Quinone-type carriers achieve redox processes across membranes2-4 and are involved in mitochondrial and photo-synthetic electron and proton transport (ubiquinones, plasto-quinones, vitamin K)4-6. They perform cotransport (symport) of two electrons and two protons through their reduced hydroquinone form. Ferrocene type carriers4 operate by way of oxidation to the ferricinium cation. Electron transfer experiments have been conducted in a cell in which an aqueous solution of a reducing agent (RED: sodium dithionite or ascorbic acid, reduction potentials Eo= - 1.13 and +39 V respectively)7 is separated from an oxidising aqueous solution (OX: potassium ferricyanide, Eo=+0.36V)7 by a diphenyl ether membrane supported on a Millipore filter and containing a suitable carrier molecule. The carriers used were either menadione (vitamin K3; 2-methyl-l ,4-naphthoquinone, NQ, Eo=+0.44 V)7 or dibutyl ferrocene (DBF, Eo= +0.23 V)8. The electron transfer process is followed by monitoring the reduction of ferricyanide to ferrocyanide at 420 nm, where only the former absorbs. In the absence of carrier in the membrane no reduction is observed. When either NQ or DBF are dissolved in the membrane, ferrocyanide is formed in the OX phase showing that electron transfer occurs through the NQ/NQH2 and DBF /DBF redox carrier systems (Table 1). The transport rate increases with increasing concentrations of NO in the membrane although not linearly (Table 1, runs 3-6). The absence of leakage through the membrane was checked at the end of the experiments by ensuring that no iron hexacyanide had passed into the RED compartment.Coupling electron transport to light irradiation should be possible in a system where irradiation would generate a reducing species which would then be able to transfer an electron to a carrier molecule through the interphase, but which would not itself cross the membrane. Methyl viologen (MV2 , I, 1'-dimethy4,4'-bipyridinium chloride) seemed to be a suitable substance; it has been used in many physicochemical and biological (for example, in photosynthesis) redox studies. Reduction of the colourless oxidising form (MY2+) gives a blue cation (MV+) (Eo= --0.44 V)7, which is able to reduce quinones like vitamin K12 and being monocharged should remain in the water phase. Furthemore, MV2+ may be photoreduced in aqueous solution by a photosensitised reaction involving visible light, a sensitising dye, proflavine (PF, Eo= - 0.73V; lmax = 445 nm), and an electron donor, ethylenediamine tetraacetic acid (EDTA)9. Table 1 Electron transport by carrier molecules across an artificial membrane in chemical (runs 1-9) and light-coupled (runs 10-15) redox systemsCarrier in Transport rate Run RED membrane(mM) (mg) (h-1) 1 Asc. 5 NQ(6.8) 1.32 Asc. 5 - 0 3 Asc. 50 NQ(1.5) 0.44 Asc. 50 NQ(3.2) 1.8 5 Asc. 50 NQ(5.8) 3.16 Asc. 50 NQ(10.9) 5.0 7 Dit. 10 NQ(7.2) 24.78 Dit. 10 - 0 9 Asc. 10 DBF(7.2) 3.0PF MV2+ EDTA (M) (mM) (Mm) 1040 1.0 20.5 NQ(6.9) 9.2 110 1.0 20.7 NQ(7.2) 5.0 120 0 20.0 NQ(7.3) 5.0 130 0 0 NQ(7.3) 0 1454 1.0 20.5 NQ(7.5) 7.9 150 0 20.5 NQ(7.2) 0 All experiments were performed at 25 C under argon in a cell consisting of three phases: RED, reducing aqueous solution (40 ml) membrane OX, oxidising solution (40ml); the two aqueous solutions are strirred continuously. Membrane: 75 : 3 mg carrier solution in diphenylether on a Millipore filter (cellulose nitrate 4 cm diameter, 500holes). Carriers: NQ, 2-methyl-l,4-naphthoquinone menadione; DBF, dibutylferrocene. Runs 1, 2:OX, 5 mM potassium ferricyanide, 0.2 M potassium phosphate buffer at pH 7.05 in both RED and OX. Runs 3-8 and 10-15 OX, 10 mM potassium ferricyanide, 0.1 M potassium phosphate buffer at pH 7.0 in OX and at pH 6.2 (runs 3-6) or pH 7.0 in RED. Run 9: OX, 20 mM potassium ferricyanide, 0.1 M sodium citrate buffer at pH 6.37 and 1.2mM sodium bromide in both RED and OX. Asc., ascorbic acid; Dit, sodium dithionite; PF, proflavine; MV2+, methyl viologen dichloride; EDTA, ethylenediamine tetraacetic acid. In the photoreduction experiments (runs 10-15) the concentration of each component of the RED phase is indicated.Light source: runs 10-13: 150W halogen lamp; runs 14 and 15: 200 mW argon laser at 476.5 nm. We have used this system (PF + MV2+ + EDTA) as the RED phase in transport experiments with menadione as the electron carrier in the membrane and ferricyanide in the OX solution, as in previous experiments. When the transport cell is kept in the dark no change in colour or in ferricyanide concentration was observed. When irradiated with visible light from a halogen lamp the RED phase turns rapidly blue, showing that MV+ is produced, and the ferricyanide concentration decreases. If the light is turned off, the ferricyanide concentration levels off; on reillumination it continues to drop. No change in ferricyanide concentration is observed in the absence of menadione in the membrane. These results show that light-coupled electron transport has indeed been achieved by the photoactivated process shown in Fig. 1a. Table 1 (runs 10-13) and Fig. 1b show some typical experiments.Fig. 1 a, Diagrammatic representation of the light coupled transport of electrons across a membrane. The reducing species in the RED aqueous phase at left is the radical cation MV+ produced via proflavine (PF) sensitised photoreduction of methyl viologen MV2+ by the electron donor EDTA. The membrane contains vitamin 1K3 as carrier molecule and the oxidising agent in OX is ferricyanide. b, Typical experiments showing the variation in ferricyanide concentration in the OX phase in three conditions: in the dark ( - - ); under irradiation but with no menadione in the membrane ( . . .); with the complete system shown in Fig. 1a (-). In the latter experiment the conditions were: RED: 50M PF - 5 mM MV2+ L 10 mM EDTA in 0.1 M phosphate buffer at pH 6.5; membrane: 6.9 mg menadione in 72 mg diphenylether on a Millipore filter; OX: 5 mM potassium ferricyanide in 0.1 M phosphate buffer at pH 7.0. From the linear section of the curve the transport rate is 7.1M h-1; the light was switched off at the point indicated by the arrow; the rest of the curve corresponds to the ferriycanide reduced by the MV+ and the hydroquinone present at the time where the light was switched off. The efficiency of the photocoupled electron transport (run 10) is comparable with that of the previous chemical redox systems. A problem with this system is, however, that the RED phase slowly bleaches in the course of the experiment but addition of fresh PF restores the original colour. Because bleached solutions continued to effect electron transport, control experiments were performed. The transport rate is the same in the absence of either PF alone (run 11) or of both PF and MV2+ (run 12), being half the rate of run 10. Thus the complete system is necessary for highest efficiency and MV2+ has no effect in the absence of PF. With EDTA alone, photoactivated electron transport still occurs (run 12), so experiments were conducted under irradiation with an argon laser at 476.5 nm where menadione does not absorb. The transport rate, although similar to that obtained with white light when the whole RED system is present (run 14), now becomes zero with only EDTA in the RED phase (run 15). This confirms that the naphthoquinone carrier may indeed be photo-reduced by EDTA under visible light irradiation with the halogen lamp. We have demonstrated that electron transport processes across artificial membranes using suitable carrier molecules can be coupled to light irradiation. The search for a more stable photosystem (for example metal cation complexes) may lead to applications in energy storage. On the other hand, coupling of either chemically-induced or photo-activated electron transport to cation (other than proton) symport and anion antiport may be envisaged, providing models for biological systems5 as well as new chemical processes and potentially useful transport cells.