Photocurrents from photosystem II in a metal oxide hybrid system: Electron transfer pathways

Abstract

We have investigated the nature of the photocurrent generated by Photosystem II (PSII), the water oxidizing enzyme, isolated from Thermosynechococcus elongatus, when immobilized on nanostructured titanium dioxide on an indium tin oxide electrode (TiO2/ITO). We investigated the properties of the photocurrent from PSII when immobilized as a monolayer versus multilayers, in the presence and absence of an inhibitor that binds to the site of the exchangeable quinone (QB) and in the presence and absence of exogenous mobile electron carriers (mediators). The findings indicate that electron transfer occurs from the first quinone (QA) directly to the electrode surface but that the electron transfer through the nanostructured metal oxide is the rate-limiting step. Redox mediators enhance the photocurrent by taking electrons from the nanostructured semiconductor surface to the ITO electrode surface not from PSII. This is demonstrated by photocurrent enhancement using a mediator incapable of accepting electrons from PSII. This model for electron transfer also explains anomalies reported in the literature using similar and related systems. The slow rate of the electron transfer step in the TiO2 is due to the energy level of electron injection into the semiconducting material being below the conduction band. This limits the usefulness of the present hybrid electrode. Strategies to overcome this kinetic limitation are discussed.

Keywords: Photosynthesis; Photosynthetic reaction centre; Protein electrode interface; Protein film photoelectrochemistry; Quinone; Water oxidizing enzyme.

Graphical abstract

Fig. 1

(A) Schematic representation of the arrangement of cofactors involved in the electron transfer chain in Photosystem II based on the 1.95 Å crystal structure (PDB reference UB6) . The numbers represent the order of electron transfer steps after charge separation. Step 1 represents both the charge separation and the first stabilization step (see text) forming the radical pair. The black arrows indicate potential exit routes for electrons from the quinones Q_A and Q_B to the protein surface. (B) Scheme of the orientation of PSII on the TiO₂/ITO electrode and indication of the electron transfer steps after charge separation. The two possibilities of electron transfer from the enzyme to the electrode are indicated.

Fig. 2

Photocurrent response from PSII multilayers (A) and monolayers (B) adsorbed onto a nanostructured TiO₂/ITO electrode surface in the absence (left traces) and presence (right traces) of the redox mediator 100 μM DCBQ. Note the bigger scale for right trace in A (multilayers plus mediator). Note: The bar at the top shows the length of the illumination periods, 10 s for the trace on the left and 20 s for the one on the right (see experimental section).

Fig. 3

Photocurrent response from PSII multilayers (A) and monolayers (B) adsorbed onto a nanostructured TiO₂ film in the absence and presence of the mediator DCBQ and the herbicide DCMU. (A) The photocurrent recorded from PSII multilayers (first trace), the presence of 100 μM of DCBQ (second trace) and 10 μM of DCMU (third trace). (B) The photocurrent recorded from a PSII monolayer (first trace) in the presence of 100 μM of DCBQ (second trace) and in the presence of 100 μM DCBQ and 10 μM DCMU (third trace). Note: The bar at the top shows the length of the illumination periods, 10 s for the trace on the left and 20 s for the others (see experimental section).

Fig. 4

(A) Oxygen evolution measurements of PSII in the presence of 0.5 mM of DCBQ, 1 mM of ferricyanide (1); 0.5 mM of DCBQ, 1 mM of ferricyanide, 50 μM of DCMU (2); 0.5 mM of DCBQ (3); 0.5 mM of DCBQ, 50 μM of DCMU (4); 1 mM of ferricyanide (5); 1 mM of ferricyanide, 50 μM of DCMU (6); 0.5 mM of quercetin (7); 0.5 mM of quercetin, 50 μM of DCMU (8); 0.5 mM of PpBQ (9); 0.5 mM of PpBQ, 1 mM of ferricyanide (10) and 0.5 mM of PpBQ, 50 μM of DCMU (11). Error bars are indicated in gray. (B) Photocurrent response from PSII immobilized onto TiO₂/ITO electrode as a monolayer; unmediated, in the presence of 100 μM of DCBQ, 100 μM of quercetin or 100 μM of PpBQ in the measuring buffer. Note: The bar at the top shows the length of the illumination periods.

Fig. 5

(A) Model illustrating the DCBQ-mediated electron transfer from PSII to ITO when a blocking layer of crystalline TiO₂ was present between the nanoporous TiO₂ layer and the ITO surface. Upon illumination electron transfer occurs between Q_A and the mesoporous TiO₂ (1). The function of the blocking layer is to block the interaction of the mobile redox mediator DCBQ (2) with the ITO surface (3). (B) Bar chart showing the mediated and non-mediated photocurrent from the immobilized PSII in the absence and presence of a blocking TiO₂ layer on the ITO electrode. Photocurrent density recorded from (from left to right) in the absence of the blocking layer, without and with the mediator 100 μM DCBQ and in the presence of the blocking layer, without and with 100 μM DCBQ. Error bars are indicated in gray.

Fig. 6

Model describing the electron transfer from PSII to ITO in the absence (A) and presence (B) of the electron mediator DCBQ. (A) Upon illumination electron transfer occurs from Q_A to the TiO₂ (1) and slow electron transfer occurs on the TiO₂ surface (2) due to the poor conducting properties of the material at an applied bias of + 644 mV vs SHE. (B) The redox mediator DCBQ provides an alternative pathway for the electrons by picking up electrons from the TiO₂ surface (2) and transferring them directly to ITO (3).

Fig. 7

Schematic representation of the electron transfer steps from PSII to the TiO₂/ITO electrode following the charge separation event and upon the application of a bias potential of + 0.644 V vs SHE. The energetic levels of the PSII , , cofactors involved in the electron transfer process are shown with respect to the position of the conduction band (E_c), the Fermi level (E_f) imposed by the bias and the valence band (E_v) of TiO₂.