laws to provide an expression for the excited-state concentration based on the kinetics of light absorption, relaxation, and ground-state concentration: 2 [Fe(III)]*=σIkb+kHom[Fe(III)] Simplified Electro-Photocatalytic Rates and Mechanism As the rate of turnover is limited by the excited-state concentration, we express the rate of turnover in terms of the catalytic rate constant and the excited-state concentration: 3 Rate of Turnover=kHom[Fe(III)]*=σIkHomkb+kHom[Fe(III)] This relationship demonstrates that the overall reaction kinetics are best described by the ratio kHomkb+kHom , which is the definition of the quantum yield (ΦCV) for this photocatalytic process. , Effectively, this condenses the three-step mechanism into a two-step EC′ reaction, allowing us to derive an overall first-order rate constant: 4 kobs=σIkHomkb+kHom=σIΦCV Since the limiting CV current provides the k obs values required for a direct measurement, steady-state conditions must hold. Thus, the observed rate of the e-PC reaction, and consequently, the catalytic current of the voltammogram, is dependent on the LMCT quantum yield when the e-PC reaction is light-limited as in the TBACl model system. Finally, by combining eqs and we arrive at an expression that describes the plateau current’s dependence on photon flux (I), the catalyst’s absorbance cross section (σ), and the quantum yield (ΦCV): 5 icip=10.446RTnFvσIΦCV Equation enables the direct quantification of photoredox quantum yield (i.e., catalyst turnover) via CV. Since eq utilizes k obs values, methods such as chronoamperometry, rotating disk electrochemistry, and foot-of-the-wave analyses that obtain related k obs values suggest possible applications of the measurement to other techniques beyond CV in the future. We note, however, that systems where current is limited by mass-transfer contributions are not as straightforward as the case outlined here. The relationship between k obs and the photoexcitation rate constant demonstrated by eq was utilized to determine the quantum yield of e-PC for the Fe­(II/III)­Cl x redox couple. Experimentally derived k obs values from different light intensities were plotted against the measured rate of photoexcitation (σI). Figure shows a linear correlation with a slope of 0.39 ± 0.07, equivalent to an 39% quantum yield that was independent of TBACl concentration (Figure S12), which is consistent with the nonunity quantum yields reported for these reactions. , We note that the transient absorbance measurements for ΦHom by other researchers directly report on the [Fe­(III)­Cl4]−* bond homolysis whereas the quantum yield determined by CV (ΦCV) in our work reports on the efficiency of Fe­(II) turnover from the rate-limiting step. We emphasize our electrochemical measurement may be impacted by a number of factors not present in homogeneous reactions such as 1) low concentrations of Fe­(II) at the electrode surface which would suppress recombination, 2) more significant ion-pairing and cage-effects due to supporting electrolyte, and 3) difficulty in accurately measuring light intensity at the electrode surface where catalysis takes place. Quantum yield determination by plotting electrochemically determined k obs against the photoexcitation rate constant (absorption cross section and photon flux) where the ΦCV = 0.39 ± 0.07. Although quantum yields obtained through transient absorbance spectroscopy and CV are probing fundamentally different processes, the aggregate ΦCV value is a practical metric that can be used when comparing relative LMCT efficiencies under different rate-limiting conditions or to obtain quantum yield values for poorly luminescent catalysts. We also acknowledge that ΦCV could be affected by radical chain propagation after the chlorine radical is generated which would contribute nonphotocatalytic increases to the current. Since CVs of Fe­(II/III)­Cl x return to the expected CV shape after illumination and catalytic currents are not impacted by the inclusion of a radical trap (BrCCl3, Figure S13), we conclude that radical chain propagation does not contribute significantly to the rate of Fe­(II) turnover on the time scale of the CV experiments. Costentin et al. arrived at a similar expression to eq (for bimolecular catalysis), however they did not frame the equation in terms of quantum yield. Fantin et al. determined the correlation between quantum yield and the light intensity dependent step of the reaction but were only able to estimate the value of the quantum yield with very few excitation events leading to a productive reaction in their system. Thus, this work demonstrates the first successful direct measurement of molecular quantum yield using CV. Finite element simulations of the three-step model (Scheme ) were conducted to support our proposed mechanism and quantum yield measurement. To start, voltammograms at varying combinations of k b and k Hom were compared to the experimental current (Figure a). The dashed white line represents values that provided approximately zero difference between experimental and simulated current. Consistent with the derived expression (eq ), a ratio of k b and k Hom produced results that accurately reflected the experimental data. Experimental data was fit according to kHomkb+kHom (i.e., ΦCV). ΦCV was incorporated into the model by rearranging the quantum yield definition and writing k Hom in terms of ΦCV and k b . k b was fixed at 1 × 109 s–1 which is consistent with experimentally determined nanosecond lifetimes of Fe­(III) complexes. Thus, ΦCV was the only adjustable parameter for fitting simulations to experimental data. (a) Color plot of the simulated catalytic current’s accuracy at varying k b and k Hom . The dashed white line indicates the regions of best fit (smallest difference between experimental catalytic current and simulated current). The simulated scan rate was set to 50 mV/s, and the simulated light intensity was set to 186 mW cm–2 to reflect the experimental CV of a solution of 1 mM FeCl2 in 0.1 M TBACl in MeCN. (b) Comparison of light dependency at 50 mV/s of simulated CVs (dashed) and experimental CVs (solid). Simulated quantum yield was 0.39 for all irradiances. (c) The normalized plateau current of simulated CV’s plotted against the square root of σI for varying quantum yields. The experimental points (solid dots) were taken at 50 mV/s for solutions of 1 mM FeCl2 in 0.1 M TBACl in MeCN. The fitted quantum yield was determined by tuning the parameter ΦCV until the simulated plateau currents were within 5% of the experimental current (Figure b). A consistent quantum yield of 0.39 was obtained across all light irradiances, demonstrating that ΦCV was independent of irradiance and consistent with the experimental value. Furthermore, a ΦCV of 0.37 was a better fit across all scan rates and was still within error of our experimentally determined quantum yield of 0.39 ± 0.07. The agreement of the fitted quantum yield to the experimental value supports that we can isolate the quantum yield from the observed rate constant. We performed additional simulations to support the derived light-dependent plateau equation (eq ). To examine the model’s dependence on irradiance, we simulated CVs at different quantum yields. The normalized plateau currents of simulated CVs were plotted against the square root of σI (Figure c), as suggested by eq . The experimental normalized currents (solid dots) fall on the line corresponding to the fitted quantum yield 0.39, which demonstrated that the model captured the catalytic current’s square root dependency on σI. Furthermore, the simulations predicted that the system continues to display square root dependency on light at higher quantum yields and σI. These simulations support the validity of the light-dependent plateau expression. The model’s sensitivity to k b was probed by simulating varying magnitudes of k b while holding ΦCV constant at 0.37 (Figure S14). At k b values greater than 100 s–1, the model is insensitive to changes in k b when a constant value for ΦCV is maintained. The model’s insensitivity indicated that k b must be faster than 100 s–1 to maintain a low concentration of the excited species which allows for strict kinetic control. The predicted k b falls around 109 s–1, based off reported lifetimes of the excited states. Therefore, the k b of our system is sufficiently fast, fulfilling the assumption that the concentration of the excited species must be constant and small (Figure S15). Similarly, the simulated concentration profiles of the oxidized species and excited species at the electrode surface were constant (at potentials sufficient to oxidize [Fe­(II)­Cl3]−, Figures S16, S17), confirming the steady-state assumption of our derivation. In summary, these simulations support the direct measurement of quantum yield enabled by eq and support the assumptions made in its derivation. While the simulated CVs accurately reflected the experimental plateau current across all scan rates, the model failed to capture the peak current magnitude of the experimental CVs at higher scan rates (Figure S18). The magnitude of the peak is determined by the competition between mass transport (i.e., diffusion or convection) and kinetic control. At higher scan rates the potential is scanned more rapidly than the catalytic reaction proceeds, resulting in mass transport governing the current before kinetic control is reached. Although the peak current does not provide quantitative information about catalysis, it serves as a baseline for modeling the competition between catalysis and mass transport. In the simulated CVs, diffusion and the photocatalytic reaction are the only replenishing sources of [Fe­(II)­Cl3]− at the electrode surface, thus the mismatch in the peak currents between experiment and simulation at higher scan rates indicated that there was another factor contributing to overall CV behavior (vide infra). Photothermal Convection Correction Considering the rate of relaxation k b is larger than the turnover step k Hom , most absorbed photons lead to nonradiative decay, releasing thermal energy near the electrode where the oxidized species, [Fe­(III)­Cl4]−, is concentrated. As noted by others, ,, the resultant temperature gradient at the electrode surface can lead to photothermally induced convection. Convection replenishes [Fe­(II)­Cl3]− near the electrode, which obfuscates catalytic contributions to the current. While illuminated CV experiments were performed in a water bath to minimize bulk temperature increases (Table S2), the localized molecular heating is not fully eliminated by a water bath. Therefore, to quantify the rate of catalysis more accurately, it was necessary to measure the convective effects of the system. Convection contributions to the CVs were investigated using the TEACl control system discussed in section , which exhibited more reliable solubility than the PPh4Cl electrolyte. Since it demonstrated no photocatalysis but otherwise retained the same properties as the Fe­(II/III)­Cl x model system, it would allow for comparable electrochemical and photophysical properties without catalysis, providing a control experiment to assess photothermal convection. Since CVs of the TEACl control system under illuminated conditions were scan rate dependent even with higher concentrations of TEACl (Figure S19), the changes in the CVs could be mostly attributed to photothermal effects rather than catalysis. The quantum yield was estimated as 0.024 ± 0.009 for the TEACl control system (Figure S20), which is sufficiently lower than with TBACl. However, we emphasize that this quantum yield measurement is an estimation only as it is highly convoluted by thermal contributions to the current. Nonetheless, the quantum yield value for the Fe­(II/III)­Cl x system in TEACl compared to TBACl corroborate the conclusion that there is minimal photocatalysis in the control system. We next modeled convection in our experimental data with finite element simulations. In the absence of convection and catalysis, redox molecules are under semi-infinite linear diffusion conditions (Figure a). The introduction of convection shifts solution conditions toward finite linear diffusion. This effect is typically modeled as a hydrostatic layer at the electrode surface (Figure b). Within this layer, molecular species are free to diffuse to the electrode surface while outside of this layer, the concentration of the redox couple is replenished to its bulk value by convection. The thickness of this diffusion layer (δ) is controlled by the amount of convection (higher convection, thinner layer and vice versa). Thus, the effects of convection were simulated by altering the length of δ for a finite element model that included only electron transfer at the electrode and no catalysis. δ was adjusted so that the simulated limiting currents matched those of the TEACl data (Figure c). CVs of the system with TEACl supporting electrolyte at 50 mV/s were chosen to capture convective effects. Because δ is intensity-dependent, each irradiance had a corresponding δ value. The δ values obtained for each irradiance were 0.014, 0.0115, 0.0092, and 0.0082 cm, respectively. (a) Concentration profile of [Fe­(II)­Cl3]− under semi-infinite diffusion conditions. The x-axis represents the distance from the electrode, and the y-axis represents the concentration of [Fe­(II)­Cl3]−. (b) Concentration profile of [Fe­(II)­Cl3]− where the diffusion width is compressed, modeling the effects of convection. (c) Comparison of simulated CVs (dashed) at varying intensity and experimental CVs of 1 mM FeCl2 in 0.1 M TEACl in MeCN at a 50 mV/s scan rate. The simulated CVs were fitted to the experimental CVs by compressing the width of the diffusion layer (δ). A δ was determined for each irradiance, with dark blue corresponding to 39.9 mW cm–2, light blue 92.3 mW cm–2, orange 141 mW cm–2, and red 186 mW cm–2. (d) Color plot depicting the accuracy of simulated catalytic current of [Fe­(II)­Cl3]− in TBACl at different combinations of quantum yield (ΦCV) and δ. Dark blue indicates the smallest difference between simulated and experimental current. The dashed line with the TEACl label indicates the diffusion layer obtained from TEACl control of δ = 0.0082 cm. The dashed line with ferrocene label indicates the diffusion layer obtained from ferrocene control of δ = 0.014 cm. Simulated CVs were set to 50 mV/s at 186 mW cm–2 to reflect experimental parameters. (e) Concentration profile of [Fe­(II)­Cl3]− under catalytic conditions in which the catalysis layer (μ) is significantly smaller than δ. While experiments with TEACl as a supporting electrolyte are useful to establish a baseline estimation of photothermal convection, they do not completely exclude slow catalytic contributions to current enhancements. To establish a general photothermal control, ferrocene (Fc) was utilized as it absorbs light at a similar wavelength to [Fe­(III)­Cl4]−. CVs of Fc were run under dark and illuminated conditions with TBAPF6 as supporting electrolyte, and δ was then simulated in the same fashion as above (Figure S21). Since the ferrocene system is not capable of photocatalysis, the differences between the CVs under illuminated and nonilluminated conditions could be completely attributed to photothermal effects. Taken together, the TEACl data and the Fc data provide a range for δ, which at 50 mV/s and at 186 mW cm–2 was 0.0082 cm ([Fe­(III)­Cl4]−) to 0.014 cm (Fc). The competition between photocatalysis and convection was probed by exhaustively simulating CVs for the catalytic TBACl system at different combinations of quantum yield and diffusion layer thickness, δ. A comparison of the simulated plateau current to the experimental plateau current revealed the pairs of ΦCV and δ that produced a dark blue region corresponding to the correct plateau current (Figure d). The plateau current is convection-controlled at small δ. Consequently, slight changes in δ induce dramatic differences in the determined quantum yield. In contrast at sufficiently large δ, the current becomes quantum yield-controlled and is insensitive to further increases in δ. In essence, all δ within this region produce the same steady-state current response when paired with the same ΦCV indicating minimal variations in photothermal current contributions. To reach this level of insensitivity to δ, the catalytic reaction must be sufficiently faster than convection. To compare the respective magnitudes of catalysis and convection, the catalytic rate can be parametrized as a catalytic reaction layer. The catalytic reaction is contained within a thin layer at the electrode surface, with an estimated width, μ=DσIΦCV . , At 50 mV/s and 186 mW cm–2 irradiance, the stabilization of the quantum yield is observed when the catalysis layer is approximately half the thickness of the convective diffusion layer (Figure e). We also note that the exhaustive experimental-simulated comparison of quantum yield and convection captures the behavior predicted by Costentin et al., who derived a hyperbolic tangent dependence on the ratio between the convection layer and the catalysis layer. Finally, the experimental interval of diffusion layer thickness provided by the TEACl and ferrocene controls, indicated by the dashed lines on Figure d, falls within the quantum yield stable region. This indicates that the catalytic reaction is sufficiently fast, and thus the CV current is tolerant to small fluctuations in photothermal convection. The average between the ferrocene δ and TEACl δ for each intensity was integrated into the catalytic three-step model, and the ΦCV was adjusted for CVs taken in TBACl (Figure a). While the currents observed in the TEACl control system under illumination could be due to background activation of MeCN, the consistent shape between simulated and experimental CVs when photothermal corrections are implemented indicates that convection is the dominant contribution and background reactivity is below the limit of detection for the Fe­(II/III)­Cl x system. A photothermally corrected quantum yield of 0.29 was obtained across all irradiances and scan rate permutations which remains close to the experimentally obtained quantum yield of 0.39 ± 0.07. The fitted quantum yield retained its intrinsic properties, demonstrating that our photothermal correction does not interfere with the ability to isolate the quantum yield. Furthermore, the incorporation of a compressed diffusion layer enabled the model to accurately capture the experimental CV behavior at all scan rates of the TBACl system (Figure b), indicating a more accurate simulation of the competition between the CV time scale and the turnover of [Fe­(II)­Cl3]−. Photothermal corrections are incorporated into the CV simulations providing a more accurate fit under all light irradiances and scan rates. (a) Comparison of the light dependence at 50 mV/s of simulated CVs (dashed) and experimental CVs (solid) when photothermal corrections were incorporated into the model by changing δ for each intensity. Simulated quantum yield was set to 0.29 for all intensities. The experimental CVs are with 1 mM FeCl2 in 0.1 M TBACl in MeCN at a 50 mV/s scan rate. (b) Comparison of scan rate dependence at 186 mW cm–2 of simulated CV with photothermal corrections (dashed) and experimental CVs (solid). Experimental CVs were with 1 mM FeCl2 in 0.1 M TBACl in MeCN under 390 nm illumination at 186 mW cm–2 irradiance. Quantification of Chlorine Radical HAT and Recombination via Cyclic Voltammetry Given our ability to resolve LMCT photocatalytic behavior, we next sought to extend our measurement to follow-up reactions. After [Fe­(III)­Cl4]− bond homolysis, the generated chlorine radical can activate C–H bonds through hydrogen atom transfer (HAT) or recombine with [Fe­(II)­Cl3]−. As conditions with the TEACl photothermal control exhibited little to no photocatalysis, we reasoned that these conditions could also be used to investigate the C–H activation step via our CV-derived quantum yield measurement with the addition of H atom donors. To examine this, cyclohexane was chosen as the H atom donor as it has readily available H atoms and is a commonly utilized substrate for C–H activation studies. Cyclohexane was added to a solution of 1 mM FeCl2 with 0.1 M TEACl in MeCN to afford a cyclohexane concentration of 0.1 M. Addition of cyclohexane yielded no change in the dark CVs, which still exhibited a reversible one-electron oxidation at −0.405 V vs Fc (Figure S22). However, upon illumination with 390 nm light, the anodic current increased with a magnitude dependent on light intensity (Figure a), thus demonstrating that the addition of cyclohexane restored photocatalysis in the TEACl control system. The quantum yield measured from these CVs was 0.24 ± 0.02 (Figure S23), which is lower than the quantum yield of the TBACl system. The differences in quantum yield between the two conditions suggested that our measurement is sensitive to the nature of the HAT step and thereby sensitive to the kinetics of the intermediate chlorine radical. (a) CVs with 0.1 M cyclohexane (C6H12) added as the substrate to 1 mM FeCl2 in 0.1 M TEACl in MeCN under 390 nm illumination at a 25 mV/s scan rate (b) CVs of cyclohexane concentrations of 0, 0.1, 0.2, 0.4, and 0.6 M in 1 mM FeCl2 in 0.1 M TEACl in MeCN under 186 mW cm–2 irradiance 390 nm illumination at a 25 mV/s scan rate. (c) k obs with lamp off, 39.9 mW cm–2, 92.3 mW cm–2, 141 mW cm–2, and 186 mW cm–2 was plotted against 0, 0.1, 0.2, 0.4, and 0.6 M cyclohexane concentrations in triplicate. (d) Quantum yield for 0, 0.1, 0.2, 0.4, and 0.6 M concentrations of cyclohexane where the error for each cyclohexane point was determined by taking the standard error of quantum yield values obtained from triplicate measurements. For concentrations of 0, 0.1, 0.2, 0.4, and 0.6 M cyclohexane-d12, ΦD = 0.024 ± 0.009, 0.223 ± 0.006, 0.282 ± 0.005, 0.380 ± 0.008, and 0.462 ± 0.006. (e) CVs of 0.6 M cyclohexane-d12 (blue) and cyclohexane (red) in 1 mM FeCl2 in 0.1 M TEACl in MeCN under 186 mW cm–2 390 nm light at a 25 mV/s. (f) Schematic showing the overall transient chlorine radical reaction kinetics where either radical recombination (k r ) or HAT is favored depending on H-donor concentration. To determine the catalytic current dependence on the H atom donor concentration, additional CVs were conducted with 0.2, 0.4, and 0.6 M cyclohexane. As the concentration of cyclohexane increased from 0 to 0.4 M, the plateau current increased (Figure b). However, the current increased minimally between 0.4 and 0.6 M cyclohexane. At all light intensities k obs increased linearly with concentration until saturating between 0.4 and 0.6 M cyclohexane concentrations (Figure c). The quantum yield followed a similar trend, increasing linearly with cyclohexane concentration and saturating between 0.4 and 0.6 M with ΦCV = 0.024 ± 0.009, 0.24 ± 0.02, 0.320 ± 0.007, 0.468 ± 0.008, and 0.52 ± 0.03 for each concentration of cyclohexane, respectively (Figure d). To further understand the impact of the HAT step on the overall kinetics, we evaluated the kinetic isotope effect (KIE) with cyclohexane-d12 under the same conditions. CVs with cyclohexane-d12 had lower catalytic plateau currents which produced overall lower quantum yields than the system with nondeuterated cyclohexane (Figure d, e). The KIE values at 0.1, 0.2, 0.4, and 0.6 M concentrations were small but sensitive to concentration with ΦH/ ΦD = 1.07, 1.14, 1.23, and 1.12. Taken together, these results are consistent with a mechanism involving an intermediate chlorine radical (Figure f). The saturation of the k obs and ΦCV between 0.4 and 0.6 M is indicative of competition between radical recombination (k r ) and HAT. This presents two k obs dependencies based on the availability of H atom donors. The first occurs when the H atom donor concentration is low. Under these conditions, the transient chlorine radical has fewer opportunities to abstract a hydrogen. Thus, radical recombination with [Fe­(II)­Cl3]− is favorable and prevents catalyst turnover. With limited H atom donors, k obs and ΦCV are dependent on H atom donor concentration. KIE values were closer to 1 at lower concentrations thus supporting a recombination limited turnover interpretation. Alternatively, between 0.4 and 0.6 M concentrations of cyclohexane, k obs and ΦCV are insensitive to increases in H atom donor concentration. At these higher concentrations, there are sufficient H atom donors for the transient chlorine radical to favor HAT. KIE measurements at these concentrations were increased relative to the low concentration KIEs thus suggesting a higher influence of HAT kinetics on the transient chlorine radical concentration. Finally, the observed rate maintained a linear dependence on light intensity at all concentrations of cyclohexane. This linear dependence indicated that the LMCT chlorine radical generating process was overall rate limiting regardless of the competition between HAT and recombination. Gratifyingly, these nonlinear rate dependencies determined by substrate concentration and KIE measurements are in excellent agreement with rate law analyses of cerium-based LMCT chlorine radical photocatalysts. Based on the complete Fe­(II/III)­Cl x mechanism, Scheme can be expanded to include the kinetics of the transient chlorine radical, k r and k HAT . Since ΦCV is a composite measurement of rates contributing to turnover in the Fe­(II/III)­Cl x system, it can be written as 6 ΦCV=ΦLMCTΦHAT where the two quantum yields for each process can be related via the kinetics of generation and consumption of the chlorine radical: 7 ΦCV=kHomkb+kHom·kHAT[D]kr[Fe(II)]+kHAT[D] where D is the concentration of H atom donors and Fe­(II) is the concentration of Fe­(II)­Cl x generated near the electrode due to bond homolysis or diffusion from the bulk solution. The full kinetic model and derivation of eq via the intermediate chlorine radical concentration is included in the SI. Eq demonstrates that ΦCV is dependent on both the LMCT and HAT rates. When the rate of HAT is sufficiently faster than the rate of recombination, differences in ΦCV values are mostly determined by the LMCT efficiency. Overall, these voltammetric experiments demonstrate that ΦCV can be used to probe specific rates within a photocatalytic system providing information about reaction mechanisms, substrate reactivity, and rate-limiting processes. The complete reaction mechanism determined by CV is shown in Figure . Our measurement is conducive to future studies of photocatalytic reactions where the simple experimental design and single analytical metric could be combined with high-throughput experimentation and analysis to accelerate the development of reactions through broad substrate scope analyses. − Proposed mechanism for the e-PC Fe­(II/III)­Cl x LMCT reaction. Extending the CV Measurement of Photocatalysis to a Ce­(III/IV)­Cl x System Since the utility of the CV quantum yield measurement was demonstrated in a Fe­(II/III)­Cl x model system, we reasoned that the measurement could be extended to other LMCT photocatalysts. For this purpose, we chose the Ce­(III/IV)­Cl x system which has demonstrated utility in C–H activation reactions due to the generation of highly reactive radical intermediates, similar to the Fe­(II/III)­Cl x mechanism. Although the excited state [Ce­(IV)­Cl6]2– can undergo a similar LMCT bond homolysis as [Fe­(III)­Cl4]−, cerium chloride complexes have a different electronic structure leading to different reactivities that would affect the ΦCV value. Electrochemical measurements were performed under analogous conditions to the Fe­(II/III)­Cl x model system where CVs were run with 1 mM CeCl3 in 0.1 M TBACl electrolyte in MeCN. Under dark conditions the CVs demonstrated a reversible one-electron oxidation of [Ce­(III)­Cl5]2– to [Ce­(IV)­Cl6]2– at −0.022 V vs Fc+/0 (Figure S24). Upon illumination with 390 nm light at 39.9 mW cm–2, 92.3 mW cm–2, 141 mW cm–2, and 186 mW cm–2, the anodic current increased as a function of light intensity, and the CVs became sigmoidal with the loss of the reverse wave (Figure a). (a) Cyclic voltammetry depicting the catalytic current dependence on irradiance of 1 mM CeCl3 in 0.1 M TBACl in MeCN under 390 nm LED illumination at a 25 mV/s scan rate. (b) Plot of i c/i p against υ –1/2 where individual data points are an average of triplicate measurements with the associated standard error. The slope of each line is used to determine k obs at different irradiances. (c) Quantum yield determination by plotting electrochemically determined k obs against the photoexcitation rate constant. The CVs demonstrated the loss of the diffusion-controlled peak with scan rate independent catalytic plateaus (Figure S25) which allowed for the ic/ip analysis to be applied. The k obs from 39.9 mW cm–2 to 186 mW cm–2 was 0.227 ± 0.004 s–1, 0.653 ± 0.007 s–1, 1.07 ± 0.01 s–1, and 1.446 ± 0.006 s–1 respectively (Figure b). Compared to the Fe­(II/III)­Cl x model system, the observed rates for Ce­(III/IV)­Cl x at each irradiance were higher. The ΦCV was determined by plotting the experimentally determined k obs against σI where the σ value was obtained using the molar absorptivity from UV–vis spectroscopy (Figure S26). The ΦCV for the Ce­(III/IV)­Cl x system was 0.28 ± 0.01 (Figure c), which was lower than the value obtained for the Fe­(II/III)­Cl x system. Overall, these results allow for a comparison between Ce­(III/IV)­Cl x and Fe­(II/III)­Cl x systems under analogous conditions. Since k obs values were higher for Ce­(III/IV)­Cl x than Fe­(II/III)­Cl x at all irradiances, the rate of turnover to regenerate Ce­(III) at the electrode was faster for the cerium chloride complex compared to regeneration of Fe­(II) at the electrode indicating that more Ce­(IV)­Cl x was being photoexcited, respectively. Although the k obs was faster for Ce­(III/IV)­Cl x , the lower quantum yield indicated it had less productive turnover per photon compared to Fe­(II/III)­Cl x . For the Ce­(III/IV)­Cl x system, pathways that led to unproductive outcomes such as relaxation or recombination were favored over homolysis and HAT reaction pathways. While we discourage interpretation of ΦCV as reporting only on the efficiency of bond homolysis, we note that our measurements are consistent with ultrafast transient absorbance measurements that Ce­(III/IV)­Cl x has generally a lower per photon LMCT efficiency than Fe­(II/III)­Cl x . Taken together, these results indicate an important distinction between the two LMCT photocatalysts where there are differences between reaction rates and efficiency. Guidelines for Utilizing Light-Dependent Electroanalytical Expressions for Photocatalysis The light-dependent plateau current expression described by eq relates the observed rate of turnover k obs to ΦCV, which enables CV as a method to obtain direct ΦCV measurements for an electro-photocatalyst. While this expression has been thoroughly demonstrated for the Fe­(II/III)­Cl x photocatalyst, the following guidelines are provided to aid its implementation across different photocatalyst families:1.The current response must be light-limited. In other words, the rate-limiting step must be downstream of the light absorption step so the impact of light intensity on k obs can be evaluated.2.Just as the original plateau expression assumes reversible, Nernstian behavior, these expressions also assume that both states of the redox couple are stable and that electrode kinetics do not contribute to the observed rates. Treatment of quasi-reversible systems may be achieved via numerical simulation.3.Only one redox state of the catalyst is responsible for photon absorption and catalysis. Starting with the photocatalytically inactive form of the redox couple in solution allows for a simplified treatment of light absorption (incident photon intensity is essentially constant at the electrode surface) and ensures that all catalysis takes place near the electrode surface. Additionally, the inactive state of the catalyst cannot strongly absorb at the same excitation wavelength which would prevent photons from reaching the photocatalyst localized at the electrode surface.4.Catalysis must be sufficiently faster than photothermally induced convection. As explored extensively in section , convection can contribute to the overall current and thus must be accounted for if catalysis is slow. We note that this point necessitates some control conditions in which catalysis does not occur or is minimized under illumination so the photothermal contributions may be quantified.5.The photocatalytic reaction must be sufficiently fast at scan rates above 20 mV/s so that the limiting current of the CV is scan-rate independent over at least an order of magnitude. The original formulation of the plateau expression necessitates that the limiting current be under kinetic control (i.e., the current is limited by reaction kinetics) and the method here relies on the same assumption. In the zone diagram formulation this is limited to zone KS and zone KD. By retaining sufficiently fast scan rates (25 mV/s and above), the catalytic current increases can mostly be attributed to catalysis rather than convection.6.The catalytic mechanism must be known to sufficient degree to simplify the mechanism to a rate-limiting EC’ expression. For example, here we demonstrate that the kinetics of chloride coordination do not limit catalysis and the rate-limiting consumption of the chlorine radical can be written as a simplified first order rate-constant. Thus, k obs is sufficient to describe the rate-limiting step.7.To interpret the kinetics and reactivity of the system determined with the quantum yield CV measurement, the overall reaction mechanisms of the system should be sufficiently understood and products identified. For the Fe­(II/III)­Cl x system, we extrapolated the products and reactivities under electrochemical conditions using previously published reaction mechanisms and products that have been studied by other groups. , For novel systems or to determine the exact reaction products under electrochemical conditions, controlled potential electrolysis under illumination should be used. 1.The current response must be light-limited. In other words, the rate-limiting step must be downstream of the light absorption step so the impact of light intensity on k obs can be evaluated. The current response must be light-limited. In other words, the rate-limiting step must be downstream of the light absorption step so the impact of light intensity on k obs can be evaluated. 2.Just as the original plateau expression assumes reversible, Nernstian behavior, these expressions also assume that both states of the redox couple are stable and that electrode kinetics do not contribute to the observed rates. Treatment of quasi-reversible systems may be achieved via numerical simulation. Just as the original plateau expression assumes reversible, Nernstian behavior, these expressions also assume that both states of the redox couple are stable and that electrode kinetics do not contribute to the observed rates. Treatment of quasi-reversible systems may be achieved via numerical simulation. 3.Only one redox state of the catalyst is responsible for photon absorption and catalysis. Starting with the photocatalytically inactive form of the redox couple in solution allows for a simplified treatment of light absorption (incident photon intensity is essentially constant at the electrode surface) and ensures that all catalysis takes place near the electrode surface. Additionally, the inactive state of the catalyst cannot strongly absorb at the same excitation wavelength which would prevent photons from reaching the photocatalyst localized at the electrode surface. Only one redox state of the catalyst is responsible for photon absorption and catalysis. Starting with the photocatalytically inactive form of the redox couple in solution allows for a simplified treatment of light absorption (incident photon intensity is essentially constant at the electrode surface) and ensures that all catalysis takes place near the electrode surface. Additionally, the inactive state of the catalyst cannot strongly absorb at the same excitation wavelength which would prevent photons from reaching the photocatalyst localized at the electrode surface. 4.Catalysis must be sufficiently faster than photothermally induced convection. As explored extensively in section , convection can contribute to the overall current and thus must be accounted for if catalysis is slow. We note that this point necessitates some control conditions in which catalysis does not occur or is minimized under illumination so the photothermal contributions may be quantified. Catalysis must be sufficiently faster than photothermally induced convection. As explored extensively in section , convection can contribute to the overall current and thus must be accounted for if catalysis is slow. We note that this point necessitates some control conditions in which catalysis does not occur or is minimized under illumination so the photothermal contributions may be quantified. 5.The photocatalytic reaction must be sufficiently fast at scan rates above 20 mV/s so that the limiting current of the CV is scan-rate independent over at least an order of magnitude. The original formulation of the plateau expression necessitates that the limiting current be under kinetic control (i.e., the current is limited by reaction kinetics) and the method here relies on the same assumption. In the zone diagram formulation this is limited to zone KS and zone KD. By retaining sufficiently fast scan rates (25 mV/s and above), the catalytic current increases can mostly be attributed to catalysis rather than convection. The photocatalytic reaction must be sufficiently fast at scan rates above 20 mV/s so that the limiting current of the CV is scan-rate independent over at least an order of magnitude. The original formulation of the plateau expression necessitates that the limiting current be under kinetic control (i.e., the current is limited by reaction kinetics) and the method here relies on the same assumption. In the zone diagram formulation this is limited to zone KS and zone KD. By retaining sufficiently fast scan rates (25 mV/s and above), the catalytic current increases can mostly be attributed to catalysis rather than convection. 6.The catalytic mechanism must be known to sufficient degree to simplify the mechanism to a rate-limiting EC’ expression. For example, here we demonstrate that the kinetics of chloride coordination do not limit catalysis and the rate-limiting consumption of the chlorine radical can be written as a simplified first order rate-constant. Thus, k obs is sufficient to describe the rate-limiting step. The catalytic mechanism must be known to sufficient degree to simplify the mechanism to a rate-limiting EC’ expression. For example, here we demonstrate that the kinetics of chloride coordination do not limit catalysis and the rate-limiting consumption of the chlorine radical can be written as a simplified first order rate-constant. Thus, k obs is sufficient to describe the rate-limiting step. 7.To interpret the kinetics and reactivity of the system determined with the quantum yield CV measurement, the overall reaction mechanisms of the system should be sufficiently understood and products identified. For the Fe­(II/III)­Cl x system, we extrapolated the products and reactivities under electrochemical conditions using previously published reaction mechanisms and products that have been studied by other groups. , For novel systems or to determine the exact reaction products under electrochemical conditions, controlled potential electrolysis under illumination should be used. To interpret the kinetics and reactivity of the system determined with the quantum yield CV measurement, the overall reaction mechanisms of the system should be sufficiently understood and products identified. For the Fe­(II/III)­Cl x system, we extrapolated the products and reactivities under electrochemical conditions using previously published reaction mechanisms and products that have been studied by other groups. , For novel systems or to determine the exact reaction products under electrochemical conditions, controlled potential electrolysis under illumination should be used. By providing the criteria necessary to utilize this equation, readers can know when it is appropriate to use this analysis and avoid discrepancies and incorrect quantitative values of catalyst performance. Indeed, a major advantage of this methodology is that it may be carried out with equipment readily available to synthetic organic and photoredox researchers; we anticipate that adoption of these e-PC CV measurements with these guidelines will enable community wide photocatalyst benchmarking. Conclusion In summary, CV was used to directly measure the turnover quantum yield for an iron chloride LMCT photocatalyst. By utilizing a simple photocatalytic system with a light-limited homolysis step, we observed ideal sigmoidal catalytic CV behavior, thus enabling the observed rate to be related to the quantum yield through a light-dependent plateau current expression. Variations in light irradiance and scan rate yielded an observed rate with a linear dependence on photoexcitation rate giving a quantum yield of 0.39 ± 0.07. Finite element simulations further supported the photocatalytic rate measurements with fitted quantum yields that matched the experimental values. Photothermal convection was experimentally probed and incorporated into simulations to more accurately model photocatalytic CV behavior. The combination of experimental and theoretical work enables a CV-based quantum yield measurement and validated our simple experimental approach to directly study photocatalyst kinetics. Finally, the quantum yield measurement was used to probe photocatalytic C–H activation mechanisms which were consistent with an intermediate chlorine radical pathway. The CV-derived quantum yield measurement presented here lays the foundation for future studies of more complex photocatalysts. Overall, the measurement is simple, fast, and cost-effective, requiring no specialized training or equipment outside of a commercial potentiostat and LEDs. This measurement provides kinetic and mechanistic information about photocatalysts on an intermediate time scale thus providing a tool to probe specific parts of the photocatalytic reaction cycle. Critically, CV enables orthogonal mechanistic information in the absence of strong luminescent behavior, making it an attractive measurement for studying emergent first-row transition metal catalysts. We also demonstrated the practical application of this measurement for studying C–H bond activation, enabling the possibility of rapid substrate scope analyses. Future research will expand on the measurement demonstrated with the Fe­(II/III)­Cl x model system by investigating more complex and nonideal photocatalytic reaction pathways. We anticipate that a direct measurement of quantum yield through cyclic voltammetry will be impactful for future studies of photocatalysts across inorganic, organic, and biological disciplines. Supplementary Material