The Biological Photosynthetic Reaction Centers Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Biological photosynthetic reaction centers are promising materials for solar energy harvesting, due to their high quantum efficiency. A simple approach in making a photovoltaic device is to apply solubilized RCs and charge carrier mediators in the electrolyte of an electrochemical cell. However, the adsorption of analytes on the electrodes can affect the charge transfer from RCs to the electrodes. In this work, the role of adsorbed RCs from purple bacteria and ubiquinone-10 (Q2) and cytochrome c (Cyt c) (as redox mediators) on a gold electrode in the photocurrent generation and the cycle of charge transfer were studied by a series of photochronoamperometric and X-ray photoelectron spectroscopy (XPS) tests. It was shown that both redox mediators were required for photocurrent generation; hence, the RC itself is likely unable to inject electrons into the gold electrode directly. The reverse redox reactions of mediators at the electrodes generate electrical current. The midpoint potential of electron transfer (ET) cofactors at the RC-gold electrode as well as redox active parts of RC was estimated by cyclic voltammetry (CV). The midpoint potentials of Q2 and Cyt c at pH 8 were measured ~0.042 V and ~0.2 vs. normal hydrogen electrode NHE at bare gold, respectively. Meantime, it was shown that adsorbed RC is electro-inactive without Q2. The voltammogram indicated that both Q2 and Cyt c were electroactive at the RC-modified electrode. RC-adsorbed gold electrode revealed a redox couple due to RC adsorption at ~+0.5 V (vs. NHE), which confirms that the RC does not change its redox properties upon directly coupling to the gold. Further photochronoamperometric studies indicated that adsorbed RCs, strongly bound to the gold surface, are functionally active and contribute significantly to the process of photocurrent generation. Similar experiments showed the adsorption of Q2 and Cyt c on unmodified gold surfaces. It was indicated by the photochronoamperometric tests that the photocurrent comes from Q2-mediated charge transfer from RCs toward the gold electrode by Q2s shuttling forward and backward between RC and gold, while solubilized Cyt c acts as a redox mediator, interacting with the P-side of adsorbed RC and cycle the current toward Pt counter electrode. The Atomic Force Microscopy (AFM) image of the adsorbed RC film shows both RCs aggregation and individual RCs. Also, the stability of the adsorbed RCs and mediators was evaluated by measuring the photocurrent response over a period of time. It is found that ~50% of the adsorbed RCs remain active after a week in aerobic conditions. Significantly extended lifetime is expected by removing oxygen from the electrolyte and sealing the device

Keywords: Bio-photoelectrochemical cell; Rhodobacter sphaeroides; Cytochrome c; Photocurrent; Adsorption; Ubiquinone-10.

1. Introduction

The reaction center (RC) protein complex in photosynthetic bacteria harvests photons and generates spatially separated positive and negative charges with a quantum yield of nearly 100%.1,2 This property of RCs can be exploited to fabricate bio-photoelectrochemical solar cells.1-4 One approach is to immobilize RCs on an electrode using linker molecules.5 Trammell et al. have developed methods for attaching RCs to a conductive electrode with diverse orientations by using appropriate linker molecules.6-8 The electrochemical properties of protein complexes attached to modified electrodes have also been extensively reported.9-16 We have previously demonstrated a diffusion model to explain charge transport between attached RCs and a modified carbon electrode.17 In another approach, the proteins are directly coupled to the electrode (without any linker) to improve the charge transfer by reducing the distance to the electrode. Interesting electronic properties can arise from a direct coupling between proteins and metal surfaces; several studies have investigated charge transfer in RCs directly coupled to a gold surface.18-20 Our team has demonstrated the attachment of a mutant RC, with a single surface-accessible cysteine group, to a gold electrode, in which electrons can tunnel from the gold surface to the immobilized proteins.21 Hollander et al. adhered Rhodobacter Sphaeroides RCs by incubating them onto a bare gold surface, which resulted in stable structure.20 Also, it was shown that the gold surface functionalization is not required for the stable binding and the protein functionality.20 The proteins direct adsorption has been shown elsewhere to be only partially reversible since the proteins may get adsorbed with many segments onto a solid surface and undergo structural changes due to the adsorption.22,23 On the other hand, there might be also some electrostatic repulsion forces between the proteins and the solid surface, which oppose to the adsorption process.24 However, protein conformational changes, which can occur during adsorption process, greatly contribute as the driving force for the adsorption.25,26 Changes in conformation can occur immediately during adsorption or slowly over time after the protein attached to the surface. Recently Frolov et al. reported the fabrication of a photoelectronic device by direct chemical binding of the photosystem I (PS I) RCs to a gold surface through surface exposed cysteines.27

Ciesielski et al. have found that a simple bio-photoelectrochemical solar cell may be constructed by a single-step injection of an electrolyte containing protein complexes (from photosystem I, (PS I)) and charge transfer mediators between a gold cathode and an ITO anode.28 The results of chronoamperometric study indicated a multilayer assembly of PS I complexes on gold over several days.28 The kinetic for the photocurrent production by an electrode modified with a PS I monolayer in the presence of electrochemical mediators is also reported in another work.29

Here, inspired by Ciesielski et al.'s work28, a bio-photoelectrochemical cell was made by injecting wild type RCs from Rhodobacter Sphaeroides and charge carrier mediators (Q2 and Cyt c) into a cell with gold and Pt electrodes. This was intended as a simple method to build a bio-photoelectrochemical cell without the need for any extensive protein incubation. However, the RCs would likely attach to the gold surface due to the cysteine tags on the protein. This would form an RC layer which could be permeable to the mediators. Hence, the formation of an adsorbed layer consisting of RCs and both mediators on the gold surface is expected in such systems. The focus of the present study is to discern the role of adsorbed entities in the charge transfer and photocurrent generation processes. A series of photochronoamperometric, XPS analysis, and AFM tests were performed for this study.

2. Background

In purple photosynthetic bacteria, photochemical energy conversion initiates in a pigment-protein complex spanning the cytoplasmic membrane: the RC.30 The RC of Rhodobacter sphaeroides is comprised of three proteins called L, M and H. The L and M subunits ligate the pigment and other cofactors that make up the RC. The cofactors, which constitute an electron transfer pathway, include a bacteriochlorophyll (BChl) dimer (termed as P, the primary donor), two monomer bacteriochlorophylls (BChlA and BChlB), two bacteriopheophytins (BPheA and BPheB), two quinones (QA and QB) known as electron acceptors, and one non-heme iron are symmetrically arranged in the L and M subunits.19, 31, 32 All cofactors are non-covalently bound to the polypeptides. Figure 1 (a) shows a RC schematic with the protein subunits and cofactors, and the approximate location of surface-exposed cysteine residues. The C156, on the H-subunit, is the most surface-exposed cysteine, and therefore holds the greatest potential for bonding to gold surfaces 5,7,33 (see supporting information paragraph for a complete view of RC complex and all five cysteine residues).

The x-ray crystallographic structures of photosynthetic RCs have contributed significantly to the understanding of the kinetics of electron transfer, and biological electron transfer processes in general.4,34,35 Upon absorption of a photon, P is raised to an excited singlet state (P*) followed by electron transfer from the primary donor to the primary quinone (QA) along the L branch, forming the charge separated state P+QA- (figure 1(b)). Afterwards, the electron is transferred from QA to QB. In vivo, Cyt c acts as a diffusible electron transfer (ET) mediator to reach to the P-side of RC and donate an electron to P. Therefore, oxidized Cyt c is the mediator, carrying the positive charge. After absorption of two photons and receiving two protons, a quinol (QH2) is produced at the QB site. QH2 diffuses out from the protein and acts as an electron carrier mediator. The photosynthetic cycle repeats after the QB vacancy is filled with a fresh quinone (i.e. Q2) (figure 1(b)).

Figure 1. (a) Representation of the RC, protein subunits and cofactors, and approximate location of solvent-exposed cysteine groups, of which C156 is the most externally exposed. White arrow shows the electron transfer path from P to QB. (b) Charge transfer cycle in the RC.

3. Experimental

3.1. Materials

N, N-Dimethyldodecylamine N-oxide solution (LDAO), ubiquinone-10 (Q2), 2-Amino-2-hydroxymethyl-propane-1, 3-diol (Tris buffer) and cytochrome c (Cyt c) from equine heart were purchased from Sigma-Aldrich. Cyt c was reduced (Cyt c2+) by the addition of excess Na2S2O4 to 72 mg of protein dissolved in 6 mL of 0.1 M Tris-HCl buffer (pH 8.0). To remove the excess Na2S2O4, the protein sample was run through a Sephadex G-50 column and an orange/red fraction was collected and analyzed by UV-Vis spectrophotometry (see Supporting Information). The concentration of the reduced form of the protein was calculated using the absorption band at 550 nm (ε = 27.7 mM−−1).36

Wild type RCs from Rhodobacter Sphaeroides were isolated using a modified version of the method of Goldsmith and Boxer.21,37 Briefly, cells were centrifuged at 9000g and resuspended in 10 mM Tris (pH 8), 150 mM NaCl and 2 mM MgCl2. A few crystals of DNAse A were added to the suspension and the cells were broken by two passages through a French press. Broken cells were centrifuged at 9000g to pellet unbroken cells, and the supernatant centrifuged overnight at 30000 rpm in a Beckman Coulter Type 70 Ti rotor to pellet membranes. Membranes were resuspended in 10 mM Tris (pH 8) and 150 mM NaCl and solubilized with 1.5% n, N-Dimethyldodecylamine N-oxide (LDAO). Solubilized membranes were ultracentrifuged at 541 000g, and 6 His-tagged RCs were purified from the supernatant using affinity chromatography.37 The concentration of RCs after purification was determined by their absorption at 804 nm as described by Goldsmith and Boxer (see Supporting Information).38

Planar gold working electrodes (RMS roughness < 2 nm) were fabricated by evaporating an adhesion layer of 20 nm Cr (at a deposition rate of 2 Ao/s) followed by 500 nm gold (at a deposition rate of ~1 Ao/s) onto glass substrates (Varian e-beam evaporator). The gold electrodes were sonicated in isopropanol, rinsed thoroughly several times with deionized (DI) water, and dried completely prior to performing any experiment. An area of ~16.0 mm2 of gold surface was exposed to the electrolyte and the rest of the area was covered with Kapton tape to insulate it from the electrolyte. A 13 cm length of Pt wire of 0.25 mm diameter was shaped to a coil and used as the counter electrode. For the photo-electrochemical measurements, the working electrode (Au), the Pt counter electrode, and an Ag/AgCl 3M NaCl reference electrode (from BASi) were arranged in a standard disposable 4.5 ml spectrophotometer cuvette. 0.1 M Tris-HCl buffer (pH 8.0) was employed as the supporting electrolyte. In the experiments in which the mediators were applied, the concentration of Q2 and Cyt c2+ was 60 μM and 80 μM, respectively. Also, the concentration of RC in the electrolyte was 0.03 µM with 0.1% LDAO.

3.2. Characterization

Optical absorption and transmission measurements were performed using a Thermo Scientific (Evolution 201) UV-Vis spectrophotometer. Photochronoamperometric experiments were performed using a VersaSTAT 4 (Princeton Applied Research) potentiostat in a three-electrode cell setup. The current polarity convention was set in a fashion that defined cathodic current as negative and anodic current as positive. Each cell was kept in the dark until the open-circuit voltage reached a stable voltage. For the photocurrent measurements, the same potential was applied to the cell by the potentiostat such that the current in the dark was very close to zero.23 The cells were illuminated with a solar simulator (RST, Radiant Source Technology, 80 in photochronoamperometric experiments in which the working electrodes and the electrolytes were exposed to periods of darkness and irradiation with pulses of 200 seconds (s) on and 200 s off. For the photocurrent spectrum measurements, white light from a Xe lamp was passed through a monochromator to illuminate the cells for 60 second periods. Current was continuously monitored as the wavelength of illumination was increased from 650 nm to 950 nm in 6 nm steps, with 60 seconds of darkness in between each period of illumination.

For estimation of the midpoint potential of electron transfer (ET) cofactors at the RC-gold electrode, cyclic voltammetry (CV) tests were performed. The midpoint potential was calculated from anodic and cathodic peak potentials. CV measurements were performed in 0.1 M Tris-HCl buffer at pH 8 both in the dark and in light (see Supporting Information) using a three-electrode configuration. For dark measurements, each cell was kept in the dark until the open-circuit voltage reached a stable voltage. For all CV test ~1.9 cm2 of the electrode's surface area was used. In the text, all potentials are reported relative to Normal Hydrogen Electrode (NHE). The scan rate was 10 mV/s for all CV measurements. For CV measurements, the gold electrodes were cleaned by acetone, methanol, isopropanol, and DI water. For degassing, N2 was bubbled into the solution for five minutes before doing any CV test. The N2 tube then removed and left on top of the solution in such a way not to disturb the solution's surface. In this way a N2 blanket was formed over the solution, which prevented the diffusion of O2.

To investigate the adsorption of analytes on the working electrode surface, X-ray photoelectron spectroscopy (XPS) was carried out. In XPS measurements, a Mg Kα X-ray source (hν = 1235.6 eV, 20 mA emission current) was used for standard core level XPS. All samples were prepared on 1 cm2 gold substrates, which were mounted on a sample plate, ensuring good conductivity. The measurements were performed in a commercial photoemission spectroscopy set-up (SPECS GmbH, Berlin, Germany) with a base pressure of 2Ã-10-10 mbar. The data analysis was performed using Igor Pro software.

The morphology of the gold electrode after exposure to RC solution and subsequent rinsing was studied with tapping mode AFM. Samples for the AFM experiments were prepared using e-beam evaporated gold on Si wafers. The substrates were exposed to the diluted RC solution (~0.03 µM) at 4 °C for 1 h. Then, the exposed gold surfaces were rinsed with buffer and DI water and dried under N2 gas flow. The obtained samples were imaged in air using Digital Instruments AFM machine. NanoSensors SuperSharpSilicon™ probe with a typical 2 nm radius of the tip, a spring force constant (C) of 1.3 N/m, and resonant frequency of ~ 61 kHz were employed in the measurements.

4. Results

4.1. Photochronoamperometric Study

Three bio-photoelectrochemical cells were fabricated with electrolytes containing; (i) RC, Cyt c2+, and Q2; (ii) RC and Q2; (iii) RC and Cyt c2+ with the aforementioned concentrations (Experimental). Figure 2 shows the photocurrent densities for these cells measured in three-probe configuration. The photocurrent density was negligible for electrolytes containing RCs and only one mediator (figure 2 (ii) & (iii)). After adding the second mediator (Q2 or Cyt c2+), the photocurrent significantly increased, suggesting that both the photogenerated charges are transferred from RCs to the electrodes via the mediators (indirect charge transfer). The photocurrent is anodic, meaning that electrons are predominantly transferred from the RC into the gold electrode via mediator interactions.

Figure 2. Photochronoamperometric measurements (3-probe configuration) of the bio-photoelectrochemical cells containing RC and either Cyt c, Q2 or both mediators. The potential was set to the value of the open circuit potential in dark. In this case, the open circuit potential was ~0.195 V versus NHE in presence of both mediators in the system. The arrows indicate light ON (↑) and OFF (↓).

A systematic series of experiments was conducted to elucidate whether each analyte (i.e.: RC, Q2, and Cyt c) adsorbed on the working electrode, as well as the role of these adsorbed species in the process of photocurrent generation and charge transfer. The gold electrode was exposed to all three analytes in turn. After measuring the photocurrent, the electrodes were washed thoroughly several times with the buffer, then transferred to a cell containing fresh electrolyte lacking the analyte of interest. If an analyte is significantly adsorbed, the photocurrent likely would be preserved in an electrolyte that lacks the analyte in question. To test the attachment of RCs to the gold electrode, the photocurrent was first measured in a cell containing both Q2 and Cyt c2+ redox mediators but no RCs. As expected, no photocurrent was observed (figure 3 (a), step I). In step II, the photocurrent was measured ~1 hour after RC was added to the electrolyte. By adding RC (step II), a photocurrent was produced, which increased sharply and dropped gradually to ~375 nA/cm2. After step II, the electrode was removed from the cell and rinsed thoroughly several times with the buffer in a way that presumably left only the strongly bound RC attached to the gold surface. The RCs' attachment to the gold electrode was tested by measuring their ability to generate a photocurrent. In step III, the electrode was applied in a new cell with a clean counter electrode and a fresh electrolyte lacking RC and containing both mediators. Similar photocurrents to the one in step II were obtained, indicating that a reasonably stable adsorption of RCs to the gold surface had occurred. RC protein complex is thought to adsorb via the C156, near the cytoplasmic surface on the H-subunit, which resulted in the stable binding between RC and the gold surface (see Supporting Information).

Furthermore, it is indicated that RCs bound to the gold surface were functionally active and significantly contributed to the photocurrent generation. Magis et al. have previously shown that adsorbed photosynthetic membranes onto a gold surface maintain their energy and electron transferring functionality.39 Our results also have shown no indication of RC denaturation, as in other reports demonstrating that redox proteins, deposited directly on a gold electrode, can retain full functionality.20,27

Figure 3. Photochronoamperometric measurements to identify the contribution of adsorbed: (a) RCs, (b) Q2, and (c) Cyt c onto the gold electrode surface in the produced photocurrent. The legends in each figure show the electrolyte in each step. The testing potential was the open circuit potential in dark. The arrows indicate light ON (↑) and OFF (↓).

To test the adsorption of Q2 on the gold electrode surface and its contribution to the produced photocurrent, photochronoamperometric measurements were first carried out in a cell containing RCs and Cyt c2+ (figure 3 (b), step I). Subsequently, in step II, Q2 was added to the electrolyte; after which, the electrode was rinsed and placed in a new cuvette with an electrolyte lacking Q2 (step III). Washing out unbounded Q2s and changing the electrolyte did not reduce the photocurrent, indicating that Q2 was strongly adsorbed at the electrode surface (Figure 3 (b) steps II and III). It has also been reported that quinol (QH2) may possibly be adsorbed on a gold surface causing a reversible electrochemical oxidation which results in the quinone.40

The adsorption of Cyt c on the gold surface and its role in the production of photocurrent was investigated using the same approach as for RCs and Q2. The photocurrent was first measured in a cell having only Q2 (60 µM) and RC (0.03 µM). As expected, the photocurrent was negligible (figure 3 (c), step I). In step II, Cyt c2+ was added to the electrolyte in a concentration of 80 µM, which resulted in an increased photocurrent (figure 3 (c), step II). The electrode was then removed from the cell and rinsed with the buffer before transferring to a new cuvette with fresh electrolyte lacking Cyt c2+. An almost zero photocurrent in the cell lacking Cyt c (step III) indicates Cyt c redox mediators is needed in the bulk of an electrolyte to carry charges to the counter electrode.

4.2. Photocurrent Spectrum of the RC protein complex on the gold electrode

To determine whether the origin of the photocurrent was indeed the RC absorption and charge carrier generation, a photocurrent spectrum was obtained across the range of wavelengths where a distinctive triplet of RC cofactor absorptions are known to occur (bacteriopheophytin at 760 nm, monomeric bacteriochlorophyl at 802 nm and the Bacteriochlorophyl "special pair" at 870 nm).41 The 3-electrode cell, including RC and both mediators, was illuminated with monochromatic light between 650 nm and 950 nm in 6 nm steps, and the resulting photocurrents are presented in terms of incident photon to generated electron quantum efficiency (Figure 4; see supporting information for further details). A convincing match between the RC absorption spectrum and the efficiency of photocurrent generation across this wavelength range strongly supports the conclusion that the photocurrent stems from the light harvesting and charge generation of the RCs. The photocurrent was only generated when the monochromator illumination was focused directly onto the gold electrode, and not just into the surrounding electrolyte. This suggests that the Au-bound RCs constitute the active material.

Figure 4. Electron per photon efficiency spectrum for the 3-electrode cell containing RC and both mediators (black symbols), compared with the absorbance spectrum of RC in 0.1 M tris buffer (red line).

4.3. X-ray photoelectron spectroscopy

Valuable reports have indicated the usefulness of XPS in providing some robust qualitative estimation of the presence and electronic properties of biological materials on different surfaces.11,42 Also, efforts including presenting XPS analysis have been performed by Bourg et al. to provide a detailed insight into the actual nature of Au-S bonding by comparison of the high resolution XPS spectra of 2D and 3D self-assembled monolayers (SAMs) and reference Au (I) complexes43, which can be used for explaining the bond between protein complexes and a gold surface. Here, XPS was used to monitor the adsorption of all analytes i.e. RC complex protein, Q2, and Cyt c. To further validate the strong adsorption of RCs to the gold surface, an XPS measurement of samples was performed after they were exposed to a solution of buffer and 0.03 µM RC. Figure 5 (a-c) shows N1s, C1s, and O1s core spectra for the bare gold, electrodes exposed to only buffer, and electrodes exposed to the RC and mediators dissolved in the buffer. The RC adsorption to the gold electrode after buffer rinsing is illustrated by figure 5 (a). The N1s core level emissions from bare gold surface and buffer treated surface are similar. However, strong N1s emission peaks appear after treatment with a solution containing RC protein complex and buffer. The advent of these peaks confirms the adsorption of RC on the gold surface. The RC is a complex of several proteins and therefore contains N as a part of the polypeptide backbone and in some amino acid side chains, which is evident from the appearance of the N1s emission peak consisting of two components representing different bonding types.

Figure 5. N1s (a), C1s (b), and O1s (c) core level XP spectra measured on e-beam evaporated gold before and after exposing to either RC, pure buffer, Q2 or Cyt c, and subsequent rinsing.

The adsorption of Q2 on the gold electrode was further validated by a XPS test in which rinsed bare gold surface was exposed to a 60 µM solution of Q2. Since Q2 lacks N, the C1s and O1s core level emissions needed to be studied to verify Q2 adsorption. Figure 5 (b) indicates that a small C1s emission peak from surface contamination can be seen on the gold substrate (one observes that by exposing gold to the environment, there will always be C and O emissions). After gold surface treatment in a buffer containing Q2, a strong C1s emission peak was observed, which can be assigned to the adsorption of Q2. The O1s peak from the Q2 treated sample consists of two components.

To validate if the Cyt c binds to the gold surface, XPS analysis was carried out for a fresh gold electrode surface, which was exposed to a buffer solution containing Cyt c2+ (80 µM). XPS has been shown to be a useful instrument for the detection of a small heme protein (e.g. Cyt c) due to the sensitivity of the spectroscopy method to N and C atoms.11,44 XPS results on Cyt c exposed electrodes are shown in figure 5. Figure 5 (b) shows the 285.4 eV peak which likely arises primarily from aliphatic R groups of the polypeptide chain backbone of Cyt c, while carbons bound to oxygen and nitrogen are responsible for the asymmetry observed on the higher binding energy side.11 It is not unexpected that proteins may adsorb on the conductive electrode surface.12 The XPS data confirmed Cyt c adsorption on the electrode surface (figure 5 (a)). As shown in figure 5 (a), adventitious nitrogen is notably absent from bare gold and buffer rinsed gold surface. After treating the gold surface with solution containing Cyt c, a N1s peak (~ 400.8 eV) is obtained from gold surface. Cyt c protein is rich in lysine amino acid residues45, therefore the N1s signal can be assigned to the nitrogen-containing side chains of Cyt c such as lysine and histidine.11 Earlier results by photochronoamperometric tests showed that the adsorbed Cyt c does not contribute to the cycle of charge transfer.

4.4. UV-Vis spectrophotometry

To estimate the amount of RC adsorbed on the gold electrode, the absorption spectrum of the electrolyte containing RC before and after insertion of gold electrode was obtained by a UV-Vis spectrophotometer. The results are shown in figure 6.

Figure 6. Absorption spectra of the electrolyte containing RC (a) before and (b) after insertion and removal of a gold electrode. The original concentration of RC before exposing to the gold electrode (curve a) was ~0.03 µM. The concentration dropped to ~0.0066 µM in (b).

The gold electrode was taken out from the electrochemical cell in less than 30 minutes. The concentration of RCs in the bulk of electrolyte before and after insertion and removal of the gold electrode was estimated by their absorption peak at 804 nm. Figure 6 clearly demonstrates that a significant amount of solubilized RC in the electrolyte were attached to the gold electrode during the aforementioned period, which resulted in a major decrease in the concentration of RCs in the bulk of electrolyte. The concentration dropped from ~0.03 µM to ~0.0066 after exposing the electrolyte to the gold surface. Hence, almost 78% of the solubilized RCs in the bulk of electrolyte were directly adhered to the gold electrode's surface.

4.5. Stability

The arrival of RC protein complex at the gold electrode surface could be driven by diffusion46 and the adsorption likely occurs through the cysteine tags on the protein. For device applications, the stability of the adsorbed RCs is crucial. Naturally, RCs may desorb or denaturize with time. Baszkin and Norde have suggested different techniques to study the protein adsorption/desorption phenomena as well as structural changes.47 In this work, photocurrent amplitude and UV-Vis spectroscopy of the electrolyte are applied for studying the protein stability. The stability of the fabricated bio-photoelectrochemical cells can be evaluated by measuring the photocurrent in the cells over a period of time. Ciesielski et al. also employed similar method to study the stability of their bio-photoelectrochemical cells.28 The photocurrent response was monitored over time for a course of a week, during which the cell was stored at 4-6 oC in aerobic conditions. In the meantime, the possibility of releasing RCs from the gold electrode's surface was investigated by absorption spectroscopy of the electrolyte. UV-Vis spectrophotometer was also employed to study the denaturation of redox mediators.

For stability study, a bio-photoelectrochemical cell was fabricated with an electrolyte containing RC, Cyt c2+, and Q2 with the aforementioned concentrations (Experimental). The gold electrode was kept in the fabricated cell for almost an hour. During this period, RC protein complexes assembled at the gold surface. The electrode then was removed from the cell, rinsed several times with the buffer, and finally was applied in a new cell with a clean reference and counter electrode and a fresh electrolyte lacking RC and containing both mediators. The photocurrent density of ~398 nA/cm2 was measured. Then the electrode was removed and absorption spectrum of the electrolyte was measured to monitor the potential amount of released RCs from the gold electrode's surface. The measurements were repeated over a period of seven days and the results are reported in figure 7 (see supporting information for further details).

Figure 7. Bio-photoelectrochemical cell stability. a) Photocurrent density response over a period of seven days. Peak photocurrent production occurs on the fabrication day, after which it decreases, due to the adsorbed RCs denaturation as well as redox mediators' degradation. In this particular study the cell was refilled with a new electrolyte after seven days and b) Absorption spectra of the electrolyte over a period of seven days, which shows no indication of RCs' release from the surface of the gold; the inset shows Cyt c degradation during time. The black arrow in inset shows the decreasing trend of Cyt c peak (550 nm) from maximum (0 day) to minimum (after 6 days).

As shown in figure 7 (a), the photocurrent density was decreased from ~398 nA/cm2 to ~139 nA/cm2 over the period of a week. In our system three different reasons might be considered for the reduction in photocurrent response over time: release of RCs from the gold electrode's surface, structural/conformational changes of the adsorbed RC protein complex on the gold surface, and denaturation of natural redox mediators. The absorption spectra of the electrolyte imply the existence of a strong bond between the adsorbed RCs and the gold's surface since no absorption peak of native RCs (804 nm) in the electrolyte was observed (figure 7 (b) - the detail of the plots around 804 nm is presented in the supporting information). Therefore, the photocurrent reduction can be attributed to the denaturation of some of the RCs and/or degradation of the redox mediators. To evaluate the effect of mediators' degradation, the cell was refilled with a fresh electrolyte containing only Cyt c and Q2 on the 7th day after the fabrication. The photocurrent density response increased to ~184 nA, which suggests that the decrease in photocurrent overtime was partially due to redox mediators' denaturation. Also, the change in the peak at 550 nm in the absorption spectrum (figure 7(b)), which corresponds to the concentration of Cyt c, confirms the degradation of the mediator. However, the photocurrent production never returned to its initial value after refilling the cell with the fresh electrolyte containing only redox mediators, which shows that the denaturation of a number of RCs adsorbed on the gold's surface is likely due to conformational/structural changes. A quick quantification on the stability of the adsorbed layer shows that ~50% of the adsorbed RCs remain active after a week in aerobic conditions. As shown by Ciesielski et al., it is expected to extend the device lifetime significantly by removing oxygen from the cell (anaerobic condition) and sealing.28

4.6. Atomic force microscopy

In order to characterize the adsorbed RC, the morphology of the gold electrode surface was studied using Atomic Force Microscopy (AFM) after exposure to 0.03 µM RC solution. AFM micrograph of the surface is displayed in figure 8. As it can be seen, gold surface after exposure to the RCs and subsequent rinsing and drying (under N2 stream) exhibits a relatively rough surface (figure 8 (a)) with the RMS (root-mean-square) roughness of ~3.4 nm, which could be an implication of protein adsorption to the surface. The topographic image of the adsorbed RC film reveals RCs aggregation on the gold surface as well as RC particles with height of ~6-7 nm (figure 8(b)), which could be an individual RC protein. As explained by Trammell et al., the broadening of single proteins in the AFM image is caused by the finite size of the apex of the cantilever.8 In general, the apparent lateral dimensions of all AFM-imaged proteins are overestimated when the geometry of the probe tip is comparable to the size of a protein molecule.48 Meantime, one should note that the larger particles than what demonstrated in figure 8 (a), show the aggregation of several proteins and not the individual RCs.

Similar studies were recently performed by Mukherjee et al. on the morphology of PS I from aqueous buffer suspensions onto alkanethiolate SAM/Au substrates, which showed the formation of complex columnar structures rather than a uniform monolayer formation due to solution phase protein aggregations.16,49 In this work, additional AFM tests were also performed with higher concentration of RC protein stock (~0.8 µM and ~ 14.5µM) to show the effect of protein solution's concentration on surface morphology and roughness (see supporting information for further details).

Further morphology and section analysis of the adsorbed RC film (from 0.03 µM RC stock) after an hour demonstrates particles with lower height (~4.7 nm), which is lower than one might expect for a RC protein complex (see supporting information). As suggested by Trammell et al.,8 this might be due to protein flattening (denaturing due to the absence of water) on the electrode and is consistent with previous observations on the RC stability in section 4.5.

Figure 8. (a) An AFM topographic image of directly adsorbed RCs on a gold surface and (b) a section analysis along the black solid line in (a) showing the heights of the protein particles. The images were obtained in air using noncontact tapping mode.

4.7. Estimation of the midpoint potential of ET cofactors at the RC-gold electrode in the dark

Due to the potential direct interaction of the mediators with the gold electrode, the midpoint potential values of redox mediators (Q2 and Cyt c) in our system and redox active parts of the RC protein complexes were assessed by CV. CV data was recorded for various combinations of cofactors with and without adsorbed RC in dark conditions. Also similar CV tests were performed in light to show the effect of illumination on the electrochemical potential values (see supporting information). Figure 9 depicts the voltammograms for 60 µM Q2 in the electrolyte at a freshly cleaned bare gold electrode and at the gold electrode with directly adsorbed RC. The CV for the adsorbed-RC electrode without Q2 in solution is also shown for comparison.8 The reported CV data for electroactive Q2 in figure 9 indicates that the midpoint potential of Q2 at pH 8 is ~0.042 V versus NHE at the bare gold. The results are well consistent with the earlier reported values by different groups.7,40,50 As shown previously by Trammell et al., cyclic voltammograms for Q2 at the RC-adsorbed gold electrode illustrate catalytic anodic and cathodic currents, which suggest the RCs have adsorbed on gold. Also, the adsorbed RCs considerably block the background current between Q2 and electrode.8 One should note that the difference between the value of the peak currents in this work and those reported by Trammell et al.,8 is due to the difference in the electrode area and the Q2 concentration. Meantime, as it has been shown elsewhere, adsorbed RC is electroinactive without Q2.

Cyclic voltammograms over a more oxidizing voltage region for the RC-adsorbed gold electrode reveal a redox couple due to the RC adsorption. Figure 9 (b) demonstrates a redox couple at ~+0.5 V (vs. NHE), which is similar to that reported by Trammell et al.8 This confirms that the RC does not change its redox properties upon directly coupling to the gold. One should note that the CV tests were performed instantly after RC incubation and keeping the electrochemical cell in dark for half an hour. Figure 9 (c) depicts voltammogram for a RC-adsorbed gold electrode after addition of 80 µM Cyt c to the electrolyte, with 0.1 V.s-1 and 0.5 V.s-1 scan rates. As shown in figure 9 (c), upon Cyt c addition to the electrolyte, new peak will be introduced, which can be assigned to Cyt c (for comparison see supporting information for CVs of RC-adsorbed electrode without Cyt and with Cyt c in solution side by side). The results are consistent with previous reports by Trammell et al.8 The voltammogram indicates that Cyt c is electroactive at the RC-modified electrode. The CV of Cyt c at bare gold surface indicates midpoint potential of ~+0.2 V (data are not shown here).8 Millo et al. reported 0.244 V midpoint potential for Cyt c immobilized on smooth gold surfaces chemically modified with 11-mercaptounodecanoic acid.51 Arrows in figure 9 (c) show redox couple of the primary donor and Cyt c. The slight shift in RCs' primary donor peak position after Cyt c addition is due to the formation of the super-molecular complex between RC and Cyt c.8 The peak cathodic current of RCs' primary donor in figure 9 (c) was decreased significantly after keeping the cell in dark for 9 hours (see supporting information for further details).

Figure 9. CVs of (a) 60 µM Q2 in solution measure with a clean bare gold electrode (blue trace) or gold electrode with adsorbed RC (red trace). CVs of gold electrode with adsorbed RC protein complex but without Q2 (black trace), (b) CV of a RC-modified electrode, and (c) CVs of the RC-modified electrode after addition of 80 µM Cyt c to the electrolyte with different scan rates; 0.1 V.s-1 (bold black line) and 0.5 V.s-1 (narrow red line).


5. Discussion

The UV-Vis spectrophotometry study (figure 6) showed that about 78% of the RCs in the electrolyte adsorb on the gold surface. Also, the results from both XPS (figure 5) and AFM (figure 8) studies indicate that the adsorbed RC layer is fairly stable with a strong adhesion to the gold. Due to the presence of the cysteine C156 on the surface of the H-subunit, it is likely that RCs would be attached to the gold electrode from the H-subunit side (see figure 10) with a preferential orientation.5,33 The rest of the RCs (~22%) would be adsorbed at the counter electrode (Pt) and be free floated in the electrolyte. However, the small differences between the amount of photocurrents in steps II and III of figure 3 (a) indicates that the contribution of the RCs in the electrolyte and adsorbed on the counter electrode in generating photocurrent at the gold electrode (working electrode) is negligible.

Despite the adsorption of RCs on the gold electrode, the measurements in figure 2 suggest that the photogenerated charges in the RCs are not able to transfer to the electrode directly. Hence, both types of mediators are required for the photocurrent generation. As explained, Cyt c3+ and QH2 act as the positive and negative charge transfer mediators, respectively (see figure 1(b)). In an electrochemical cell containing two mediators in the electrolyte, the photocurrent would be the superposition of anodic (QH2Q+2e-+2H+) and cathodic (Cyt c3++e-Cyt c2+) reactions at the electrode surfaces.17,52 Since in this work the current polarity convention was set in a fashion that defined anodic current as positive; the observed anodic currents imply domination of electron transfer from the RCs to the gold electrode via QH2/Q reaction. The shape of the photocurrent with a rapid increase and gradual drop suggests concurrency of both anodic and cathodic reactions again with domination of anodic reaction from conversion of QH2 to Q. Similar shapes of the photocurrent also were observed in another work when using two electrochemical mediators.53 The kinetic of the reactions in a bio-photoelectrochemical cell with one mobile mediator has been studied by Ciesielski et al.28 However, studying the kinetics in a system with more than one redox reaction is complicated and may require designing additional experiments to measure the reaction rates separately. Furthermore, previous studies suggest that the anodic current in the cell can be limited by the diffusion of QH2 but not the kinetic of the redox reaction.17 Therefore, before analyzing the thermodynamic of the system, an appropriate model for the reactions in the device must be developed, which is not in the scope of this work. Nevertheless, the net anodic current in all the photochronoamperometric measurements indicates the domination of QH2/Q reaction at the gold electrode.

The XPS results indicate the capability of gold to adsorb Q2. Also, the results in figure 3(b) suggest that the charge transfer is mediated mainly by the quinone adjacent to the electrode, while the contribution of dissolved Q2 in the electron transfer between the attached RCs and gold is negligible. This implies that Q2s shuttle back and forth between the RCs and the electrode17,20 (figure 10).

Although XPS results show that Cyt c can be adsorbed on gold, the photocurrent measurements in figure 3(c) indicate that only dissolved Cyt c contributes to the photocurrent cycling toward counter electrode. A solubilized redox mediator (i.e. Cyt c) is required in the electrolyte to complete the charge transfer pathway. As long as there is ample Cyt c in the solution, current cycling is enabled via Cyt c in the bulk of electrolyte toward the Pt counter electrode. In this system, solubilized Cyt c interacts with the P-side of the RC protein complex, taking the positive charges and diffusing toward the Pt electrode surface to be reduced again.28 Therefore, the adsorbed Cyt c on the gold electrode surface does not contribute to the current cycling. However, adsorbed Cyt c can significantly contribute to the photocurrent cycling when modified RCs are attached to the gold surface from the P-side as it was shown by Lebedev et al.54 In that case, the use of chemical coupling agents may possibly allow for a covalent attachment of metalloproteins to electrode surfaces, which can be of prime importance in the construction of biosensors55, optoelectronic devices56, and other applications.57-59

Figure 10 shows a summary of the discussion in the form of a cartoon of electron transfer events between RCs, electrochemical mediators, and the electrodes surface

Figure 10. The electron transfer events between RCs, electrochemical mediators, and the gold electrode. RCs are most likely attached to the gold surface through C156 cysteine. The path of electron transfer inside RCs is shown by with arrows. The observed anodic photocurrent is generated by dominated electron carrier mediators (QH2/Q) which shuttle back and forth between the RCs and the gold electrode. The positive charge at the P-site is transferred to the counter electrode by Cyt c in the bulk of the electrolyte. Interaction of Cyt c with the gold electrode limits the photocurrent.

It should be stated that since both mediators are required to sustain the photocurrent, the electrode can interact with both mediators to donate and accept electrons, which is inefficient for photocurrent generation. However, we expect significant improvements will be made by employing semiconducting electrode materials with suitable energy levels for selective charge transfer.60 A detailed study of the structure and electrochemical properties of such protein based photovoltaic devices when using wide band gap semiconductors to increase the photocurrent is underway in our group and will be reported in forthcoming papers.

5. Conclusions

A bio-photoelectrochemical cell has been constructed by injection of RCs suspension, Q2, and Cyt c between gold and Pt electrodes. Here, we described formation of an adsorbed layer containing RCs and both mediators on the surface of the gold electrode. Cysteine C156 on the surface of H-subunit facilitated the expected preferential orientation of RCs on the gold surface. Photochronoamperometric results suggested Q2-mediated charge transfer on gold electrodes from RCs toward gold, while current cycling is enabled via solubilized Cyt c in the bulk of the electrolyte toward the Pt counter electrode. In the absence of either one of the mediators (Q2 or Cyt c2+), the photocurrent was almost zero. Hence, it appears that indirect charge transfer dominates in these experimental bio-photoelectrochemical cells. The attachment of RCs and mediators to the gold surface was strong enough and was further approved by XPS, AFM, and stability analysis. The stability of the adsorbed RCs and mediators was evaluated by measuring the photocurrent response produced by the same cell over a period of time. A quick quantification on the stability of the adsorbed layer shows that ~50% of the adsorbed RCs remain active after a week in aerobic conditions. Further stability can be achieved by applying anaerobic conditions and sealing the device. RC-adsorbed gold electrode revealed a redox couple due to RC adsorption at ~+0.5 V (vs. NHE), which confirms that the RC does not change its redox properties upon directly coupling to the gold. The voltammogram indicated that both Q2 and Cyt c were electroactive at the RC-modified electrode. The topographic image of the adsorbed RC film, formed from 0.03 µM diluted RC solution, reveals the presence of large particles, which resulted from RCs aggregation on the gold surface as well as RC particles with height of ~6-7 nm, which is typical for a RC protein.

Supporting Information Available: Additional optical characterizations of the electrochemical system, including absorbance spectra of the RC suspension and the Cyt c charge carrier mediator, the complete view of RC with all five cysteine residues, detailed explanation on determination of the photocurrent action spectrum, the influence of illumination on cyclic voltammograms at 10 mV/s for Q2 at bare gold and adsorbed RC (pH 8 buffer), and Q2 at adsorbed-RC gold electrode, photocurrent plots for the cell stability study, absorption spectra of the electrolyte for the detail around 804 nm in stability study, an AFM topographic image of directly adsorbed RCs on a gold surface and a RC protein particle after one hour exposure to aerobic conditions, as well as changes in RMS roughness of the adsorbed RC film on the gold as a result of increasing the concentration of RC solution, and 3D AFM topographic images of the adsorbed RC film on a gold surface from 0.03 µM, 0.8 µM, and 14.5 µM RC stock are shown. This material is available free of charge via the Internet at