Review
Biofuel cells and their development

https://doi.org/10.1016/j.bios.2006.01.030Get rights and content

Abstract

This review considers the literature published since 1994 on microbial and enzymatic biofuel cells. Types of biofuel cell are classified according to the nature of the electrode reaction and the nature of the biochemical reactions. The performance of fuel cells is critically reviewed and a variety of possible applications is considered. The current direction of development of biofuel cells is carefully analysed. While considerable chemical development of enzyme electrodes has occurred, relatively little progress has been made towards the engineering development biofuel cells. The limit of performance of biofuel cells is highlighted and suggestions for future research directions are provided.

Introduction

Biological fuel cells (biofuel cells) have been defined as fuel cells that rely on enzymatic catalysis for at least part of their activity (Palmore and Whitesides, 1994). A broader definition would consider biofuel cells as those fuel cells which use biocatalysts, which includes systems utilising non-enzyme proteins as well. In the broadest sense, we would define biological fuel cells are devices capable of directly transforming chemical to electrical energy via electrochemical reactions involving biochemical pathways.

The connection between biology and electricity has been known since the experiments of Galvani in the 1780s (Galvani, 1791), when it was discovered that current from a static electricity generator could cause a severed frog's leg to twitch, revolutionising the understanding of the nervous system. The fuel cell has been known for almost as long, since Grove successfully reversed the action of the electrolysis of water, recombining hydrogen and oxygen to produce water and electrical current (Grove, 1839). Given the early nature of these two discoveries it is perhaps surprising that a half cell using microorganisms was not demonstrated until 1910, when M.C. Potter, a professor of Botany at the University of Durham observed electricity production by E. coli (Potter, 1910). These results were not widely reported until the experiments of Cohen in 1931 who demonstrated a voltage greater than 35 V from microbial fuel cells connected in electrical series (Cohen, 1931).

The expansion of interest in fuel cells triggered by the USA space program, in the late 1950s and early 1960s, led to the development of microbial biofuel cells as a possible technology for a waste disposal system for space flights that would also generate power. Also in the late 1960s, the biofuel cell using cell-free enzyme systems began to be used, with the early goal of a power supply for a permanently implantable artificial heart (Kreysa et al., 1990, Wingard et al., 1982) and references within).

In a fuel cell, an oxidation reaction occurs at the anode and a reduction reaction at the cathode. The oxidation releases electrons, which travel to the cathode via the external circuit doing electrical work. The circuit is completed by the movement of a compensating charge through the electrolyte often in the form of positive ions.

Conventionally, fuel cells operate using relatively simple inorganic chemistries, taking such fuels as hydrogen or methanol (MeOH), and producing energy, water and carbon dioxide (in the case of methanol). Other fuels, such as other lower order alcohols and alkanes, are also used, but they are frequently reformed to produce hydrogen before the fuel cell process (Mitsos et al., 2004, Vielstich et al., 2003).

Conventional fuel cells are regarded as ’low temperature’ if they operate in the region of 80 °C and typically require expensive p-group catalysts (Larminie and Dicks, 2000). A summary of the main types of inorganic fuel cell is presented in Table 1.

Biofuel cells use enzymes as catalysts (alone or within an organism) and tend operate under mild conditions (20–40 °C, near-neutral pH). These properties make biofuel cells an attractive development prospect for use in applications where generating high temperatures is difficult, or where harsh reaction conditions are undesirable. Biofuel cells are not, in theory at least, limited to these mild conditions as extremophile organisms or enzymes derived from them should be able to operate under a wide variety of reaction conditions. Also the variety of reactions able to be catalysed by enzymes makes the use of a much wider range of fuel substances possible (in some systems, up to the use of molecules as large as soluble starches). Biocatalysts, either protein, enzyme or whole organism, can also offer cost advantages over metallic catalysts, although this is not likely to be the case until the consumption of the enzyme is sufficient to merit large scale production. Widespread usage of a biocatalyst would however tend to lower the cost of producing it, while this is not true of transition metal catalysts. Previous review papers pertaining to biofuel cells or relevant similar technologies are summarised in Table 2. This paper seeks to collate descriptions and performance information for all biofuel cells reported since 1994.

Presently there are two practically applied systems; a test rig operating on starch plant wastewater (microbial fuel cell system), which has been operating for at least 5 years has been demonstrated as a bioremediation method (Gil et al., 2003) and as a biological oxygen demand (BOD) sensor (Kim et al., 2003a), and also a biofuel cell has been employed as the stomach of a mobile robotic platform ‘Gastronome’, designed as the precursor to autonomous robots that can scavenge their fuel from their surroundings (gastrobots). The original Gastronome ‘eats’ sugar cubes fed to it manually (Wilkinson, 2000), but other groups have refined the concept somewhat to produce predators consuming slugs (Kelly and Melhuish, 2001), or flies (Ieropolous et al., 2004), although so far they both still require manual feeding. Many applications have been suggested however, and several of these are in varying stages of development.

The most obvious target for biofuel cells research is still for in vivo applications where the fuel used could be withdrawn virtually without limit from the flow of blood to provide a long-term or even permanent power supply for such devices as pacemakers, glucose sensors for diabetics or small valves for bladder control. The early goal of powering an artificial heart has been abandoned as the difficulty of realising a functioning implant has become more apparent. Once small systems are in place and functioning, this challenge may well be revisited. A biofuel cell capable of operation implanted in a grape (Mano et al., 2003a) shows that progress is being made in this. The challenge of biocatalysis over a suitably long period is particularly problematic in these areas, where surgical intervention could be required to change over to a new cell and ethical constraints are paramount.

Ex vivo proposed applications are diverse. The large scale is represented by proposed power recovery from waste streams with simultaneous remediation by bioelectrochemical means, or purely for power generation in remote areas, the medium scale by power generating systems for specialist applications such as the gastrobot above, and perhaps of greatest potential the small scale power generation to replace battery packs for consumer electronic goods such as laptop computers or mobile telephones. The larger scale applications tend to be organism based and the smaller scale ones more likely to be enzymatic. In the case of enzymatic fuel cells, at least, the major barrier to any successful application is component lifetime, particularly in view of the limited enzyme lifetime and problems of electrode fouling/poisoning.

For any fuel cell the power output, Pcell is likely to be the bottom line determining its value.Pcell=EcellIdtOr, if the current is constant Pcell = EcellI.

The overall cell voltage that can be derived from any electrochemical system is given by:Ecell=ECEAIReEC and EA are experimental potentials of the cathode and anode, respectively. These electrode potentials can be related to the equilibrium potential for the couple by:η=EEe

The equilibrium potential is in turn related to the formal potential Ee0 for the couple (the potential when the reduction and oxidation potentials are equal) by:Ee=Ee0+RTzFlncOcRTo obtain an optimal voltage from a cell it is desirable to maximise the driving force (EC  EA) and to minimise the ohmic resistance losses, IR. The latter can be achieved through appropriate cell design considerations such as minimising the inter-electrode gap or operating at low current density. Low current density operation is, however, counter to the generation of a high power density.

For a fuel cell to be viable, its total energy budget (power output from the cell minus power consumed by pumping, stirring, sparging and any other ‘utilities’) must be positive. The utility costs must be minimised by appropriate chemical or physical design or economies of scale, or may be eliminated altogether in some situations (a conventional battery has no pumping costs, relying purely on diffusion, an implanted biofuel cell in vivo could rely on the heart to perform the pumping duty, etc.).

As one of the target applications for biofuel cells is portable power (particularly for enzymatic cells) consideration must be given to the size of a proposed device. For the fuel cell itselfPowerdensity=PcellAandVolumetricpower=PcellVe=AeAPcell

For a complete fuel cell system it is also necessary to consider the fuel as adding volume and, for man-portable systems especially, weight. Another useful metric to consider is:Specificpower=Pcellmcell,fuel(s),utilities

It is important to review the fundamental definitions relevant to biofuel cells. A generalised schematic of a biofuel cell half cell is presented in Fig. 1, and an overview of different types of power generating electrochemical devices is presented in Fig. 2. This overview allows us to consider the different definitions that a device has to conform to in order to be a biofuel cell, and allows the characterisation of different biofuel cell-like devices.

Electrochemical devices convert energy directly from a chemical to an electrical form. This allows them to bypass the thermodynamic limitations of combustion based power cycles. Furthermore, in an electrochemical system, the rate of energy production can be finely controlled through an external circuit.

A battery is an electrochemical device that allows a reaction to occur at an anode and a cathode, the fuel being contained internally by the battery housing (Gileadi, 1993).

A fuel cell is a similar device in which fuel is supplied from outside the housing and wastes resulting from a spent fuel are removed. It is not a requirement that the fuel is added continuously, merely that it is stored and supplied from outside the reaction chambers (Gileadi, 1993).

Bridging the gap between batteries and fuel cells are hybrid devices and redox flow cells. Hybrid devices such as metal-air cells are batteries in the anode compartment, oxidising metals to their ions in solution while being fuel-cell-like in the cathode compartment with air from the atmosphere supplying oxygen to be reduced to form hydroxyl ions (Haas et al., 2003, Vielstich et al., 2003). Redox flow cells are composed of an electrochemical cell with large volumes of liquid fuel stored locally but externally in a closed loop. Power can be supplied to the cell to drive a charging set of reactions. The energy can be stored in chemical form until the cell is discharged to generate current. As the charged chemical is stored outside the main body of the electrochemical cell this can be considered as resembling a fuel cell but it cannot truly be fuelled indefinitely (Divisek and Emonts, 2003).

Any type of electrochemical device will require both a fuel and a catalyst, and by considering these, further definitions can be arrived at:

Inorganic devices utilise inorganic chemistries and usually employ metals as catalysts (Palmore and Whitesides, 1994), for example the classic Pt-catalysed H2/O2 fuel cell (Ecell0=1.23V),H2Pt/anode2H++4e12O2+2H++4ePt/cathodeH2Oor an Al/Air hybrid battery.

Organically fuelled devices operate using more complex chemicals as fuel sources, but still rely on inorganic catalysts to achieve the reactions as in, for example, the direct methanol fuel cell (DMFC) (Ecell0=1.20V).CH3OH+H2Oinorganic_anodeCO2+6H++6e32O2+6H++6einorganic_cathode3H2O

Biologically catalysed devices (usually also organically fuelled) utilise biological molecules as catalysts for achieving their redox reactions (in at least one half cell), either as purified enzymes (or enzyme derivatives) to catalyse a specific reaction, or through the use of whole organisms (Palmore and Whitesides, 1994). For methanol oxidation (11) a nicotinamide adenine dinucleotide (NAD) mediated sequence of reactions is (Palmore et al., 1998):CH3OH+2NAD+alcohol_dehydrogenaseCH2O+2NADHCH2O+H2O+2NAD+aldehyde_dehydrogenaseHCOOH+2NADHHCOOH+2NAD+formate_dehydrogenaseCO2+2NADH3NADH+6Mediatorreddiaphorase3NAD++6Mediatorox6Mediatoroxanode6Mediatorred+6e

In practice, working cell voltages tend to be ≪0.5 V due to kinetic restrictions.

This latter category combined with the various types of electrochemical device allows for the definition of the biological battery (biobattery) and biological fuel cell (biofuel cell). A bio-redox flow cell has not yet been demonstrated, nor has any suitable chemistry for such a device been proposed.

Biologically catalysed devices can be divided according to the source of the biocatalysts (usually enzymes) used to perform the catalysis.

Microbially catalysed systems use whole living organisms as the source of complete enzyme pathways. These are generally robust systems that can operate on variable feedstocks and are resistant to poisoning due to being living systems, and are usually capable of oxidising the substrate completely to carbon dioxide and water. For example direct electrochemical communication of Shewanella putrefaciens with an anode via membrane cytochromes (Kim et al., 2002a).Lactatemicrobial_metabolismwastes+(membrane_cytochrome)(membrane_cytochrome)anode(membrane_cytochrome)+e

Enzymatically catalysed systems use isolated and purified enzyme proteins to perform specific reaction catalysis (reactions (13)–(17)). This generally gives a better defined system and several enzymes can be combined to completely oxidise the substrate, although structuring the enzymes so that the different reactions can occur in the right order is problematic. In some cases protein units smaller than complete enzymes can be used as catalysts, but these cases can be treated as enzymes for fuel cell purposes.

Electrons can transfer between the reaction site and the electrode through the diffusion of a secondary fuel, via a mediator molecule that repeatedly cycles or via a direct electron transfer (DET) between the reaction site and the electrode.

There has been some variation in the literature on how to define these types of systems. Aston and Turner (1984) define a system in which a secondary fuel for the electrode is generated by the biological reaction as indirect biofuel cells, with cases involving electron-shuttling reversible mediators or direct electron transfer between the biological component and the electrode being direct biofuel cells. However Palmore and Whitesides (1994) define indirect biofuel cells as any that include a diffusive component, hence according to their definition a system involving diffusing reversible mediators is indirect, but one involving non-diffusing mediators is direct. Higgins and Hill (1985) defined three categories; product type (organisms or enzymes convert electrochemically inactive species to electrochemically active species), regenerative type (organisms or enzymes regenerate redox compounds which in turn carry out the electrochemical reaction) or depolariser type (organisms or enzymes act as catalysts of the electrochemical reactions at the electrode, or themselves diffuse to the electrode and transfer electrons directly). These definitions, while internally consistent and comprehensive are unsatisfactory, particularly the use of the word ‘depolariser’ which is misleading in the context of electrochemistry.

Enzymatic cells have been divided into mediated electron transfer (MET) and direct electron transfer (DET), with DET covering only systems where the electron tunnels directly from the active site fixed in the enzyme to the electrode, and MET encompassing all forms of regenerative mediation whether diffusive or non-diffusive (Calabrese Barton et al., 2004). Product type operation was not covered.

This paper will also consider DET in the case using whole microorganisms where electrons transfer directly between the surface of a microorganism and an electrode. MET will similarly be used to describe microbiological fuel cells where a mediator molecule is used as an electron relay.

Section snippets

Bioelectrochemical cells involving a whole organism

The use of whole organisms as a biofuel cell system allow multiple enzymes and hence multiple substrates (or mixed substrates) to be used, with the organisms contained in the fuel cell system acting as micro-reactors providing appropriate conditions for each enzyme to act optimally. Recent examples of biofuel cells making use of whole organisms as catalysts are tabulated in Table 3.

Evolution has produced highly efficient electron control systems within living organisms. This informs a search

Fuel cells utilising a purified enzyme

Fuel cells making use of a purified enzyme have arisen from a desire to have specific and defined reaction(s) occurring in the fuel cell. Many proposed uses for this type of system are in vivo applications, such as self-powering glucose sensing for use in diabetics (Katz et al., 2001, Quinn et al., 1997) or the long-standing goal of the implantable power supply for a cardiac pacemaker. As such, several technologies for DET and MET biofuel cells of this type have developed in parallel or

Review of progress

Palmore and Whitesides (1994) have identified the overall objective of biofuel cell research as to establish whether biofuel cells are real contenders for practical use. It was noted that the stability and activity of a hydrogen-oxidising anode provide a high performance; hence it is difficult for biochemical fuels to compete with the relatively high power density of gaseous hydrogen (Tsujimura et al., 2001a) and liquid fuels such as methanol. Biofuel cell research has shown success in

Acknowledgements

The authors are grateful to DSTL and EPSRC for the award of a CASE studentship to RAB.

References (103)

  • N.L. Akers et al.

    Electrochim. Acta

    (2005)
  • F.A. Armstrong et al.

    Electrochim. Acta

    (2000)
  • S.C. Barton et al.

    Biosens. Bioelectron.

    (2002)
  • P. Bianco et al.

    J. Electroanal. Chem.

    (1990)
  • Y. Choi et al.

    Bioelectrochemistry

    (2003)
  • M.J. Cooney et al.

    Enzyme Microb. Technol.

    (1996)
  • S. Cosnier

    Biosens. Bioelectron.

    (1999)
  • D. Das et al.

    Int. J. Hydrogen Energy

    (2001)
  • G.C. Gil et al.

    Biosens. Bioelectron.

    (2003)
  • T. Ikeda et al.

    Biochim. Biophys. Acta Proteins Proteomics

    (2003)
  • T. Ikeda et al.

    J. Biosci. Bioeng.

    (2001)
  • E. Katz et al.

    Anal. Chim. Acta

    (1999)
  • E. Katz et al.

    J. Electroanal. Chem.

    (1999)
  • H.J. Kim et al.

    Enzyme Microb. Technol.

    (2002)
  • N. Kosaric et al.

    FEMS Microbiol. Rev.

    (1995)
  • S.V. Morozov et al.

    Bioelectrochemistry

    (2002)
  • J. Niessen et al.

    Electrochem. Commun.

    (2004)
  • G.T.R. Palmore

    Trends Biotechnol.

    (2004)
  • G.T.R. Palmore et al.

    J. Electroanal. Chem.

    (1998)
  • G.T.R. Palmore et al.

    J. Electroanal. Chem.

    (1999)
  • H.S. Park et al.

    Anaerobe

    (2001)
  • A. Pizzariello et al.

    Bioelectrochemistry

    (2002)
  • C.A.P. Quinn et al.

    Biomaterials

    (1997)
  • S. Tsujimura et al.

    Enzyme Microb. Technol.

    (2001)
  • S. Tsujimura et al.

    Electrochem. Commun.

    (2003)
  • I. Willner et al.

    Bioelectrochem. Bioenerget.

    (1998)
  • L.B. Wingard et al.

    Enzyme Microb. Technol.

    (1982)
  • T. Wong et al.

    Curr. Opin. Biotechnol.

    (2003)
  • N.L. Akers et al.

    Preprint Pap. Am. Chem. Soc. Div. Fuel Chem.

    (2003)
  • R.M. Allen et al.

    Appl. Biochem. Biotechnol.

    (1993)
  • W.J. Aston et al.

    Biotechnol. Genetic Eng. Rev.

    (1984)
  • H.P. Bennetto et al.

    Biotechnol. Bioengin.

    (1983)
  • S. Calabrese Barton et al.

    Chem. Rev.

    (2004)
  • S.K. Chaudhuri et al.

    Nat. Biotechnol.

    (2003)
  • T. Chen et al.

    J. Am. Chem. Soc.

    (2001)
  • Y. Choi et al.

    Bull. Korean Chem. Soc.

    (2003)
  • B. Cohen

    J. Bacteriol.

    (1931)
  • J. Divisek et al.

    Energy storage via electrolysis/fuel cells

  • L. Galvani

    De bononiensi scientiarum et artium instituto atque academia Comentarrii

    (1791)
  • E. Gileadi

    Electrode Kinetics for Chemists, Chemical Engineers and Materials Scientists

    (1993)
  • B.A. Gregg et al.

    J. Phys. Chem.

    (1991)
  • B.A. Gregg et al.

    J. Phys. Chem.

    (1991)
  • W.R. Grove

    Philos. Mag. Ser.

    (1839)
  • O. Haas et al.

    Metal/air batteries: the zinc/air case

  • L. Habermuller et al.

    Fresen. J Anal. Chem.

    (2000)
  • A. Heller

    J. Phys. Chem.

    (1992)
  • A. Heller

    Phys. Chem. Chem. Phys.

    (2004)
  • I.J. Higgins et al.

    Essays Biochem.

    (1985)
  • I. Ieropolous et al.
  • K. Kano et al.

    Electrochemistry

    (2003)
  • Cited by (0)

    View full text