Redox reaction characteristics of riboflavin: A fluorescence

0 downloads 0 Views 822KB Size Report
spectroelectrochemical analysis and density functional theory calculation .... In situ elec- trochemical fluorescence was recorded on a fluorescence spectrometer.
Bioelectrochemistry 98 (2014) 103–108

Contents lists available at ScienceDirect

Bioelectrochemistry journal homepage: www.elsevier.com/locate/bioelechem

Redox reaction characteristics of riboflavin: A fluorescence spectroelectrochemical analysis and density functional theory calculation Wei Chen 1, Jie-Jie Chen 1, Rui Lu, Chen Qian, Wen-Wei Li, Han-Qing Yu ⁎ Department of Chemistry, University of Science and Technology of China, Hefei 230026, China

a r t i c l e

i n f o

Article history: Received 22 January 2014 Received in revised form 13 March 2014 Accepted 21 March 2014 Available online 29 March 2014 Keywords: Spectroelectrochemistry Riboflavin Proton coupled electron transfer Redox chemistry Disproportionation process

a b s t r a c t Riboflavin (RF), the primary redox active component of flavin, is involved in many redox processes in biogeochemical systems. Despite of its wide distribution and important roles in environmental remediation, its redox behaviors and reaction mechanisms in hydrophobic sites remain unclear yet. In this study, spectroelectrochemical analysis and density functional theory (DFT) calculation were integrated to explore the redox behaviors of RF in dimethyl sulfoxide (DMSO), which was used to create a hydrophobic environment. Specifically, cyclic voltafluorometry (CVF) and derivative cyclic voltafluorometry (DCVF) were employed to track the RF concentration changing profiles. It was found that the reduction contained a series of proton-coupled electron transfers dependent of potential driving force. In addition to the electron transfer-chemical reaction-electron transfer process, a disproportionation (DISP1) process was also identified to be involved in the reduction. The redox potential and free energy of each step obtained from the DFT calculations further confirmed the mechanisms proposed based on the experimental results. The combination of experimental and theoretical approaches yields a deep insight into the characteristics of RF in environmental remediation and better understanding about the proton-coupled electron transfer mechanisms. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Electron transfer reactions are ubiquitous in environmental remediation of pollutants, geochemical cycles of Fe and Mn, and microbial fuel cells that produce electricity from wastes [1–3]. Flavins are redox active species widely distributed both in nature and in vivo, and play a key role in mediating electron transfer in biogeochemical systems [3–5]. Also, flavin derivatives can act as photosensitizers to produce reactive oxygen species [6] and molecular catalysts for water-oxidation reaction [7], and can be chelated with metals easily [8,9]. Riboflavin (RF) is the primary redox active component of flavin and possesses three accessible oxidation states similar to quinones, each with protonation and deprotonation forms (Scheme 1) [10,11]. The elucidation of electron transfer reactions and redox properties of RF is critical to understand its biochemical function. It is well known that under hydrophilic conditions the redox behavior of RF is pH-dependent and RF is directly reduced to hydroquinonoid RF (RFredH2) in one step via a two-electron/two-proton transfer pathway [10–14]. For the flavoenzymes that function as hydrophobic sites [15], however, an aprotic organic environment is suitable for the ⁎ Corresponding author. Fax: +86 551 63601592. E-mail address: [email protected] (H.-Q. Yu). 1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.bioelechem.2014.03.010 1567-5394/© 2014 Elsevier B.V. All rights reserved.

investigation into the electron transfer process. Cyclic voltammograms (CV) of RF in aprotic solvents have shown more complex redox behavior than those of simple quinonoid structures [16,17]. However, there are still controversies on the two deduced mechanisms involved in the reduction of RF in aprotic media, the electron transfer-chemical reaction-electron transfer (ECE) and the disproportionation (DISP1) mechanisms. In addition, the influence of potential driving force on the redox progress remains unclear yet. Thus, efforts should be made to elucidate the redox mechanisms involved and better understand its environmental behaviors. Spectroelectrochemical methods have been widely used to explore the redox properties of compounds analogous to quinonoid structure [18–20]. A combination of electrochemical regulation and spectroscopic monitoring helps to correlate the electron transfer processes in electrode redox reactions to spectroscopic observations. Some spectroelectro-chemistry-based analytical methods have been developed. Among them, cyclic voltabsorptometry (CVA) and derivative cyclic voltabsorptometry (DCVA) are useful tools to track the changes of reactants, intermediates and products in the electrochemical process [21,22]. With high sensitivity and salient structural information fluorescence can provide, spectrofluoroelectrochemistry offers an efficient and sensitive approach to investigate redox properties of species [23,24]. On the other hand, the density functional theory (DFT) calculations can also provide molecular level insights into the change of RF

104

W. Chen et al. / Bioelectrochemistry 98 (2014) 103–108

e

RFox

RFrad2

e

H+

H+

RFox

H+

e

RFoxH+

e

RFrad

e

H+

H+

H+

H+

e e

e

H+

RFradH H+

H+

e e

H+

RFradH2

RFred2

RFredH H+

e e

H+

H+

RFredH2

Scheme 1. Possible species of riboflavin in the steps of sequential electron and proton addition reactions.

reduction in aprotic organic environment to further interpret the experimental results. However, so far information regarding this respect is limited. With strong fluorescence and variable intensity under distinctive redox states, RF is an appropriate compound for fluorescence spectroelectrochemical analysis. In this contribution, we report our UV–Vis and fluorescence spectroelectrochemical studies of electron transfer of RF in DMSO. Cyclic voltafluorometry (CVF) and derivative cyclic voltafluorometry (DCVF), similar to CVA and DCVA respectively, were employed to elucidate the mechanisms of the redox reactions and to find out whether the reduction process was reversible. With a combination of spectroelectrochemical analysis with the DFT calculation, the compounds formed in the reduction and their correlations with the peaks in CVs were identified, and thus the mechanisms of the electrochemical redox reaction of RF in aprotic media were elucidated. 2. Experiments and computations 2.1. Reagents n-Bu4NPF6 was purchased from J&K Tech. Ltd., China, while RF was purchased from Sigma Co., USA. DMSO (Sinopharm Chemical Reagent Co., China.) was used as solvent. All chemicals were reagent grade and were used without further purification. 2.2. Spectroelectrochemical instruments and analytical procedures CVs were recorded on a CHI760D electrochemical workstation (CHI Instruments Co., China). Working electrodes were a 1-mm diameter planar platinum disk for electrochemical study and a piece of 6 mm × 7 mm Pt mesh for spectroelectrochemical experiments. In addition, a Pt wire counter electrode and an Ag/Ag+ reference electrode (a silver wire as quasi reference electrode for spectroelectrochemical experiments) were used. Before each experiment, the Pt disk electrode was polished (alumina polishing paste), washed with water, sonicated and activated by H2SO4. The Pt mesh electrode was cleaned by immersing it in 0.1 M HNO3 for 3 min and rinsed extensively, followed by H2SO4 activation. Accurate potentials were obtained using ferrocene as an internal standard. All solutions were deoxygenated by

nitrogen purging before the tests and all experiments were conducted under ambient temperature (22 ± 2 °C). A thin-layer quartz glass spectroelectrochemical cell (CHI Instruments Co., China) was used for complete electrolysis in a short time and eliminating the influence of diffusion at very low scan rates (v). 300 μL of the solution was added into the thin-layer cell for all spectroelectrochemical experiments. The configuration of the three-electrode arrangement in the thin-layer cell for spectroelectrochemical measurements is shown in Scheme S1. The UV–Vis spectra obtained at different applied potentials were recorded on a UV-2450 spectrometer (Shimadzu Co., Japan). In situ electrochemical fluorescence was recorded on a fluorescence spectrometer (LS 55, Perkin-Elmer Co., USA) fitted with a pulsed Xenon excitation source. Unless otherwise stated, the excitation and emission slits were both 5 nm, and the scanning speed was set at 600 nm/min for all measurements. 2.3. DFT calculations The structures of RF molecules in DMSO solution are studied by DFT computation. In the calculation, an all-electron method within the Perdew, Burke, and Ernzerhof (PBE) forms of generalized gradient approximation (GGA) [25,26] for the exchange-correlation term was employed, as implemented in the DMol3 code [27,28]. The double precision numerical basis sets including p polarization (DNP) were adopted. The energy in each geometry optimization cycle was converged to within 1 × 10− 5 Hartree with a maximum displacement and force of 5 × 10−3 Å and 2 × 10−3 Hartree/Å, respectively. The solvent effect of the DMSO medium was described using the conductorlike screening model (COSMO) [29,30]. COSMO is a continuum solvent model where the solute molecule forms a cavity within the dielectric continuum of permittivity that represents the solvent. 3. Results and discussion 3.1. Electrochemical results compared with previous experiments CVs of RF reduction in DMSO in different scan ranges are shown in Fig. 1a. Two oxidation peaks (Peaks 2 and 3) were observed upon reversal of the scan direction after the first reduction process (Peak 1). At more negative potential, other two reduction peaks (Peaks 4 and 5) appeared. Peak 2 decreased significantly when the forward scan was extended to pass Peak 5, compared with that when it just passed Peak 1. Meanwhile, the size of Peak 3 was enlarged when the forward scan direction was reversed after Peak 5. The reduction process was diffusion-controlled as peak potential of Peak 1 was independent of the scan rate and the peak current increased linearly with the square root of scan rate (Fig. S1). Fig. 1b shows the CVs of 0.2 mM RF at different scan rates in a scan range from −0.4 to −1.4 V vs. Ag/Ag+. To clearly compare the peak change, the current data are scaled through multiplying them by v−0.5. With an increase in v, Peak 3 appeared and grew larger and then shrunk, while Peak 2 shifted conversely in magnitude. The changes in the relative magnitude of Peaks 2 and 3 at v ≥ 100 mV/s were similar to the results reported by Tan et al. [17], who failed to report the results at v lower than 100 mV/s. The reduction of RFox produced the radical anion RF•− rad (Eq. 1, Peak 1), followed by homogeneous protonation to form RFradH• (Eq. 2). The protonated radical RFradH• was further reduced to produce RFredH− and oxidized (Eq. 3, Peak 3) back when the CV scan was reversed. − The protonation step (Eq. 2) governs the ratio of RF•− rad and RFredH , as well as the oxidation peak intensities of Peaks 2 and 3. At a higher scan rate (5000 mV/s), there was no sufficient time for RF•− rad to be chemically protonated. Thus, the protonation of RF•− rad was suppressed and the reoxidation of RF•− rad (Peak 2) dominated when scanning was back. As v was decreased from 5000 to 100 mV/s, the protonation reaction began to dominate and more RF•− rad radicals were turned into RFradH•, which could be reduced into RFredH− immediately. When v

W. Chen et al. / Bioelectrochemistry 98 (2014) 103–108

P3

P2

a

1

Current (µA)

Current (µA)

-1 0 -1

-2

105

20 ms 50 ms 100 ms 1000 ms

-2

P4 P5

-3

-3 P1

-2.0

-2.5

-1.5

-1.0

-0.5

0.0

-2.0

Potential (V, vs. Ag/Ag+)

Current (µA)

b

CVs (Fig. 1c) of RF at different scan rates were recorded in a wider scan range, where other two reduction peaks appeared. And the differential pulse voltammograms (DPVs, Fig. 2) under different pulse periods were also investigated. The intensity of Peak 4 was observed to increase with the decreasing scan rate (increasing pulse period), while that of Peak 5 decreased. Besides, Peak 1 also decreased with the increasing pulse period (decreasing scan rate). As mentioned above, the protonation reaction (Eq. 2) began to dominate with the decreasing scan rate, resulting in the increasing concentration of prod• ucts (RF− ox and RFradH ) and the decreasing concentration of reactants ). Since the peak intensity is proportional to the concen(RFox and RF•− rad tration of species responsible for the reduction, Peak 4 could be attributed to the reduction of RF− ox (Eq. 4) and Peak 5 was associated with a 2− − further reduction of RF•− rad to RFred (Eq. 5). The consumption of RFox by Peak 4 (Eq. 4) moved the reaction equilibrium of Eq. 2 forward, and thus more RFredH− was formed by the rapid reduction of RFradH•. This could be used to explain why the current ratio of Peak 2 to Peak 3 decreased in a wider scan range (Fig. 1a).

5 mV/s 10 mV/s 100 mV/s 500 mV/s 5000 mV/s

-2

-1.2

-1.0

-0.8

-0.6

-0.4

Potential (V, vs. Ag/Ag+)

Current (µA)

3

c

0

20 mV/s 100 mV/s 500 mV/s 5000 mV/s

-3

3.2. Spectroelectrochemical insights into RF reduction

-6

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

Potential (V, vs. Ag/Ag+) Fig. 1. (a) CVs of 0.2 mM RF at a Pt disk electrode in DMSO with 0.05 M n-Bu4NPF6 at scan rate 0.5 V/s. Scan range: −0.4 to −1.4 V vs. Ag/Ag+ (Dashed line); 0.0 to −2.3 V vs. Ag/ Ag+ (Solid line). CVs of RF recorded at different scan rates. Scan range: −0.4 to −1.4 V vs. Ag/Ag+ (b); 0 to −2.3 V vs. Ag/Ag+ (c). The current data are scaled through multiplying them by v−0.5.

became very slow (≤10 mV/s), the homogeneous proton transfer reaction was considered as a reversible process and no longer a rate-limiting step. Thus, the electrochemical redox reactions remained dominant in the entire CV scanning with such a low v. –

RFox þ e ⇌RFrad

–

ð1Þ 

þ RFox ⇌RFrad H þRFox 



RFrad H þ e ⇌RFred H –

RFox þ e ⇌RFrad

RFrad

–

0.0

0

-1.4



-0.5

Fig. 2. DPVs obtained in the reduction of RF at different pulse periods in DMSO with 0.05 M n-Bu 4 NPF 6 . Scan range: 0 to − 2.0 V vs. Ag wire; amplitude: 0.05 V; pulse width: 5 ms; potential increment: 5 mV.

-4

–

-1.0

Potential (V, vs. Ag wire)

2

RFrad

-1.5





2–

2–

þ e ⇌RFred :



ð2Þ ð3Þ

ð4Þ ð5Þ

In situ UV–Vis spectra of RF at different applied potentials (Fig. 3a) were obtained to spectroscopically observe the species change during the reduction. Each potential was applied until the spectra did not change at all to ensure complete electrolysis. As the potential moved negatively, the absorption peaks at 446 nm, 344 nm and 271 nm were observed to decrease, along with the increase of absorption peaks at 330 nm and 262 nm. These changes corresponded to the transformation of oxidized RF to its reduced form. Absorption at 373 nm was observed to dramatically increase firstly and decrease after the applied potential negative than −1.0 V vs. Ag wire (Fig. 3b), corresponding to the characteristic absorption peak of the flavin semiquinone radical [16]. On the other hand, a shoulder absorption around 475 nm appeared due to a significant amount of RFredH− in solution with RF•− rad [17]. The intensity of these two absorption peaks, 373 nm and 475 nm, changed slightly but not obvious after the electrolysis, indicating that the radicals formed during the reduction were long-lived. The excitation–emission matrix of RFox in DMSO was measured (Fig. S2), and the fluorescence intensity at λem = 516 nm remained almost unchanged over time under illumination at λex = 440 nm, indicating that no photodegradation occurred. In the repetitive CV scans in a wider scan range (Fig. 4a), the CVF of RFox was recorded (Fig. 4b) and compared with the corresponding CV in Fig. 4a. The reduction of RFox was paralleled by the disappearance of fluorescence and the fluorescence returned to the original level after the cycle, indicating that the compounds produced in the reduction process could be oxidized back into RFox without forming any other decomposition products.

106

W. Chen et al. / Bioelectrochemistry 98 (2014) 103–108

a

0.6

Absorption (a.u.)

0.5

Initial -0.4 V -0.6 V -0.8 V -1.0 V -1.2 V -1.4 V -1.6 V -1.8 V Reoxidation

0.4 0.3 0.2 0.1 0.0 -0.1 200

250

300

350

400

450

500

550

600

650

Wavelength (nm)

b Absorption (a.u.)

0.14

0.12



ð6Þ

0.10

3.3. DFT calculation of RF reduction Absorption at 373 nm

0.0

-0.5

-1.0

-1.5

-2.0

Potential (V, vs. Ag wire) Fig. 3. (a) In situ UV–Vis spectra of 0.2 mM RF in DMSO containing 0.05 M n-Bu4NPF6 obtained at each potential step and a comparison of the oxidized form, reduced form and re-oxidized form (solid lines). (b) Plot of Absorption at 373 nm vs. controlled potential.

To obtain the derivative of fluorescence vs. time, the corresponding DCVF was calculated from the CVF data in Fig. 4b. The DCVF data in Fig. 4c agree well with the CV data in Fig. 4a. The fluorescence intensity was proportional to the compound concentration and the current was the first derivative of the charge flowing through the system [23]. Thus, the peaks of the DCVF coincided with the peaks of the CV corresponding to the reduction and oxidation of RFox. A comparison between

Potential (V, vs. Ag wire) 0 0

-2

0

-2

0

a

The charge distribution and geometry structure change of the tricyclic ring can significantly affect the electron-accepting process. The N3, N5, and O4 atoms are active sites of the RF structure in DMSO shown in Fig. 6a. Table S1 lists the typical geometry parameters and charge distribution of the active sites of intermediate species in RF reduction. The charge distribution results from charge equilibration (QEq) approach [31] indicate that the different redox state causes a charge difference in N3, N5, and O4 atoms. The change of the charge distribution for the active sites is consistent for the four main reduction steps (Eqs. 1, 3, 4, and 5). The N3, N5, and O4 atoms become less electrophilic. In addition, the bond lengths of C4a–N5, N3–C4 and C4–O4 increase in the reduction processes, and the variation of bond angle of C5a–N5–C4a is opposite to C4a–C4–N3, resulting in the changes of geometry symmetry in the isoalloxazine ring. Furthermore, the butterfly bend angles (θb) of the RF are considered to be a key feature that distinguishes flavin's reduced state from the oxidized one [32]. The θb vary significantly from RFradH• to RFredH− (Fig. 6b, d) in the reduction step of Eq. 3. This reduced and protonated anion (FlredH−, the front view in Fig. 6c) is very long-lived [17], and the bending of the tricyclic isoalloxazine ring plays an important role in the modulation of the flavin's reduction potentials [33] and the interaction of enzyme environment.

-20

800

-40 200

-1.0 V vs. Ag wire Fluorescence Intensity (a.u.)

Fluorescence (a.u.) Curerent (µA)



–

RFrad þ RFrad H ⇌RFred H þ RFox :

0.08

dF / dt

Fig. 4a and c indicates that the two CV peaks, i.e., Peaks 1 and 2 in Fig. 1a, should be attributed to the reduction of RFox and its reverse reaction at a low scan rate. In the scan range from 0 to −1.0 V vs. Ag wire, CV, CVF and DCVF of RF were also investigated (Fig. S3), showing an irreversible fluorescence decrease because of the consumption of RFox. There are two possible pathways in the reduction of RF, the ECE (Eqs. 1–3) or DISP1 (Eqs. 1, 2 and 6) mechanisms. To find out whether the latter process was involved in the regeneration of RFox or not, the fluorescence change of RFox was recorded in a potential step from a constant potential to the open circuit (Fig. 5). The fluorescence intensity decreased largely at −1.0 V vs. Ag wire, showing the consumption of RFox by the electrolysis. When the potential was stepped to the open circuit, a slight increase in fluorescence was observed, indicating the regeneration of RFox, which was not likely to be caused by the electrochemical oxidation. Also, due to the complete electrolysis and the solutions in the thin-layer cell all exposed to the optical path, the diffusion effect can be excluded. A homogeneous electron transfer (Eq. 6) could be used to explain the regeneration of RFox. Thus, the DISP1 process was confirmed to be responsible for the formation of RFox and RFredH−.

b

160 120 80 2 c 0

600

400

Open Circuit

200

-2 0

200

400

600

800

t (s)

0

400

800

1200

Time (s) Fig. 4. Multicycle thin-layer CV (a), CVF (b) and DCVF (c) of 0.2 mM RF at a Pt mesh electrode. Scan range: 0.0 to −2.0 V vs. Ag wire; scan rate: 10 mV/s; circling number: 2; excitation/emission wavelength: 440/516 nm. The curves were smoothed by the Savitzky–Golay method.

Fig. 5. Fluorescence intensity (excitation/emission wavelength: 440/516 nm) vs. time for the consumption and regeneration of RFox in a potential step from −0.1 V vs. Ag wire to open circuit. The excitation and emission slits were set as 10 nm.

W. Chen et al. / Bioelectrochemistry 98 (2014) 103–108

107

Fig. 6. Optimized structures of RFox and RFredH− in the front (a, c) and top (b, d) views as the representative structures with the butterfly bend angles (θb), the θb is defined as the angle between the planes of the pyrimidine and benzene six-membered rings.

The reduction potentials in DMSO for both one- and two-electron pathways of RF reduction were calculated by DFT methods with COSMO to determine the solvent effect. The standard electrode θ θ vs. EFc ) of the half-reaction is defined as potential (Ecal θ

θ

θ

Ecal ¼ –ΔG =nF–EFc

ð7Þ

where n is the number of electrons on the left side of the reaction, F is the Faraday constant, which equals 23.06 kcal/mol V, and EθFc is the standard reduction potential of ferrocene (Fc) electrode with a calculated value of 4.89 V provided in Table S2. The calculated values of the redox potentials of RF obtained from the DFT calculations were used to verify the experimentally measured results. All the calculated reduction potentials (Eθcal) were referenced to the reduction potential of ferrocene (EθFc) to allow a comparison among the data from the CV experiments (Eexp) and the simulated values reported in 17 (Esim). The results show that the Eθcal of RF for each step agrees well with the Eexp and Esim shown in Table 1. Therefore, the calculation results further confirm the mechanisms of electrochemical reduction of RF in DMSO, and imply that the fluorescence spectroelectrochemistry can be effectively applied to explore the redox reactions of the derivatives with tricyclic heteronuclear organic ring. Table 1 Electrochemical values of RF reduction in DMSO with 0.05 M Bu4NPF6 associated with heterogeneous electron transfer steps and homogeneous proton transfer steps. Reaction −

RFox + e



RF•− rad

+ RF•− ⇌ RFradH• rad + H RFradH• + e− ⇌ RFredH− RFox − H+ ⇌ RF− ox − RF− ⇌ RF•2− ox + e rad − •− RFrad + e ⇌ RF2− red

Eθcal (V)

Eexp (V)[a] & CV peak

Esim (V) (Ref. [17])

−1.08

−1.254R, P1 −1.166O, P2

−1.22

−1.09

−0.957O, P3

−1.05

−1.52 −2.25

R

−1.814 , P4 −2.214R, P5

−1.62 −1.98

([a] Peak potential measured at v = 500 mV/s: E (vs. Fc/Fc + ) = − 0.183 V + E(vs. Ag/Ag+ ). R Reduction process. O Oxidation process).

The redox potential of the reactions is governed by the thermodynamic driving force, which substantially affects the electron transfer kinetics [34]. The free energy changes of RF in the reactions of electron and proton transfers (Scheme 1) are shown in Table S3. The results reveal that, in all cases, the free energy changes for the electron additions are substantial and make a dominant contribution to the overall free energy of the reduction. The free energy change caused by the first electron transfer to neutral RFox in DMSO is −87.966 kcal/mol, which is quite negative compared to that of the second electron transfer to anionic semiquinone (RF•− rad), − 60.871 kcal/mol. If a coupling of the individual electron and proton transfer reactions is taken into account, the first coupled 1e−/1H+ transfer reaction appears to be more favorable than that of the second 1e−/1H+ transfer reaction (Table S3). Furthermore, the free energy change (−60.871 kcal/mol) of RF•− rad + •− e− ⇌ RF2− red is more negative than the value (−0.470 kcal/mol) of RFrad + H+ ⇌ RFradH•, suggesting that the second electron transfer can occur more easily with a stronger thermodynamic driving force than the proton addi• − − tion for RF•− rad. The reaction of RFradH + e ⇌ RFredH with a free energy change of −87.680 kcal/mol will spontaneously occur and change the re+ • action equilibrium to favor the reaction of RF•− rad + H ⇌ RFradH . These results can explain why some of RF•− rad species can be transformed into RFradH•. Thus, the peaks present in the CVs are in agreement with the theoretical calculations. A combination of spectroelectrochemical analysis and theoretical calculation in our work can be used as an efficient method to determine the radicals in the electron transfer processes. With the results of the spectroelectrochemical analysis and DFT calculations above, we propose the redox reaction scheme of RF in DMSO. RFox is initially reduced on the electrode via a reversible one-electron reduction pathway to RF•− rad (Eq. 1), which can either be further reduced to RF2− red through another one-electron electrochemical reduction reaction (Eq. 5), or be protonated to form RFradH• (Eq. 2, the proton is provided by RFox itself, RFox − H ⇌ RF− ox), depending on the scan range and scan rate selected. The protonated RFradH• is immediately reduced to RFredH− (Eq. 3). In addition, RFredH− and RFox can be regenerated via a DISP1 pathway (Eq. 6). The deprotonated RF− ox is also able to undergo one-electron reduction at more negative potentials (Eq. 4).

108

W. Chen et al. / Bioelectrochemistry 98 (2014) 103–108

RF is capable of undergoing a series of 1e− reduction associated with proton transfer, depending on the potential driving force. 4. Conclusions In the present work, a combination of electrochemical observation with spectroelectrochemical analysis as well as DFT calculation was successfully used to explore the mechanism of RF reduction. The effective integrated utilization of spectroscopy, electrochemistry and computational chemistry provides new possibilities to probe proton-coupled electron transfer mechanisms of redox active molecules, and the results provide direct evidence about the redox properties of flavins present within hydrophobic flavoproteins. A better understanding of the redox mechanisms of flavins would be useful to regulate and tune the electron transfers in environmental remediation. In addition, the integrated approach used in this work may also be utilized to elucidate the redox reaction mechanisms of other species in natural and engineered biosystems. Acknowledgments We thank the National Natural Science Foundation of China (51129803) and the Program for Changjiang Scholars and Innovative Research Team in University of the Ministry of Education of China for the support of this study. The numerical calculations were performed on the supercomputing system in the Supercomputing Center at the University of Science and Technology of China, China. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bioelechem.2014.03.010. References [1] P.P. Edwards, H.B. Gray, M.T.J. Lodge, R.J.P. Williams, Electron transfer and electronic conduction through an intervening medium, Angew. Chem. Int. Ed. 47 (2008) 6758. [2] M.E. Nielsen, D.M. Wu, P.R. Girguis, C.E. Reimers, Influence of substrate on electron transfer mechanisms in chambered benthic microbial fuel cells, Environ. Sci. Technol. 43 (2009) 8671. [3] H. von Canstein, J. Ogawa, S. Shimizu, J.R. Lloyd, Secretion of flavins by Shewanella species and their role in extracellular electron transfer, Appl. Environ. Microbiol. 74 (2008) 615. [4] E. Marsili, D.B. Baron, I.D. Shikhare, D. Coursolle, J.A. Gralnick, D.R. Bond, Shewanella secretes flavins that mediate extracellular electron transfer, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 3968. [5] R. Li, J.M. Tiedje, C. Chiu, R.M. Worden, Soluble electron shuttles can mediate energy taxis toward insoluble electron acceptors, Environ. Sci. Technol. 46 (2012) 2813. [6] C.K. Remucal, K. McNeill, Photosensitized amino acid degradation in the presence of riboflavin and its derivatives, Environ. Sci. Technol. 45 (2011) 5230. [7] E. Mirzakulova, R. Khatmullin, J. Walpita, T. Corrigan, N.M. Vargas-Barbosa, S. Vyas, S. Oottikkal, S.F. Manzer, C.M. Hadad, K.D. Glusac, Electrode-assisted catalytic water oxidation by a flavin derivative, Nat. Chem. 4 (2012) 794. [8] M.J. Pouy, E.M. Milczek, T.M. Figg, B.M. Otten, B.M. Prince, T.B. Gunnoe, T.R. Cundari, J.T. Groves, Flavin-catalyzed insertion of oxygen into rhenium–methyl bonds, J. Am. Chem. Soc. 134 (2012) 12920. [9] Z. Shi, J.M. Zachara, L. Shi, Z.M. Wang, D.A. Moore, D.W. Kennedy, J.K. Fredrickson, Redox reactions of reduced flavin mononucleotide (FMN), riboflavin (RBF), and anthraquinone-2,6-disulfonate (AQDS) with ferrihydrite and lepidocrocite, Environ. Sci. Technol. 46 (2012) 11644.

[10] M. Cable, E.T. Smith, Identifying the n = 2 reaction mechanism of FAD through voltammetric simulations, Anal. Chim. Acta. 537 (2005) 299. [11] Y. Astuti, E. Topoglidis, P.B. Briscoe, A. Fantuzzi, G. Gilardi, J.R. Durrant, Protoncoupled electron transfer of flavodoxin immobilized on nanostructured tin dioxide electrodes: thermodynamics versus kinetics control of protein redox function, J. Am. Chem. Soc. 126 (2004) 8001. [12] Y.H. Bi, Z.L. Huang, B. Liu, Q.J. Zou, J.H. Yu, Y.D. Zhao, Q.M. Luo, Two-photon-excited fluorescence and two-photon spectrofluoroelectrochemistry of riboflavin, Electrochem. Commun. 8 (2006) 595. [13] V.C. Diculescu, A. Militaru, A. Shah, R. Qureshi, L. Tugulea, A.M.O. Brett, Redox mechanism of lumazine at a glassy carbon electrode, J. Electroanal. Chem. 647 (2010) 1. [14] B.C. Thompson, O. Winther-Jensen, B. Winther-Jensen, D.R. MacFarlane, A solid-state pH sensor for nonaqueous media including ionic liquids, Anal. Chem. 85 (2013) 3521. [15] D. Zhong, A.H. Zewail, Femtosecond dynamics of flavoproteins: charge separation and recombination in riboflavine (vitamin B2)-binding protein and in glucose oxidase enzyme, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 11867. [16] A. Niemz, J. Imbriglio, V.M. Rotello, Model systems for flavoenzyme activity: oneand two-electron reduction of flavins in aprotic hydrophobic environments, J. Am. Chem. Soc. 119 (1997) 887. [17] S.L.J. Tan, R.D. Webster, Electrochemically induced chemically reversible protoncoupled electron transfer reactions of riboflavin (vitamin B-2), J. Am. Chem. Soc. 134 (2012) 5954. [18] Y. Hui, E.L.K. Chng, C.Y.L. Chng, H.L. Poh, R.D. Webster, Hydrogen-bonding interactions between water and the one- and two-electron-reduced forms of vitamin K1: applying quinone electrochemistry to determine the moisture content of nonaqueous solvents, J. Am. Chem. Soc. 131 (2009) 1523. [19] R. Petrucci, G. Marrosu, P. Astolfi, G. Lupidi, L. Greci, Cyclic voltammetry, spectroelectrochemistry and electron spin resonance as combined tools to study thymoquinone in aprotic medium, Electrochim. Acta 60 (2012) 230. [20] R. Sokolova, I. Degano, S. Ramesova, J. Bulickova, M. Hromadova, M. Gal, J. Fiedler, M. Valasek, The oxidation mechanism of the antioxidant quercetin in nonaqueous media, Electrochim. Acta 56 (2011) 7421. [21] J.B. He, Y. Wang, N. Deng, X.Q. Lin, Study of the adsorption and oxidation of antioxidant rutin by cyclic voltammetry–voltabsorptometry, Bioelectrochemistry 71 (2007) 157. [22] B.K. Jin, L. Li, J.L. Huang, S.Y. Zhang, Y.P. Tian, J.X. Yang, IR spectroelectrochemical cyclic voltabsorptometry and derivative cyclic voltabsorptometry, Anal. Chem. 81 (2009) 4476. [23] M. Dias, P. Hudhomme, E. Levillain, L. Perrin, Y. Sahin, F.-X. Sauvage, C. Wartelle, Electrochemistry coupled to fluorescence spectroscopy: a new versatile approach, Electrochem. Commun. 6 (2004) 325. [24] C. Lei, D. Hu, E.J. Ackerman, Single-molecule fluorescence spectroelectrochemistry of cresyl violet, Chem. Commun. (2008) 5490. [25] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh, C. Fiolhais, Atoms, molecules, solids, and surfaces: applications of the generalized gradient approximation for exchange and correlation, Phys. Rev. B 46 (1992) 6671. [26] J.P. Perdew, Y. Wang, Accurate and simple analytic representation of the electrongas correlation energy, Phys. Rev. B 45 (1992) 13244. [27] B. Delley, An all-electron numerical method for solving the local density functional for polyatomic molecules, J. Chem. Phys. 92 (1990) 508. [28] B. Delley, From molecules to solids with the DMol[sup 3] approach, J. Chem. Phys. 113 (2000) 7756. [29] A. Klamt, V. Jonas, T. Bürger, J.C.W. Lohrenz, Refinement and parametrization of COSMO-RS, J. Phys. Chem. A 102 (1998) 5074. [30] A. Klamt, G. Schuurmann, COSMO: a new approach to dielectric screening in solvents with explicit expressions for the screening energy and its gradient, J. Chem. Soc., Perkin Trans. 2 0 (1993) 799. [31] A.K. Rappe, W.A. Goddard, Charge equilibration for molecular-dynamics simulations, J. Phys. Chem. 95 (1991). [32] M.A. North, S. Bhattacharyya, D.G. Truhlar, Improved density functional description of the electrochemistry and structure–property descriptors of substituted flavins, J. Phys. Chem. B 114 (2010). [33] S. Bhattacharyya, M.T. Stankovich, D.G. Truhlar, J. Gao, Combined quantum mechanical and molecular mechanical simulations of one- and two-electron reduction potentials of flavin cofactor in water, medium-chain acyl-CoA dehydrogenase, and cholesterol oxidase, J. Phys. Chem. A 111 (2007). [34] M.H.V. Huynh, T.J. Meyer, Proton-coupled electron transfer, Chem. Rev. 107 (2007) 5004