Direct Harvesting of Electricity from Plant Photosynthesis

Takane Usui
November 28, 2010

Submitted as coursework for Physics 240, Stanford University, Fall 2010

Fig. 1: Calculated power density versus light intensity with a linear fit showing efficiency.

Bioenergy is becoming a popular research subject in a field of alternative energy. Biomass is generally viewed as an important contribution to an abundant, carbon-neutral renewable energy resource. [1] Studies indicate that by the year 2050, the demand for bioenergy can potentially increase above 1020 joules per year. [2]

The photosynthetic apparatus absorbs excitation energy and generates chemical bound energy, much of which is in the form of fixed carbon. [3] So far, most attempts to utilize photosynthetic plants focus on usage of such fixed carbon, mostly saccharides, as a fuel. However, this approach inherently suffers from efficiency losses because every process after photosynthesis involves non-perfect efficiency. In this regard, extracting electrons as close as possible to the plant's chloroplast would maximize the efficiency. There are a few attempts of this nature - extracting electrical energy directly from photosynthesis, before the energy is used for carbon fixation. In this report, these attempts are reviewed and the efficiencies of these approaches are discussed.

Chloroplast Photosynthesis Mechanism

Light-dependent reactions occur in several complexes on a membrane of thylakoid, a compartment inside a chloroplast. Photon energy is first absorbed in photosystem II (PSII), splitting water and exciting two electrons. These electrons reduce a plastoquinone, which carries electrons through the lipid membrane to a complex called cytochrome b6f. The electrons are then carried through inside of the plant cell by a plastocyanin to photosytem I (PSI). At PSI under light, an electron and a hydrogen ion are added to NADP to NADPH. This enzyme NADPH, together with an ATP, generates an organic compound Glyceraldehyde 3-phosphate (C3H7O6P), which is the most primary unit of fixed carbon. [4]

Enzyme Biosensors Embedded in Cactus

Flexer et al. proposed a method of directly measuring oxygen and glucose (C6H12O6) generated during photosynthesis. [5] In their experiment, a set of macro-scale electrodes were inserted in a living cactus plant under light. The electrodes were biosensors based on immobilized redox enzymes, i.e., they are designed so that the concentration of O2 and glucose can directly be monitored. When the cactus leaf was illuminated with a light of an intensity 250 W/cm2 for 200 seconds, the currents from both the oxygen sensing and glucose sensing electrodes showed a clear increase. When the light was turned off, the currents exponentially decayed to the baseline values.

If this system was to be used as a power source, it can be thought of as a fuel cell with two half reactions:

Anode: 2C6H12O6 → 2C6H10O6 + 4H+ + 4e-
Cathode: O2 + 4H+ + 4e- → 2H2O

resulting in the net reaction:

2C6H12O6 + O2 → 2C6H10O6 + 2H2O

The plotted power indeed showed a round peak that resembles to that of a fuel cell. The reported peak power achieved was, at 0.4 V bias, 9 μW/cm2 under 250 W/cm2 illumination. Thus the corresponding conversion efficiency of the proposed power source, analogous to solar cells, can be calculated as:

η = 9 × 10-6 / 250 = 3.6 × 10-8

Even if this was sustainable, it is apparent that the efficiency is far inferior to commercial solar cells on the market; commercial thin film modules routinely achieve 4-10%, and wafer-based modules achieve 10-16% [6].

Nanoprobe Inserted in a Single Algal Cell

While Flexer's method was originally optimized for observing and understanding photosynthesis kinetics rather than extracting electricity, Ryu et al. attempted to probe photosynthesis with their main focus on energy extraction. [7] In this work, the authors separated a single cell of algae called chlamydomonas reinhardtii, trapped the cell in a microchannel and inserted a nanoscale probe into the cell using an atomic force microscopy apparatus. This is a more controlled way of experiment in a sense that they are only looking at a single cell, and that a single cell of chlamydomonas only contains one chloroplast (it is still questionable, however, how effective this step is. After all there are more than one thylakoid stacks in a chloroplast, and it was unclear if the probe tip was located inside or outside the thylakoid membrane).

Symbol Unit Value(s) Description Justification
J pA {0.6, 1.5, 2.3, 4.1} Current extracted from one cell Read from plot (Fig. 5(a) of ref [7].)
i μmol photon / m2s {20, 51, 86, 134} Light intensity Read from plot (Fig. 5(a) of ref [7].)
V mV 400 Bias voltage Described in published text
λ nm 1000-1500 Wavelength Approximation of halogen lamp spectrum
n m-2 4 × 108 Approximate number of cells trapped per unit area Estimated from photograph of experimental setup (Fig. 3 of ref. [7]).
Table 1: Summary of parameters used in the calculations.

Neither the measured power density nor efficiency was reported in this publication. We attempt to estimate the areal power density and the conversion efficiency using factors given in the publication, and making reasonable assumptions on other variables, which are summarized in Table 1. One important piece of information that was missing from this publication is the spectrum of their light source, necessary to convert the unit of light intensity from moles of photons to watts. Because their light source was described as a halogen lamp, the spectrum was approximated as a rectangle between wevelengths 1000 nm and 1500 nm, based on an actual spectrum of halogen lamps. [8] The wavelength range was varied to estimate the error induced by this approximation, and the error was shown to be only as much as 20%. Also it was assumed that all the cells trapped in their microchannel apparatus contributed to the total power density.

Using these values, the light intensity I in W/m2 and the power density p of this system can be estimated as:

I = ∫ i (NA/Δ λ) h (c/λ) dλ = i (NA/Δ λ) hc ln[λ]
p = J V n

where NA is Avogadro's number, h is Planck constant, and c is the speed of light. The result of this calculation is plotted in Fig. 1, with a linear fit indicating the efficiency of this system, η = 1.1 ×10-4.

The difference in the efficiencies of Ryu's study compared to Flexer's study is understandable. In calculating the areal power density in Ryu's case, we assumed that every chlamydomonas cell trapped in the microchannel can be utilized as a power source; this will not practically be the case since the size of their electrode was larger than the spacing between the trapped cells. In Flexer's case, a macroscopic electrode was inserted in a cactus leaf, meaning not every chloroplast in the cactus leaf was utilized as a power source. In either case, however, the conversion efficiency is again far below commercial solar cells.

Array Fabrication and Electrical Connection

Our calculation attempt in the previous section illuminates the practical need for arraying photosynthetic cells with high number density, allowing electrical connection. Otherwise the conversion efficiency dramatically drops just because much of the incoming light is wasted by passing in between cells. Although the electrical wiring is still a challenge, there is an interesting effort to fabricate an array of photosystems on a semiconducting substrate.

Frolov et al. reported that they successfully deposited a layer of PSI from cyanobacteria on a GaAs substrate. [9] In this work, small (10-20 nm in length) linker molecules formed a self-assembled monolayer, and chemisorbed on a GaAs surface through their carboxyl end. Cyanobacteria PSI then formed covalent bonding to the linker molecules with a specific orientation. So far they have only demonstrated these cells as photodetectors, but by changing the operating bias regime, there is a possibility that this can operate as a solar cell.


By looking at the conversion efficiencies, these attempts to directly extract electrical energy from plants are far from being useful as an alternative energy source, although the efforts are interesting. Contrary to the expected high efficiency [7,9], the overall efficiency in terms of light intensity and generated power density are very low. Many photons are wasted passing through the inactive area of the experimental setup, and many electrons go not collected by the external circuit. The design of an apparatus, where the photosystems can be densely arrayed and electrically connected, will be the key challenge.

© Takane Usui. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] A. J. Ragauskas et al, "The Path Forward for Biofuels and Biomaterials," Science 311 (2006).

[2] G. Berndes, M. Hoogwejk and R. van den Broek, "The Contribution of Biomass in the Future Global Energy Supply: A Review of 17 Studies," Biomass and Bioenergy 25, 1 (2003).

[3] R. E. Blankenship, Molecular Mechanisms of Photosynthesis (Oxford, 2002).

[4] G. M. Cooper and R. E. Hausman, The Cell: A Molecular Approach, Fourth Edition (Sunderland, 2006).

[5] V. Flexer and N. Mano, "From Dynamic Measurements of Photosysthesis in a Living Plant to Sunlight Transformation into Electricity," Anal. Chem. 82, 4 (2010).

[6] M. A. Green, "Thin-film Solar Cells: Review of Materials, Technologies and Commercial Status," J. Mater. Sci.: Mater. Electron S15, 18 (2007).

[7] W. Ryu et al., "Direct Extraction of Photosynthetic Electrons from Single Algal Cells by Nanoprobing System," Nano Lett. 1173, 10 (2010).

[8] M. Ohmi and M. Haruna, "Ultra-High Resolution Optical Coherence Tomography (OCT) Using a Halogen Lamp as the Light Source," Opt. Rev. 5, 10 (2003).

[9] L. Frolov et al., "Photoelectric Junctions Between GaAs and Photosynthetic Reaction Center Protein," J. Phys. Chem. C 112 (2008).