Combined Spectroelectrochemical and Simulated Insights into the

2 hours ago - Electrogenerated chemiluminescence (ECL) based sensors have the intrinsic advantage of having zero theoretical background signal, derive...
0 downloads 5 Views 703KB Size
Subscriber access provided by Kaohsiung Medical University

Article

Combined Spectroelectrochemical and Simulated Insights into the Electrogenerated Chemiluminescence Coreactant Mechanism Andrew Scott Danis, Karlie Paige Potts, Samuel Charles Perry, and Janine Mauzeroll Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 14 May 2018 Downloaded from http://pubs.acs.org on May 14, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Combined Spectroelectrochemical and Simulated Insights into the Electrogenerated Chemiluminescence Coreactant Mechanism Andrew S. Danis, Karlie P. Potts, Samuel C. Perry, Janine Mauzeroll* Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal H3A 0B8, QC, Canada. ABSTRACT: Electrogenerated chemiluminescence (ECL) based sensors have the intrinsic advantage of having zero theoretical background signal, derived from the electrochemical initiation of the luminescence process. Since the limit of detection (LOD) for sensors is defined as three times the noise of the background over the sensitivity of the system, further improvement to an ECL based detection limit is tied to improving sensitivity. Enhancing ECL sensitivity can be achieved through optimizing the mechanistic or kinetic performance of the reagents. While the mechanism for many luminophore-coreactant pairs have been established, the kinetics for the competing homogeneous reactions responsible for photon emission have not been directly resolved. This is due to the difficulty in experimentally probing and isolating a single homogeneous reaction while multiple simultaneous heterogeneous and homogeneous reactions are occurring. Combining the techniques of spectroelectrochemistry and finite element modelling, we monitor the homogeneous reactions for the coreactant pair, tris(2,2’-bipyridine)ruthenium(II) (Ru(BPY)32+) and tripropylamine (TPA). Corresponding trends found in the experimental absorbance and theoretical concentration profiles demonstrated that the reaction between Ru(BPY)33+ and TPA• intermediates proceeds significantly faster than the other available pathways. The identification of the oxidized intermediates as the dominant electron transfer pathway implies that the screening of luminophore and coreactant pairs that increase the stability of these kinetically labile intermediates would increase ECL sensitivity and ultimately performance.

INTRODUCTION Electrogenerated chemiluminescence (ECL) has evolved into a detection strategy,1–3 a platform for light emitting devices (LEDs),4–7 and an avenue for electron transfer (ET) research.8,9As a detection methodology, ECL is implicitly attractive as it is fundamentally limited by the detector’s noise.10 As a substitute material for semiconductors used in LED/organic light emitting device (OLED) fabrication, access to a tunable emission profile through luminophore design provides an entirely novel host of substrates for R&D.11 As a probe to study ET, the process of ECL is a physical manifestation of Marcus’s difficult to experimentally access inverted region.12 Despite ECL permeating research from fundamental to application, there are still fundamental gaps in our understanding of the rates at which competing homogeneous mechanisms generate photons in the most prominent ECL systems. The inherent complexity of ECL arises from the requirement that the electrochemically generated species undergo an exergonic ET to form an excited state molecule with relaxation kinetics that enables photon emission rather than vibrational relaxation.13 Early ECL research explored the annihilation pathway whereby the exergonic ET occurs between the reduced and oxidized form of the same luminophore.14,15 The ECL luminophore, tris(2,2-bipyridine)ruthenium(II) (Ru(BPY)32+) is the most

commonly employed for evaluating novel luminophore performance in the annihilation mechanism.16 

Ru(BPY)3+ + Ru(BPY)33+ ⇌ Ru(BPY)32+ + Ru(BPY)32+* (1) The annihilation pathway is ideal mechanistically, as it involves a single homogeneous reaction (eq. 1), but it is limited by experimental requirements such as a large solvent window to produce stabile forms of the radical cation and radical anion intermediates of the luminophore.17 In the coreactant mechanism one intermediate of the luminophore is substituted with another coreactant species.18 The coreactant is oxidized or reduced at the electrode and degrades to form a strong reducing or oxidizing intermediate, respectively. The coreactant intermediates then serve as the ECL compliment to the electrochemically activated luminophore species in solution.19 An example of a commonly employed oxidative-reductive coreactant is tripropylamine (TPA), which is oxidized at the electrode before undergoing a hydrogen abstraction in solution to become a strong reducing agent (TPA•), serving as a substitute for Ru(BPY)3+ in eq. 1.20. 

Ru(BPY)33+ + TPA• ⇌ Ru(BPY)32+* + TPA+

(2)

The coreactant pathway greatly complicates the overall ECL mechanism. Depending on the strength of the reducing or oxidizing agent, side products are formed that leads to additional pathways that also generate ECL. In the case of the Ru(BPY)32+ and TPA coreactant system,

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

coreactant driven homogeneous reactions (eqs 3 & 4) occur and contribute to the ECL emission in addition to eqs 1 & 2.21, 22 

Ru(BPY)32+ + TPA• ⇌ Ru(BPY)31+ + TPA+ +• 

Ru(BPY)31+ + TPA ⇌ Ru(BPY)32+* + TPA

(3) (4)

Scheme 1: Coreactant ECL mechanism. The intertwined heterogeneous and homogeneous reactions occurring simul2+ taneously in the Ru(BPY)3 / TPA system that generate ECL.

The development of higher performing luminophore and coreactant species requires an in depth knowledge of the nature and rates of all reactions involved in the ECL mechanism. While the broad mechanism of a number of luminophore-coreactant pairs are relatively well understood, thus far, the determination of homogeneous kinetics of the Ru(BPY)32+ and TPA coreactant system has not been achieved. This is largely due to modeling efforts requiring simultaneous treatment of all four (eqs 1-4) reaction pathways coupled to direct experimental evidence of species undergoing homogeneous reactions at the working electrode solution interface.22-24 Consequently, indirect electrochemical methodologies were developed to study kinetic rates for homogenous reactions that ensue electrochemical reactions. One such approach relied on the homogeneous reaction of ECL intermediates with a sacrificial agent forming a product that is both electroactive and generated in proximity to the electrode.25 In our work, hyphenation of electrochemistry and spectroscopy, spectroelectrochemistry (SEC), circumvents the need to use indirect sacrificial agents and provides a means to directly interrogate species formation or consumption in the vicinity of the electrode solution interface. Specifically, we monitor the concentration of Ru(BPY)32+ in solution and extract the rate of Ru(BPY)32+ consumption using a cuvette-based SEC setup. To extract mechanistic information we developed a model that encapsulated all proposed heterogeneous and homogeneous reactions within Scheme 1. In comparing experimental SEC and Finite El-

ement Modeling (FEM) based ECL simulations, we can identify what is the dominant homogeneous pathway during an ECL experiment. Elucidating the relationship between the homogenous reaction kinetics of coreactant based ECL system is a prerequisite to achieve a complete understanding of ECL and further the capability of ECL sensor detection.

EXPERIMENTAL SECTION Materials. All experiments were conducted using tris(2,2’-bipyridine)ruthenium(II) hexafluorophosphate (Ru(BPY)32+) (Sigma-Aldrich), tetrabutylammonium hexafluorophosphate (TBAPF6) (Sigma-Aldrich), tripropylamine (TPA) (Sigma-Aldrich), HPLC grade acetonitrile (MeCN)(Fisher), chloroplatinic acid (Sigma-Aldrich), ethanol (EtOH) (Commercial Alcohols), acetone (ACP), and distilled water. Equipment. The optical components leading to the cuvette consisted of an SLS 201L broadband source (Thorlabs), two Ø25 mm color balancing filters, LB-165 and LB-200 (Thorlabs), a fiber optic cable, M92L01 (Thorlabs), and a collimator. The collimator was composed of a Ø 12.7 mm plano convex lens, LA1074 (Thorlabs), in a housing assembly. The cuvette holder CVH100 (Thorlabs), contained two Ø 25 mm neutral density filters, ND02B and ND20B, with optical densities of 0.2 and 2.0 (Thorlabs), respectively. The cuvette used was a high precision quartz cuvette (Hellma Analytics). The light collection components consisted of a Ø 10 mm aspherical condenser lens, ACL108 (Thorlabs), inside a housing assembly, a fiber optic cable, M92L01 (Thorlabs) and a Czerny Turner model spectrometer, Maya LSL (Ocean Optics). The potentiostat used was a VSP-300 (Bio-Logic) in a three electrode setup. The working electrode was fabricated from a 70 x 9 x 2.2 mm piece of machine cut FTO glass (MSE Supplies LSE) that had platinum electrodeposited in a 40 x 9 mm area.26 The counter electrode was a platinum mesh (Goodfellow) and the reference electrode was a commercial Ag/Ag+ (10 mM Ag+ +e-→Ag) (CH Instruments). The cuvette cap holding the electrodes was a custom 3-D printed piece of Z-PETG and was printed using a Zortrax M200. Polymer shims were composed of Versaflex CL30 (GLS Corp). Ru(BPY)32+ Calibration Curve. Prior to the experiment, all glassware (less cuvette) was soaked overnight in an ethanolate bath, washed with distilled H2O, EtOH, acetone, and allowed to dry. A 1 mM stock solution of Ru(BPY)32+ in acetonitrile was diluted to provide sample solutions between 1 μM and 100 μM. The cuvette was rinsed with distilled water, EtOH, acetone, and sample solution and then filled with 3 mL of sample solution. An hour prior to the experiment, the emission source was turned on to allow it to stabilize. The spectrometer recorded data for 100 scans with an integration time of 450 ms at λmax for Ru(BPY)32+ (450 nm).27 The solution was then discarded and the cuvette was washed and rinsed as described previously in preparation for the next aliquot of sample solution at this concentration. All concentration measurements were performed in triplicate. The samples

ACS Paragon Plus Environment

Page 2 of 7

Page 3 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry were analyzed in order of least to most concentrated so as to minimize any possible contamination effects. Spectroelectrochemistry. A solution of 1 mM Ru(BPY)32+ in 0.1 M TBAPF6 was prepared in acetonitrile. The cuvette and electrodes were washed and rinsed as described above. The working, counter, and reference electrodes were positioned in the cuvette using a custom 3D printed cap, along with a 2.5 mL aliquot of analyte solution. An hour prior to the experiment, the emission source was turned on to give time for the lamp to stabilize. The cuvette was placed in the cuvette holder so as to have the electrodes parallel with the beam line. The potential was stepped between 0 V vs 1.05 V vs Ag/Ag+ at 5 min intervals for two cycles. The potentials of 0 V and 1.05 V vs Ag/Ag+ were chosen because 0 V vs Ag/Ag+ is sufficiently negative as to not oxidize Ru(BPY)32+ and 1.05 V vs Ag/Ag+ is sufficiently positive to cause mass transport limited oxidation of Ru(BPY)32+. The spectrometer acquired the spectrum for the duration of the waveform with an integration time of 450 ms. Absorption data for Ru(BPY)32+ was collected at 450 nm. Spectroelectrochemistry under ECL conditions. Solutions of Ru(BPY)32+ were prepared with varying ratios of TPA known to elicit ECL.28 The experimental setup was the same as described in the spectroelectrochemistry procedures, however, this time the potentiostat was set to apply a potential of 0 V vs Ag/Ag+ for 1 min followed by 1.05 V vs Ag/Ag+ for 6 min and 40 s. The potentials of 0 V and 1.05 V vs Ag/Ag+ were chosen because 0 V vs Ag/Ag+ is sufficiently negative as to not oxidize Ru(BPY)32+ and 1.05 V vs Ag/Ag+ is sufficiently positive to cause mass transport limited oxidation of Ru(BPY)32+.

RESULTS AND DISCUSSION SEC instrument and cell design. To track the concentration of Ru(BPY)32+ under ECL conditions, we developed a cuvette based SEC instrument with an integrated molecular absorption system requiring the addition of a light source, light focusing optics, light collecting optics, and filters (See Figures 1, S1). For ECL experiments the electrode orientation is perpendicular to the fiber optic aperture,29 in the SEC configuration the electrodes are parallel to the incident beam allowing light transmission and maximizing the available path length. After transiting the cuvette, the parallel rays are focused via an aspherical condenser lens [6] into a fiber optic [7] connected to a spectrometer [8]. Beam alignment close to the electrode solution interface required careful optimization.

Figure 1. Instrumentation used for SEC measurements. Starting from the point of origin for the light, the setup is composed of a Broadband 360-2600 nm source [1], color balancing filters (CBF) [2], fiber optic cable [3], plano convex lens [4], and neutral density filters (NDFs) [5] that generate, shape, guide, collimate, and decrease emitted light intensity, respectively, prior to light entering the cuvette. The SEC, inset rotated 90 degrees to show cross section, consists of a quartz cuvette, a light blocking mask, and uses a 3D printed cap to position a platinized FTO working electrode (WE), a + Pt mesh counter electrode (CE), and a Ag/Ag reference electrode (RE). The working and counter electrode positions are secured by polymer shims. The half reactions occurring behind the working electrode are benign to the interrogation source due to the incorporation of a light blocking mask on the exterior portion of the cuvette. After the light passes through the cuvette it is collected by an aspheric condenser lens [6] and directed to the spectrometer [8] via a fiber optic cable [7].

Due to the physical presence of the electrodes in the beam, less light impinges on the detector, leading to an overall decrease in signal intensity. Even a minor shift in electrode positioning and/or orientation within the beam can therefore cause detectable shifts in light intensity. While electrode movement is most likely negligible in a single fixed experiment, it becomes prevalent in our triplicate analysis, where the entire assembly is deconstructed, washed, rinsed, refilled, and reassembled. Our SEC cell mitigates these variations through two strategies. First, the positions of the electrodes were fixed at both their point of entry and point of termination in the cell via a custom cuvette cap and polymer shims respectively. Second, a light blocking mask was applied to the rear of the cuvette to negate any shift in absorbance caused by species formed at the counter electrode. SEC quantification of Ru(BPY)32+ concentration. The analytical performance of the cuvette based SEC setup to detect our target analyte, Ru(BPY)32+, was evaluated through an absorbance calibration curve (Figure 2) using the metal to ligand charge transfer (MLCT) absorption maximum at 450 nm.30 With the inclusion of 100 fold post measurement signal averaging, a limit of detection (LOD)

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 7

(3σ) of 26.3 nM was ascertained with a linear dynamic range (LDR) from 1 μM to 100 μM. While existing setups achieve a lower LOD and wider LDR for the detection of Ru(BPY)32+,31 the intent of designing this spectrometer based setup was for spectral resolution, which was germane to our application of SEC.

2+

2+

Figure 2. Calibration curve obtained for Ru(BPY)3 at   , to validate fundamental capability to conduct molecular absorption measurements. Spectra were recorded with 450 ms integration time and each trial was recorded for 100 scans allowing for 100 fold signal averaging. Each data point was collected in triplicate using a fresh aliquot of solution. Error bars represent standard error.

SEC quantification of the rate of Ru(BPY)32+ consumption by electrochemical oxidation. To validate that our cuvette based SEC setup could monitor in situ changes of Ru(BPY)32+ concentration, we assembled our SEC cell and collected the absorption signal during a series of potential steps above and below the standard redox potential of Ru(BPY)32+, 0.975 V vs Ag/Ag+ in acetonitrile (Figure 3). It has been shown that the electrochemical oxidation of Ru(BPY)32+ to Ru(BPY)33+ can be monitored via SEC as a decrease in absorbance32 or emission.33 The applied upper potential step was sufficiently positive to oxidize Ru(BPY)32+ at the mass transport limited rate. The applied lower potential was selected such that initially no Faradaic processes occurred, and following the initial oxidation step also ensured the reduction of Ru(BPY)33+ back to Ru(BPY)32+ at a mass transport limited rate. Confirmation of the capability to track in situ Ru(BPY)32+ concentration changes via SEC provided a foothold for evaluating the kinetics of (eq. 1-4) because the ECL luminophore Ru(BPY)32+ is a component in all them, either as a reactant or a precursor. In situ tracking of changes in Ru(BPY)32+ concentration (Figure 4) is evidenced by the subsequent fall and rise of absorbance correlating to the potential shifts above and below the standard redox potential of Ru(BPY)32+, respectively.

Figure 3. Cyclic voltammograms of 1 mM Ru(BPY)3 and 0.1 M TBAPF6 in acetonitrile performed in SEC cuvette with + platinized FTO WE, Pt mesh CE, and Ag/Ag RE. Scan rate was 0.1 V/s and potential was ramped from 0 V to 1.3 V to 0 V + (vs Ag/Ag ) for five cycles. Only the last three cycles are shown where non-faradaic current has been minimized and a reproducible reaction environment at the electrode surface has been achieved. Inset is scan rate analysis; linearity displays electrochemical reversibility of system under semiinfinite conditions.

Figure 4. Spectroelectrochemical control and detection of 2+ dynamic Ru(BPY)3 concentrations in solution. Initial con2+ centration in the SEC cell are 1 mM Ru(BPY)3 and 0.1 M TBAPF6 in acetonitrile. The applied potential waveform over time (blue trace), consists of a series of potential steps oscil2+ lating from oxidizing conditions for Ru(BPY)3 (1.05 V vs + 2+ + Ag/Ag ) to those regenerating Ru(BPY)3 (0 V vs Ag/Ag ). The absorbance intensity (black trace), at 450 nm, for 2+ Ru(BPY)3 changing over time overlaid with potential waveform. Spectrum was recorded with 450 ms integration time.

ACS Paragon Plus Environment

Page 5 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry SEC considerations and insight for the evaluation of Ru(BPY)32+ consumption during ECL. Our ECL cell operates under semi-infinite conditions to generate ECL analogous to predominant current sensor designs. We used concentration regimes of Ru(BPY)32+ and TPA that have been shown to illicit ECL21,28 to interrogate the weighed contributions of the homogenous reactions of the coreactant mechanism (k1-k4) in the reaction layer. Experimentally this was carried out through monitoring the global rate of Ru(BPY)32+ absorbance change during the electrochemical oxidation of TPA and/or Ru(BPY)32+ (Figure 5B and 5C). Conventional coreactant ECL experiments typically employ short time scale potential pulses (10-1 - 101 s) to maximize the concentration gradients and subsequent reagent mixing in order to generate high intensity ECL.13,34 Herein, extended time scale (hundreds of seconds) are used to maximize the initial absorbance signal, provide adequate sensitivity at 450 nm, and mitigate the impact of transient signals present at the selected integration time. Applying a constant potential for longer durations, forced the system into a mass transport limited environment, whereby the competing homogeneous reactions occurring in solution would be pinned at their relative maximums. Using SEC’s combined capability to track Ru(BPY)32+ concentrations in solution and to isolate which species is oxidized at the electrode via selective potential application, we were able to focus on individual components of the mechanism and examine their effect on Ru(BPY)32+ at different concentrations of TPA. In Figures 5B and 5C, more negative values of dAbs/dt correspond to a greater rate of Ru(BPY)32+ consumption in the cell. Figure 5B (red trace) shows an environment where the applied potential only oxidizes TPA (0.75 V vs Ag/Ag+) and in theory is only engaging the reactions described in eqs 3 & 4. As the ratio of TPA:Ru(BPY)32+ is increased, an acceleration in absorbance decay is observed, which indicates that the generation of Ru(BPY)3+ (eq. 3) outpaces the regeneration of Ru(BPY)32+ (eq. 4). Our FEM model depicts this as a consequence of the relatively short lifetime of the TPA+• radical (approx. 10-4 s) before deprotonation into TPA•, limiting the effective concentration of TPA+•.21 When a potential is applied that oxidizes both Ru(BPY)32+ and TPA (1.05 V vs Ag/Ag+) (Figure 5C black trace), all pathways are enabled (eqs 1-4). At low ratios of TPA to Ru(BPY)32+, the direct electrochemical oxidation of Ru(BPY)32+ is the major contributor to the rapid absorbance decay. However, upon increasing the ratio of TPA to Ru(BPY)32+, the absorbance decay rapidly decelerates, resulting in the plateau seen at high concentrations of TPA.

Figure 5. Applied electrochemical potentials and subsequent 2+ rate of 1 mM Ru(BPY)3 absorbance change with the addition of different ratios of TPA. An acetonitrile based solution of 1 2+ mM Ru(BPY)3 and 0.1 M TBAPF6 with varying ratios of TPA added to the SEC cell. An oxidizing potential of 0.75 V (•) or  1.05 V (•), was applied for 400 s and the   was  recorded. Integration time for spectrometer was 450 ms. A) Demonstration of capability to selectively electrochemically 2+ oxidize only TPA or both TPA and Ru(BPY)3 ; Cyclic Volt2+ ammetry was performed at a 10:1 ratio of TPA: Ru(BPY)3 . + Red line represents application of 0.75 V vs Ag/Ag where only TPA is oxidized, and black line represents application of + 2+ 1.05 V vs Ag/Ag where both Ru(BPY)3 and TPA are oxidized concurrently. B) Observed rate of change in the ab+ sorbance (red) of the system at 0.75 V vs Ag/Ag using different ratios of TPA and FEM simulation performed at the same potential (blue). C) Observed rate of absorbance change + (black) for the system at 1.05 V vs Ag/Ag using different

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ratios of TPA with FEM simulation performed at same potential (blue).

Elucidation of homogeneous rate constants from simulated ECL data. The cumulative rate of change in the concentration of Ru(BPY)32+ during ECL is complicated to predict, as Ru(BPY)32+ is being oxidized at the electrode surface, meanwhile in the bulk solution it is simultaneously being consumed according to eq. 3, and formed according to eqs 1, 2 and 4. The total concentration of Ru(BPY)32+ in the cell volume at any given time is therefore dependent on the relative magnitudes of the rates of consumption or formation according to eqs 1-4. This provides us with a window into this complex network of reactions, as we are able to compare simulated concentrations of Ru(BPY)32+ with the measured absorbance from Figure 5. The simulated rate of Ru(BPY)32+ consumption (d[Ru(BPY)32+]/dt) is dependent on the concentration of TPA in solution, and the relative homogeneous rate constants for eqs 1-4. If eqs 1-4 have the correct homogeneous rate constants, a simulated plot of d[Ru(BPY)32+]/dt vs [TPA] should show the exact same trend an experimental plot of dAbs/dt. The FEM model was initially constructed using established ECL simulation literature to bound the Ru(BPY)32+/ TPA mechanism.22 Diffusion coefficients and heterogeneous rate constants for Ru(BPY)32+ and TPA oxidation were calculated in house using an established methodology.35 Initially, the model assumed that all homogeneous reactions (eqs 1-4) proceeded at an equal rate. This resulted in d[Ru(BPY)32+]/dt becoming more negative with increased TPA concentration. This was the opposite of what was observed experimentally, indicating that in the experimental system, one of these reactions was kinetically favored over the others and was driving the regeneration of Ru(BPY)32+. To investigate this, individual homogeneous rate constants were parametrically swept in a model of our electrochemical cell29 until simulated changes in Ru(BPY)32+ concentration agreed with the experimental absorbance for Ru(BPY)32+ in solution. By increasing individual reaction rates, we found that only increasing the rate of reaction between TPA• and Ru(BPY)33+ (eq. 2) caused d[Ru(BPY)32+]/dt to become less negative with increasing TPA concentration (Figure S11). Further refinement showed that the slope of the curve and the concentration of TPA at which the plateau is reached could be manipulated by varying the magnitude of the rate increase (Figure S12). Parametric analysis revealed that a high-quality fit between simulated and experimental data came when the rate of TPA• reacting with Ru(BPY)33+ (eq. 2) was 100 times faster than the competing homogenous reactions. The shape of the matching simulated calibration curve is so strongly dominated by (eq. 2) that the impact of (eqs 1,3,4) becomes negligible.

CONCLUSIONS Through combining SEC and ECL based FEM, we were able to show that in the prominent mechanism for ECL research using the coreactant pair, Ru(BPY)32+ /TPA, there is a significantly dominant homogeneous reaction, eq. 2.

Using SEC and the judicious application of potentials, we were able to isolate parts of the ECL mechanism and monitor how Ru(BPY)32+ concentration varied in solution from a 0:1 to a 20:1 ratio of coreactant to luminophore. Utilizing our ECL based FEM governed by an established mechanism, we reproduced the experimental trend and extracted the homogeneous kinetics. This resulted in the finding that the recombination of Ru(BPY)33+ and TPA• outpaces the other three reactions with analogous molecules by two orders of magnitude. This opens up a host of questions for future research about the thermodynamics and kinetics behind this specific ET and the causality for the vast discrepancy. This manuscript also represents a significant step forward in the complete understanding of the coreactant ECL mechanism.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Contained within SI is: spectral radiance calculations for optics, electrochemistry, molecular absorption, spectroelectrochemistry, finite element modeling. (PDF)

AUTHOR INFORMATION Corresponding Author * Email: [email protected] Tel: 514-398-3898

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

ACKNOWLEDGMENT We would like to acknowledge Dr Sylvain Canesi and Dr. Hanadi Sleiman for continued ECL support. Dr Rustam Khaliullin and Dr Yifei Shi for helpful conversations.

REFERENCES (1) Dennany, L.; Forster, R. J.; White, B.; Smyth, M.; Rusling, J. F. J. Am. Chem. Soc. 2004, 126, 8835–8841. (2) Valenti, G.; Scarabino, S.; Goudeau, B.; Lesch, A.; Jović, M.; Villani, E.; Sentic, M.; Rapino, S.; Arbault, S.; Paolucci, F.; Sojic, N. J. Am. Chem. Soc. 2017, 139, 16830–16837. (3) Crespo, G. A.; Mistlberger, G.; Bakker, E. J. Am. Chem. Soc. 2012, 134, 205–207. (4) Kong, S. H.; Lee, J. I.; Kim, S.; Kang, M. S. ACS Photonics 2017, Article ASAP, DOI: 10.1021/acsphotonics.7b00864 (5) Chang, S.-C.; Li, Y.; Yang. J. Phys. Chem. B 2000, 104, 11650–11655. (6) Qi, H.; Zhang, C.; Huang, Z.; Wang, L.; Wang, W.; Bard, A. J. J. Am. Chem. Soc. 2016, 138, 1947-1954 (7) K. N. Swanick; S. Ladouceur; E. Zysman-Colman; Z. Ding, RSC Adv. 2013, 3, 19961. (8) Ketter, J. B.; Wightman, R. M. J. Am. Chem. Soc. 2004, 126, 10183–10189. (9) Qi, H.; Chen, Y.-H.; Cheng, C.-H.; Bard, A. J. J. Am. Chem. Soc. 2013, 135, 9041–9049. (10) Benoit, L.; Choi, J.-P. ChemElectroChem 2017, 4, 1573– 1586.

ACS Paragon Plus Environment

Page 6 of 7

Page 7 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry (11) Mitschke, U.; Bauerle, P. J. Mater. Chem. 2000, 10, 1471–1507. (12) Kapturkiewicz, A. ChemElectroChem 2017, 4, 1604– 1638. (13) Bard, A. J. Electrogenerated Chemiluminescence; Marcel Dekker, Inc.: New York, 2004. (14) Hercules, D. M. Science 1964, 145, 808–809. (15) M. Hesari; Z. Ding, J. Electrochem. Soc. 2016, 163, H3116–H3131. (16) Zhou, M.; Robertson, G. P.; Roovers, J. Inorg. Chem. 2005, 44, 8317–8325. (17) Pyati, R.; Richter, M. M. Annu. Rep. Prog. Chem. Sect. C: Phys. Chem. 2007, 103, 12–78. (18) Miao, W. Chem. Rev. 2008, 108, 2506–2553. (19) Kirschbaum-Harriman, S.; Duerkop, A.; Baeumner, A. J. Analyst 2017, 142, 2648–2653. (20) Zu, Y.; Bard, A. J. Anal. Chem. 2000, 72, 3223–3232. (21) Miao, W.; Choi, J.-P.; Bard, A. J. J. Am. Chem. Soc. 2002, 124, 14478–14485. (22) Svir, I.; Oleinick, A.; Klymenko, O. V.; Amatore, C. ChemElectroChem 2015, 2, 811–818. (23) Kaim, W.; Fiedler, J. Chem. Soc. Rev. 2009, 38, 3373– 3382. (24) Jin, J.; Takahashi, F.; Kaneko, T.; Nakamura, T. Electrochim. Acta 2010, 55, 5532-5537.

(25) Fan, F.; Cliffel, D.; Bard, A. J. Anal. Chem, 1998, 70, 2941-2948. (26) Tsekouras, G.; Mozer, A. J.; Wallace, G. G. J. Electrochem. Soc. 2008, 155, K124–K128. (27) Witt, M. D.; Roughton, S.; Isakson, T. J.; Richter, M. M. J. Lumin. 2016, 171, 118–123. (28) Liu, X.; Shi, L.; Niu, W.; Li, H.; Xu, G. Angew. Chemie Int. Ed. 2007, 46, 421–424. (29) Danis, A. S.; Odette, W. L.; Perry, S. C.; Canesi, S.; Sleiman, H. F.; Mauzeroll, J. ChemElectroChem 2017, 4, 1736– 1743. (30) Colasson, B.; Credi, A.; Ragazzon, G. Coord. Chem. Rev. 2016, 325, 125–134. (31) Chen, Y.; Zhou, S.; Li, L.; Zhu, J. Nano Today 2017, 12, 98–115. (32) Wilson, R. A.; Pinyayev, T. S.; Membreno, N.; Heineman, W. R. Electroanalysis 2010, 22 (19), 2162–2166. (33) Branch, S. D.; Lines, A. M.; Lynch, J.; Bello, J. M.; Heineman, W. R.; Bryan, S. A. Anal. Chem. 2017, 89, 7324-7332 (34) Zhang, H.; Zuo, F.; Tan, X.; Xu, S.; Yuan, R.; Chen, S. Biosens. Bioelectron. 2018, 104, 65-71. (35) Perry, S. C.; Al Shandoudi, L. M.; Denuault, G.; Anal. Chem. 2014, 86, 9917-9923.

For TOC only

ACS Paragon Plus Environment