Bioresource Technology 158 (2014) 231–238

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Using planktonic microorganisms to supply the unpurified multi-copperoxidases laccase and copper efflux oxidases at a biofuel cell cathode

http://dx.doi.org/10.1016/j.biortech.2014.02.0380960-8524/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +49 761 203 73218; fax: +49 761 203 73299.E-mail addresses: sabine.sane@imtek.uni-freiburg.de (S. Sané), katrin.richter@

kit.edu (K. Richter), claude.jolivalt@upmc.fr (C. Jolivalt), Catherine.Madzak@grignon.inra.fr (C. Madzak), zengerle@imtek.uni-freiburg.de (R. Zengerle), johannes.gescher@kit.edu (J. Gescher), kerzenma@imtek.uni-freiburg.de (S. Kerzenmacher).

Sabine Sané a, Katrin Richter b, Stefanie Rubenwolf a, Nina Joan Matschke a, Claude Jolivalt c,f,Catherine Madzak d, Roland Zengerle a,e, Johannes Gescher b, Sven Kerzenmacher a,⇑a Laboratory for MEMS Applications, IMTEK – Department of Microsystems Engineering, University of Freiburg, Georges-Koehler-Allee 103, 79110 Freiburg, Germanyb Institute for Applied Biosciences, Department of Applied Biology, Karlsruhe Institute of Technology, Fritz-Haber-Weg 2, 76131 Karlsruhe, Germanyc Sorbonne Universités, UPMC Univ Paris 06, UMR 7197 Laboratoire de Réactivité de surface, F-75005 Paris, Franced INRA, UMR 1319 Micalis, Domaine de Vilvert, F-78352 Jouy-en-Josas, Francee BIOSS – Centre for Biological Signalling Studies, University of Freiburg, 79110 Freiburg, Germanyf CNRS, UMR 7197 Laboratoire de Réactivité de surface, F-75005 Paris, France

h i g h l i g h t s

� Supply of unpurified laccase and copper efflux oxidase to a biofuel cell cathode.� Using crude culture supernatant from planktonic microorganisms.� Live cells at a cathode provide copper efflux oxidase.� Easy to operate and cost effective enzymatic biofuel cells.

a r t i c l e i n f o

Article history:Received 7 January 2014Received in revised form 7 February 2014Accepted 10 February 2014Available online 17 February 2014

Keywords:LaccaseCopper efflux oxidaseBiofuel cellCathodeBioelectrochemistry

a b s t r a c t

The feasibility to apply crude culture supernatants that contain the multicopper oxidases laccase or cop-per efflux oxidase (CueO) as oxygen reducing catalysts in a biofuel cell cathode is shown. As enzyme-secreting recombinant planktonic microorganisms, the yeast Yarrowia lipolytica and the bacterium Esch-erichia coli were investigated. The cultivation and operation conditions (choice of medium, pH) had dis-tinct effects on the electro-catalytic performance. The highest current density of 119 ± 23 lA cm�2 at0.400 V vs. NHE was obtained with the crude culture supernatant of E. coli cells overexpressing CueOand tested at pH 5.0. In comparison, at pH 7.4 the electrode potential at 100 lA cm�2 is 0.25 V lower. Lac-case-containing supernatants of Y. lipolytica yielded a maximum current density of 6.7 ± 0.4 lA cm�2 at0.644 V vs. NHE. These results open future possibilities to circumvent elaborate enzyme purification pro-cedures and realize cost effective and easy-to-operate enzymatic biofuel cells.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Biofuel cells (BFC) have great potential for the eco-friendly di-rect conversion of biochemically stored energy into electricity(Bullen et al., 2006; Osman et al., 2010, 2011). A prominent appli-cation is to combine the treatment of wastewater with the gener-ation of electricity (Logan, 2005), hydrogen, or fine chemicals(Rosenbaum et al., 2011). Furthermore, implantable, glucose

powered fuel cells which supply medical implants (Cinquin et al.,2010; Kerzenmacher et al., 2008), or miniature biofuel cells thatdigest organic matter to power energy-autonomous robots(Ieropoulos et al., 2012) are currently under development.

Usually, enzymatic catalysts or the biochemical pathways ofcomplete microorganisms are employed to catalyse the electrodereactions in biofuel cells. In this way, the use of expensive andunsustainable noble metal catalysts is circumvented and renew-able electrode materials such as carbon can be utilized as elec-trodes (Bullen et al., 2006; Lapinsonniére et al., 2012; Osmanet al., 2010, 2011).

However, to bring biofuel cells from the lab-scale into practicalapplication, in particular the cathode performance needs to beimproved (Harnisch and Schröder, 2010; Schaetzle et al., 2009).

232 S. Sané et al. / Bioresource Technology 158 (2014) 231–238

A preferred oxidizer for biofuel cell cathodes is oxygen, because ofits abundance and comparably high redox potential of 0.918 V vs.NHE at pH 5 and 30 �C (calculated according to (Logan, 2008)). Fur-thermore, the reaction product of the 4-electron reduction of oxy-gen is harmless H2O (Ivnitski and Atanassov, 2007; Schaetzle et al.,2009). In this context, enzymatic catalysts such as bilirubin oxi-dases and laccases enable the reduction of O2 at a lower overpoten-tial compared to microorganisms (Rosenbaum et al., 2011) or evenexpensive noble metal catalysts such as platinum (Lapinsonniéreet al., 2012; Mano et al., 2003; Soukharev et al., 2004), and there-fore promise higher power densities. However, the drawback ofusing enzymatic electrodes is their short lifetime in the range ofseveral weeks at best due to gradual deactivation and loss of cata-lytic activity (Rubenwolf et al., 2011). Furthermore, the prevailinguse of purified enzymes is associated with technical efforts, andtherefore additional costs (Lapinsonniére et al., 2012; Liu et al.,2014; Schaetzle et al., 2009).

In a previous publication, the use of crude culture supernatantof the white-rot fungus Trametes versicolor to supply the unpurifiedenzyme laccase to a biofuel cell cathode was successfully demon-strated (Sané et al., 2013). This operational strategy decreases timeand costs needed for purifying enzymes. In addition, it was shownthat by regular exchange of the crude culture supernatant it is pos-sible to extend the lifetime of the cathode by at least 5-fold.

However, a disadvantage of using fungi as an enzyme source isthat these organisms grow in cell agglomerates and therefore arecomparably difficult to cultivate in a continuous manner. To contin-ually provide enzymes to a biofuel cell cathode it is preferable tocultivate planktonic microorganisms such as yeasts and bacteria.

The aim of the present work is thus to explore, under whichconditions planktonic enzyme-secreting microorganisms can beused to supply multicopper-oxidases to an oxygen reduction cath-ode via untreated culture supernatant. These investigations are animportant step towards the realization of extended lifetime enzy-matic fuel cells, in which enzyme-secreting microorganisms growin an electrode-integrated micro-bioreactor to continually supplyfresh enzymes. To supply multi-copper oxidases to a biofuel cellcathode the recombinant strains of the yeast Yarrowia lipolyticaYL4 and the bacterium Escherichia coliCueO constructed in the pres-ent work are compared.

The recombinant yeast Y. lipolytica YL4 secretes laccase IIIb fromthe fungus T. versicolor (Jolivalt et al., 2005). The oxygen reductionpotential of laccase from T. versicolor can be as high as 0.856 V vs.NHE at pH 5 and 30 �C (Sané et al., 2013) which makes laccase anattractive candidate as a cathode catalyst in biofuel cells(Lapinsonniére et al., 2012). When expressed by Y. lipolytica,laccase IIIb exhibits higher activity towards the redox indicatorABTS (2,2 Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)diammonium salt) at acidic pH, and shows little to no activity ata pH of 6 and above (Madzak et al., 2006). Some electrochemicalactivity has been achieved under physiological conditions (pH7.4) with the help of osmium-based redox polymers as mediators(Barrière et al., 2006).

The recombinant bacterium E. coliCueO was constructed tosecrete the laccase-like enzyme copper efflux oxidase (CueO).Compared to laccase, CueO shows a higher overpotential towardsoxygen reduction with an onset potential of oxygen reduction ofaround 0.649 V vs. NHE (0.450 V vs. Ag/AgCl (Miura et al., 2009)).However, in experiments with rotating disc electrodes it was foundthat CueO exhibits an about 9 times higher limiting catalytic cur-rent compared to laccase (Miura et al., 2007). Similar to laccase,CueO shows no activity towards ABTS at pH 7 (Supporting informa-tion S1), but reportedly exhibits electrochemical activity in thebroad range from pH 2–8 (Tsujimura et al., 2008).

Both, laccase and CueO have the advantage of performing directelectron transfer. They do not require a mediator which shuttles

electrons between the enzyme and the electrode (Christensonet al., 2004; Miura et al., 2009). In general, mediators can increasethe overpotential (Harnisch and Schröder, 2010) and at the sametime become an additional cost factor (Lapinsonniére et al.,2012). For both microorganisms, the influence of cultivationparameters on the enzyme activity in crude culture supernatantwas investigated. Subsequently, the electro-catalytic activity to-wards the cathodic oxygen reduction was assessed by means ofrecording polarization curves for culture supernatant as well ascomplete cultures with live E. coli cells at pH 5.0 and pH 7.4.

2. Methods

2.1. Expression of laccase IIIb by the yeast Y. lipolytica

For the expression of laccase IIIb from T. versicolor the yeaststrain Y. lipolytica YL4 (Jolivalt et al., 2005) was used. The non-lac-case-producing strain Y. lipolytica Po1g (Madzak et al., 2004), whichwas also used as the host for the heterologous laccase expression ofY. lipolytica YL4, served as control. Both strains were maintained onYPD (Yeast Extract Peptone Dextrose) agar plates (10 g/l yeast ex-tract, 20 g/l vegetable peptones, 20 g/l agar, 0.1 M glucose mono-hydrate, Sigma–Aldrich, Taufkirchen, Germany).

A pre-culture (15 ml) was inoculated with Y. lipolytica from aYPD plate and cultivated overnight. For enzyme production, 50 llof the pre-culture were transferred to 50 ml of liquid medium ina 100 ml baffled Erlenmeyer flask. The cultures of Y. lipolytica wereshaken at 180 rpm and 30 �C.

For cultivation, YNB medium (yeast nitrogen based medium),containing 6.7 g/l yeast nitrogen base (Carl Roth, Karlsruhe, Ger-many), 50.5 mM glucose monohydrate, and 20 mM 2,2 dim-ethylsuccinic acid (both Sigma–Aldrich, Taufkirchen, Germany)was used. Furthermore, modified PPB medium (peptone phosphatebuffered medium) according to (Jolivalt et al., 2005) and SCL med-ium (synthetic complete laccase medium) according to (Sané et al.,2013) were investigated. All media were supplemented with 1 mMCuSO4 (Merck KGaA, Darmstadt, Germany) and the pH adjusted toa value of 5.0 with 10% acetic acid (Carl Roth, Karlsruhe, Germany).

Aliquots of the supernatants were collected daily to measurethe enzyme activity towards ABTS (Sigma–Aldrich), as describedin Section 2.3.

2.2. Cloning and expression of copper efflux oxidase by the bacteriumE. coli

The multi copper efflux oxidase CueO is a periplasmic enzymeexported via the twin-arginine-transport machinery. In order to ex-port the protein into the medium, the CueO gene was fused to thefirst 430 nucleotides of the pulA gene. PulA encodes for the pullula-nase protein of Klebisiella oxytoca, an exoprotein that degrades thepolysaccharide polymer pullulan. With the help of the Pul-secretonexpressed from a second plasmid (pCHAP710) the fusion constructcan be secreted into the medium (Francetic and Pugsley, 2005).Therefore, CueO was amplified using primer pBad_CueO_for (TCCATACCCGTTTTTTTGGGCTAGAAATAATTTT-GTTTAACTTTAAGAAGGA-GATATACATACCATGCAACGTCGTGATTTCTT) and CueO_rev (GAAGAGGATCCGTTATCCATTACCGTAAACCCTAACATCA). The primerscontain homologous regions to the pBAD plasmid and the 50 regionof the pulA gene, respectively. The first 430 nucleotides of pulA wereamplified using primer pulA_for (ATGGATAACGGATCCTCTTC) andpBAD_pulA_rev (CAGGCTGAAAATCTTCTCTCATC-CGCCAAAACAGCCAAGCTGGAGACCGTTTTTAAAACGGTCTGGTCCCAGG). The ampli-fied fragment contained sequences complementary to the pBADplasmid. After digesting the plasmid vector using the restrictionsenzymes PmeI and NcoI, the three fragments were combined in

S. Sané et al. / Bioresource Technology 158 (2014) 231–238 233

one step using the in vitro ligation method according to (Gibsonet al., 2009). For expression of copper efflux oxidase, the con-structed plasmid pBADcueO_pulA was transformed into E. coli DH5aZ1(E. coliZ1) (Lutz and Bujard, 1997) resulting in E. coli DH5aCueO(E. coliCueO). E. coliZ1 was used as a non-CueO producing controlstrain.

E. coliCueO was maintained on selective LB plates (10 g/l LB-med-ium, 10 lg/ml kanamycin, 34 lg/ml chloramphenicol (Carl Roth),10 g/l Agar (Sigma–Aldrich)). The E. coliZ1 strain used for controlexperiments was maintained as a liquid culture in non-selectiveLB medium (10 g/l LB-medium, Carl Roth).

For the expression experiments, E. coli was cultivated at 37 �C in50 ml liquid medium in a 100 ml baffled Erlenmeyer flask and sha-ken at 180 rpm. LB medium (10 g/l LB-medium, Carl Roth) wasused either without pH adjustment (pH 7.4), or buffered at pH5.6 or 5.5 with 0.1 M MES buffer (2-(N-morpholino)ethanesulfonicacid, Carl Roth).

Furthermore, minimal medium at pH 7.0 (Schuetz et al., 2009),and SCCueO medium (synthetic complete CueO medium) at pH 7.0containing 6.7 g/l yeast nitrogen base (Carl Roth) as well as 10 g/lglucose monohydrate (50.5 mM) and 1.6 g synthetic drop out sup-plement w/o uracil (Sigma–Aldrich) were used.

Whenever plasmid maintenance was necessary, media were sup-plemented with 10 lg/ml kanamycin and 34 lg/ml chloramphenicol.

A pre-culture (2 ml) was inoculated with E. coliCueO from a selec-tive LB plate. After 6 h of cultivation, the main culture (50 ml) wasinoculated with 50 ll of the pre-culture. One hour after the inocu-lation of the main culture, the CueO production was induced with1 mM arabinose and the secretion machinery of the Pul systemwas induced with 0.2% (w/v) maltose.

During the growth of E. coliCueO cultures in LB medium (pH 7.4)in triplicate, the enzyme activity towards ABTS was monitoredover a period of 24 h after induction. To be able to collect samplesevery hour, the 24 h were divided in three time intervals usingthree batches. To assess the long term CueO activity in the super-natant of growing E. coliCueO cultures in different media, aliquots ofsupernatant were collected each day for a total of 4 days.

2.3. Characterization of enzyme activity towards ABTS

The enzyme activity was quantified spectrophotometrically bymonitoring the oxidation of ABTS at 405 nm and 30 �C in 3.17 sintervals (Sané et al., 2013) using a Wallac Victor2 microplate read-er. All experiments were conducted in triplicate.

To measure laccase activity of Y. lipolytica YL4, 100 ll of 2 mMABTS in 0.1 M sodium acetate buffer (pH 5.0) were added directlybefore the measurement into 100 ll sterile filtered supernatant.

The CueO activity of E. coliCueO was quantified by injecting150 ll of 3.3 mM ABTS and 2.5 mM CuCl2 (Serva, Heidelberg, Ger-many) in 0.1 M sodium acetate buffer (pH 5.0) into 100 ll E. coliCueO

supernatant, as adapted from (Grass and Rensing, 2001). Theabsorbance measurement was started after one minute ofincubation.

One unit of enzyme activity was defined as the amount of crudeextracellular protein extract required to oxidize 1 lmol of ABTS perminute at pH 5.0 and 30 �C.

2.4. Influence of the pH value on E. coliCueO growth and CueO activitytowards ABTS

To investigate the influence of the pH value on the E. coliCueO

growth rate, 50 ml of LB medium in triplicate were adjusted to dif-ferent pH values and inoculated with 500 ll of an E. coliCueO pre-culture with an OD of 0.06. After three hours the optical density(OD) was measured at 600 nm using a Wallac Victor2 microplatereader (Perkin Elmer, Rodgau – Jügesheim, Germany).

The optimal pH value for CueO activity towards ABTS wasdetermined by using CueO in supernatants of E. coliCueO grown inLB medium (pH 7.4, 37 �C). The supernatant was divided into threealiquots for each pH value, and the pH was adjusted to values be-tween pH 3 and 8 using 0.1 M NaOH (Merck KGaA, Darmstadt, Ger-many), or 10% acetic acid. The activity was measured by injecting150 ll of 3.3 mM ABTS and 2.5 mM CuCl2 in distilled water into100 ll E. coliCueO supernatant. The absorbance was measured as de-scribed in Section 2.3. One unit of enzyme activity was defined asthe amount of crude extracellular protein extract required to oxi-dize 1 lmol of ABTS per minute at the given pH and 30 �C.

2.5. Setup of the electrochemical reactors

To characterize the electrochemical activity of the culturesupernatants, stepwise galvanostatic cathodic polarization curveswere recorded in half-cell set-ups as described elsewhere in detail(Kloke et al., 2010; Sané et al., 2013). Buckypaper electrodes (Hus-sein et al., 2011) (0.9 cm2) made of carbon nanotubes(1.4 ± 0.3 mg cm�2, Baytubes C150 HP, Bayer Material Science AG,Germany) dispersed onto a nylon filter support (Whatman, Dassel,Germany) were used as cathodes. A saturated calomel referenceelectrode (SCE, KE11, Sensortechnik Meinsberg, Germany;0.244 V vs. normal hydrogen electrode NHE) was used to recordelectrode potentials. All potentials are referenced to the normalhydrogen electrode (NHE). In accordance with the pH value atthe cathode, the counter electrode compartments were either filledwith 0.1 M sodium acetate buffer (pH 5.0) or 0.1 M potassiumphosphate buffer (pH 7.4).

For the experiments using supernatant or the correspondingcontrols, the cathode compartment was either filled with 4 ml (Y.lipolytica experiments) or 6 ml (E. coli experiments) sterile filteredcatholyte and purged with humidified air.

In addition to experiments with crude culture supernatant,E. coliCueO and E. coliZ1 were also directly cultivated in the electrodecompartment. In this case, the reactor setups adapted from (Klokeet al., 2010) were modified with a larger electrode compartment tohold a culture volume of 30 ml.

2.6. Recording of cathode polarization curves using Y. lipolytica

To compare polarization curves, supernatants of Y. lipolyticagrown in different media were used after a cultivation time oftwo weeks. Triplicate polarization curves were recorded in parallelat 30 �C by increasing the current density in steps of 1.1 lA cm�2

every hour.Supernatant of YL4 grown in SCL medium or PPB medium was

used at pH 5.0, or in case of SCL medium also adjusted to pH 7.4with 0.1 M NaOH. For comparison, supernatant of the non-laccaseproducing strain Po1g grown in SCL medium was used at pH 5.0 oradjusted to pH 7.4 with 0.1 M NaOH. Furthermore, purified laccaseIIIb (Jolivalt et al., 2005) was added to supernatant of Po1g grownin SCL medium and used at pH 5.0.

As controls, plain SCL or PPB medium, as well as purified laccaseIIIb added to either 0.1 M sodium acetate buffer, to SCL medium, orto PPB medium were used.

Prior to the recording of polarization curves, the laccase in thesupernatant was allowed to adsorb to the electrode for 12 h underopen circuit conditions. Values for current density and potentialare reported as mean ± sample standard deviation, whereby thebars in the graphs of polarization curves correspond to minimumand maximum values. The pH increase in the cathode chamber be-fore and after the experiments is typically less than 0.5 pH units inthis type of setup.

234 S. Sané et al. / Bioresource Technology 158 (2014) 231–238

2.7. Recording of cathode polarization curves using E. coli

To record polarization curves at 30 �C using E. coli grown in LBmedium both, culture supernatants or complete E. coli cultureswere supplied at the cathode after a cultivation period of 20 h. Inminimal medium, the cultivation period was extended to 44 h be-fore using the supernatant at the cathode, since an increased culti-vation time resulted in a higher CueO activity.

Supernatant of E. coliCueO grown in LB medium at pH 7.4 or buf-fered with 0.1 M MES buffer at pH 5.6 was used without pH adjust-ment. In addition the supernatant of E. coliCueO grown in LBmedium at pH 7.4 or in minimal medium at pH 7.0 was adjustedto pH 5.0 with 10% acetic acid. Furthermore, also live cultures ofE. coliCueO growing in LB medium at pH 7.4 or buffered with0.1 M MES at pH 5.5 were supplied to the cathode.

As controls served plain selective LB medium (pH 5.0, as in Sec-tion 2.2) and supernatant of E. coliZ1 grown in LB medium and usedat pH 7.4 or adjusted to pH 5.0 with 10% acetic acid. Furthermore,live cultures of E. coliZ1 growing in LB medium buffered with 0.1 MMES at pH 5.5 were used as control.

Prior to recording polarization curves, CueO in the supernatantwas allowed to adsorb at the electrode for 12 h. In cases where thecathode compartment was filled with live E. coliCueO culture, CueOwas allowed to adsorb for 28 h before a load current was applied.

For E. coli supernatants and the corresponding controls, tripli-cate polarization curves were recorded in parallel by increasingthe current density in steps of 2.2 lA cm�2 every hour in the rangeof 2–55 lA cm�2, and steps of 5.6 lA cm�2 per hour in the range of55–170 lA cm�2.

To record duplicate polarization curves of cathodes suppliedwith live E. coliCueO or E. coliZ1 culture the current density was in-creased in steps of 5.6 lA cm�2 h�1 in the range of 5.6–225 lA cm�2.

Current densities and potentials are reported in the same wayas described in Section 2.6. In this setup, the pH increase in thecathode chamber before and after experiments starting at pH 5.0is typically up to 1.4 pH units. No increase in pH was observedwhen starting at pH 7.4.

3. Results and discussion

3.1. Laccase from Y. lipolytica grown in different culture media

3.1.1. Activity towards ABTSAs shown in Fig. 1a, already after two days of cultivation in PPB

or in SCL medium, laccase activity towards ABTS in the crude cul-ture supernatant of Y. lipolytica YL4 could be detected. The highestlaccase activity of 0.09 ± 0.02 Uml�1 was obtained in PPB mediumafter about 10 days of cultivation. In SCL medium, the highest lac-case activity recorded was 0.029 ± 0.004 Uml�1 after a cultivationperiod of 7 days. No laccase activity towards ABTS could be de-tected when YL4 was grown in YNB medium. Plain media, as wellas culture supernatant of Po1g, showed no activity towards ABTS.

The results clearly demonstrate that the choice of culture med-ium can have a significant influence on the laccase activity in cul-ture supernatants of Y. lipolytica YL4.

3.1.2. Electrochemical activity using supernatant of Y. lipolytica YL4grown in different media

Fig. 1b, shows the cathodic polarization curves obtained withdifferent supernatants of Y. lipolytica YL4 compared to the respec-tive culture medium and cultures of Y. lipolytica Po1g as control.

The crude culture supernatant of YL4 grown in SCL medium (pH5.0, laccase activity towards ABTS of 0.02 Uml�1) shows the highestelectro-catalytic activity towards oxygen reduction and the highest

open circuit potential (OCP) of 0.770 ± 0.006 V vs. NHE. As no elec-tro-catalytic activity was detected using the supernatant of thenon-laccase-producing Y. lipolytica strain Po1g in SCL medium(OCP: 0.444 ± 0.024 V vs. NHE), the electro-catalytic activity inYL4 supernatant in SCL medium can clearly be attributed to the lac-case enzyme. In contrast to using supernatant of YL4 grown in SCLmedium, the supernatant of YL4 grown in PPB medium (pH 5.0)shows an OCP of only 0.504 ± 0.032 V vs. NHE, despite an almostfour times higher laccase activity towards ABTS (i.e. 0.07 Uml�1).Furthermore, the polarization curve coincides with the controlcurve recorded in pure PPB medium, indicating that the superna-tant exhibits no additional electro-catalytic activity.

To test, whether components of the PPB medium itself or se-creted byproducts of Y. lipolytica YL4 grown in PPB medium sup-pressed the electro-catalytic activity of laccase at the cathode,the polarization curves of purified laccase IIIb using the same lac-case activity of 0.02 Uml�1 at the cathode either in PPB medium orin 0.1 M sodium actetate buffer were compared. The results (Sup-porting information S2) show that the electro-catalytic activity ofpurified laccase in PPB medium is only slightly higher than PPBmedium without enzyme, whereas the same laccase activity in0.1 M sodium acetate buffer shows an increased catalytic activitythat could be clearly attributed to the laccase. This indicates thatcomponents of the PPB medium inhibit the electro-catalytic activ-ity of laccase.

3.1.3. Factors influencing the electro-catalytic activity of Y. lipolyticaYL4 supernatant grown in SCL medium

To identify the factors limiting the electro-catalytic performanceof YL4 supernatant grown in SCL medium, the corresponding polar-ization curves (pH 5.0) with a laccase activity of 0.02 Uml�1 werecompared to polarization curves recorded with purified laccase IIIbof the same activity in either 0.1 M sodium acetate buffer (pH 5.0),plain SCL medium (pH 5.0), or supernatant of the non-laccase pro-ducing strain Po1g grown in SCL medium (pH 5.0) (Fig. 1c).

Although having the same ABTS oxidizing activity, purified lac-case added to sodium acetate buffer and plain SCL medium enabledmore than 4 and 5-fold higher current densities (11.3 ± 0.3 lA cm�2

and 13.5 ± 0.1 lA cm�2 at 0.644 V vs. NHE), respectively comparedto the supernatant of YL4 grown in SCL medium (2.7 ± 0.1 lA cm�2

at 0.644 V vs. NHE). On the other hand, the addition of purified lac-case IIIb (0.02 Uml�1) to the culture supernatant of the non-laccaseproducing strain Po1g resulted in a current density of2.8 ± 0.1 lA cm�2 at 0.644 V vs. NHE, which is in the same rangeas laccase-containing supernatant of YL4 grown in SCL medium.This suggests that the lower electro-catalytic activity of Y. lipolyticaYL4 supernatant grown in SCL medium is not caused by mediumcomponents, but rather by secreted byproducts of the yeast. To as-sess whether a shorter cultivation period results in the secretion ofless inhibitory by-products, polarization curves were also recordedusing supernatant of YL4 grown in SCL medium with a cultivationtime of only four days instead of 14 days, with a laccase activityof 0.03 Uml�1. No improvement of cathode performance was ob-served (Supporting information S3).

In one case, after four days of cultivating Y. lipolytica YL4 in SCLmedium a 2-fold higher enzyme activity of 0.06 Uml�1 comparedto the average activity of 0.03 Uml�1 in this medium was obtained.The reason for this higher activity is not known, but presumablysmall deviations in the culture conditions could be responsiblefor the increase in laccase activity. To see whether a higher enzymeactivity in the supernatant leads to a higher current density, thissupernatant was also used to record polarization curves. As shownin Fig. 1d, the higher laccase activity of 0.06 Uml�1 in supernatantat the cathode resulted in a more than 2-fold higher current den-sity of 6.7 ± 0.4 lA cm�2 (at 0.644 V vs. NHE) compared to2.7 ± 0.1 lA cm�2 (at 0.644 V vs. NHE, 0.02 Uml�1 laccase activity).

Fig. 1. (a) Laccase activity in crude culture supernatant of Yarrowia lipolytica YL4 grown in PPB medium (green), SCL medium (red) and YNB medium (yellow). (b) Comparisonof polarization curves using crude culture supernatant of Y. lipolytica YL4 grown in SCL medium with a laccase activity of 0.02 Uml�1 (red) and grown in PPB medium with alaccase activity of 0.07 Uml�1 (green). The polarization curves for PPB medium (green dotted line) and SCL medium (red dotted line) as well as supernatant of the non-laccaseproducing Y. lipolytica strain Po1g in SCL medium (blue) are shown as controls. (c) Comparison of polarization curves for laccase-containing supernatant of YL4 grown in SCLmedium and purified laccase with the same laccase activity towards ABTS (0.02 Uml�1) supplied either to 0.1 M sodium acetate buffer, to SCL medium or to the supernatant ofPo1g grown in SCL medium. (d) Comparison of polarization curves using the supernatant of YL4 grown in SCL medium with different laccase activities and Po1g grown in SCLmedium without laccase activity as a control. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

S. Sané et al. / Bioresource Technology 158 (2014) 231–238 235

This clearly demonstrates that the electro-catalytic activity in thesupernatant of YL4 grown in SCL medium can be further increasedby improving the laccase production. Further research shouldtherefore concentrate on increasing the laccase production of Y.lipolytica YL4 or screening the use of alternative microorganismswith a higher laccase production and less inhibitory byproducts.

The supernatant of Y. lipolytica YL4 grown in SCL medium ad-justed to a pH value of 7.4 did not show an electro-catalytic activityfor oxygen reduction attributable to laccase (polarization curves inthe same range as when using supernatant at pH 7.4 of the non-laccase producing Y. lipolytica Po1g grown in SCL medium, see Sup-porting information S4). This is in agreement with a previous re-port, which show almost no laccase IIIB activity in the presenceof ABTS as redox indicator above a pH value of 6 (Madzak et al.,2006) and underlines the need to sustain a pH around a value ofpH 5 for enzymatic cathodes based on laccase.

3.2. Copper efflux oxidase from E. coliCueO grown in different culturemedia

3.2.1. Activity towards ABTSIn Fig. 2a the time course of copper efflux oxidase (CueO) activ-

ity towards ABTS (measured at pH 5.0 and 30 �C) in supernatants ofE. coliCueO cultivated in different media at 37 �C is compared. As

shown above for the laccase of Y. lipolytica YL4, the obtained CueOactivity towards ABTS also dependents on the choice of culturemedium. The highest CueO activity recorded was(0.034 ± 0.003) � 10�3 Uml�1 after 19 h of cultivating E. coliCueO inLB medium at pH 7.4. However, already on the second day of cul-tivation, the CueO activity decreased by more than 80%. A similarlyhigh CueO activity was recorded in the supernatant of E. coliCueO

grown in minimal medium (pH 7.0), where a maximum value of(0.030 ± 0.001) � 10�3 Uml�1 was measured after 3 days ofcultivation.

No CueO activity was detectable in the supernatant of E. coliCueO

grown in SCCueO medium. Accordingly, these cultivation condi-tions were omitted from further analyses. As expected, the med-ium itself as well as the supernatant of E. coliZ1, did not showABTS oxidizing activity.

3.2.2. Electrochemical activity using supernatant of E. coliCueO atdifferent pH values

The polarization curves using supernatant of E. coliCueO growneither in LB medium or minimal medium are compared in Fig. 2b.Here, the supernatant of E. coliCueO grown in LB medium (pH 7.4)and adjusted to pH 5.0 (CueO activity of 0.034 � 10�3 Uml�1)resulted in an OCP of 0.562 ± 0.013 V vs. NHE, and a currentdensity of 119 ± 23 lA cm�2 at 0.400 V vs. NHE. Remarkably,

Fig. 2. (a) Time course of CueO activity in the supernatant of E. coliCueO cultivated in different media and measured towards ABTS (pH 5.0, 30 �C). (b) Polarization curves forculture supernatants of E. coliCueO grown in minimal medium and LB medium. In the caption of the curves, the first pH value indicates the pH during cultivation, whereas thesecond pH value indicates the pH of the supernatant used at the cathode.

236 S. Sané et al. / Bioresource Technology 158 (2014) 231–238

the use of crude culture supernatant of E. coliCueO grown in min-imal medium yielded no electro-catalytic activity at the cathode.Presumably, components in the minimal medium or secretedbyproducts of E. coliCueO cultivated under these conditions havea negative effect on the electro-catalytic activity of CueO.

In Fig. 3, supernatants at pH 7.4 or adjusted to pH 5.0 of E.coliCueO grown in LB medium are shown, and compared to theirrespective controls. The E. coliCueO supernatant when used at pH7.4 yielded an 89 mV lower OCP of 0.473 ± 0.007 V vs. NHE com-pared to supernatant at pH 5.0. This potential shift is in good agree-ment with the expected pH-dependency of the electrode potentialas described by the Nernst equation. However, with increasing cur-rent densities, large activation losses occur, which indicate a lowerelectrochemical activity of CueO at pH 7.4 as compared to pH 5.0.

During the experiment, an increase in pH from initially 5.0 up toa final value of 6.4 was observed. This pH drift is a well-knownphenomenon in biofuel cells caused by insufficient bufferingcapacity and proton transport from the anode to the cathode(Rozendal et al., 2006). It occurs in particular at high operating cur-rents and could explain that above approx. 100 lA cm�2, the polar-ization curves of supernatant at pH 5.0 start to coincide with

Fig. 3. Comparison of polarization curves using culture supernatant at pH 5.0 andpH 7.4 of E. coliCueO and E. coliZ1. In the caption of the curves the first pH valueindicates the pH during cultivation, whereas the second pH value indicates the pHof the supernatant used at the cathode.

polarization curves of supernatant at pH 7.4. It is also likely thatlimited oxygen supply to the electrode or the limited turnovernumber of the adsorbed enzymes contribute to the increasedpolarization above approx. 100 lA cm�2. No pH increase was re-corded for supernatants at cathodes with an initial pH of 7.4, pre-sumably due to the higher buffering capacity of that system.

As expected, significantly lower electro-catalytic activities wereobserved in control experiments using plain LB medium (pH 5.0,Supporting information S5) or supernatants at pH 5.0 or pH 7.4of E. coliZ1 grown in LB medium (Fig. 3).

3.2.3. Polarization curves of cathodes supplied with complete E. colicultures

Given that, in contrast to Y. lipolytica YL4, the use of E. coliCueO

offers the advantage of higher current densities and significantelectro-catalytic activity also at neutral pH, it was further investi-gated whether it is possible to use intact cells secreting CueO in thecathode compartment, instead of sterile filtered supernatant.

As shown before, the use of the sterile-filtered supernatant ofE. coliCueO at pH 7.4 resulted in an OCP of 0.473 ± 0.007 V vs. NHE(Fig. 3). However, when directly supplying cultures of E. coliCueO

at a pH of 7.4 to the cathode, a maximum OCP of only 0.266 V vs.NHE was recorded after 12 h, and decreased afterwards (observa-tion period one week). Due to the low OCP, no polarization curveswere recorded using cultures under these conditions. The low OCPof E. coliCueO cultures growing at a pH of 7.4 can be explained byoxygen depletion in the cathode compartment due to the organ-ism’s continuous metabolism. This explanation is supported bythe polarization curves of cathodes supplied with live E. coli cul-tures at the lower pH of 5.5 (30 �C) shown in Fig. 4. At pH 5.5,the growth rate of E. coli is lower than at pH 7.4 (results of pHdependency on growth rate in Supporting information S1), andaccordingly, less oxygen is consumed by the growing cultures. Asa consequence, complete cultures of E. coliCueO growing at pH 5.5in the cathode compartment yielded an OCP of 0.523 ± 0.005 Vvs. NHE. This is similar to the OCP of sterile filtered supernatantof E. coliCueO in LB medium buffered with 0.1 M MES at pH 5.6(Fig. 4).

With increasing current density, the live E. coliCueO cultureshows initially less polarization than the sterile-filtered superna-tant. This suggests a higher activity due to the continuous supplyof CueO by the live E. coliCueO cultures. However above a currentdensity of approx. 80 lA cm�2 both curves do not differ signifi-cantly. This suggests that in the course of the experiment the CueO

Fig. 4. Comparison of cathode polarization curves using supernatant at pH 5.0 orpH 5.6 of E. coliCueO grown in LB medium at pH 7.4 (red) and pH 5.6 (green),respectively, to using live cultures at the cathode at pH 5.5 of E. coliCueO (purple) orE. coliZ1 (orange) both grown in 0.1 M MES buffered LB medium (pH 5.5). In thecaption of the curves, the first pH value indicates the pH during cultivation, whereasthe second pH value indicates the pH of the supernatant used at the cathode. (Forinterpretation of the references to colour in this figure legend, the reader is referredto the web version of this article.)

S. Sané et al. / Bioresource Technology 158 (2014) 231–238 237

activity in the E. coli culture decreases, possibly due to secretedbyproducts or because of an increase in pH value. However, CueOactivities were not analyzed in order to avoid disturbing the exper-iment by continuously taking samples, and further work is re-quired for clarification. No catalytic activity attributed to CueOwas recorded using live cultures of the non-CueO producing strainE. coliZ1 growing in LB medium (pH 5.6). This indicates again, thatthe catalytic activity recorded for E. coliCueO can be attributed to thesecreted enzyme CueO at the cathode.

In summary, the results demonstrate the possibility to useE. coli growing directly in the cathode compartment, which wouldsimplify the reactor setup to supply CueO at the cathode, in partic-ular in miniaturized integrated systems. An important parameterin this context is the oxygen depletion occurring through the cul-ture’s metabolism. This will require further optimization regardingtemperature or nutrient availability to decrease the growth ratewhile still obtaining a sufficient CueO production. Furthermore,for continuous cultivation experiments over extended periods oftime, the integration of the CueO gene into the chromosome willbe of advantage.

3.3. Evaluation of the use of supernatants for practical application

For a practical application of crude culture supernatants at abiofuel cell cathode, both the electro-catalytic activities, as wellas the operation conditions have to be considered.

In terms of electrochemical activity at pH 5.0, the OCP recordedfor supernatants of Y. lipolytica YL4 grown in SCL medium is about200 mV higher compared to the supernatant of E. coliCueO grown inLB medium. However, when comparing the current densities at0.400 V vs. NHE, a 10-fold higher current density can be achievedwith the E. coliCueO supernatant compared to supernatant of Y.lipolytica YL4. For applications demanding higher current densities,the E. coliCueO supernatant is thus clearly preferable.

The supernatant of E. coliCueO grown in LB medium also shows asignificant electro-catalytic activity at pH 7.4 and exhibits approx.200 mV higher electrode potentials over the investigated currentdensity range than the control experiment. In contrast, thesupernatant of Y. lipolytica YL4 grown in SCL medium exhibits no

electro-catalytic activity attributable to laccase when the pH wasadjusted to 7.4. The use of E. coliCueO can thus again be regardedadvantageous, since it offers a larger degree of freedom in cathodeoperation conditions and promises better compatibility to typicalmicrobial fuel cell anodes operated at around a pH value of 7.

Compared to the use of laccase-containing T. versicolor superna-tant at a cathode at pH 5.0 (Sané et al., 2013), the electro-catalyticperformance of both supernatants investigated in the present workis clearly lower. Whereas E. coliCueO enables similarly high currentdensities as T. versicolor supernatant, the electrode potential is byapprox. 300 mV lower due to the different redox potentials of CueOand laccase. On the other hand, the supernatant of Y. lipolytica YL4shows a similar OCP compared to T. versicolor but yields much low-er current densities due to limited enzyme activity. Nevertheless,the planktonic nature of the microorganisms investigated hereopens more options in terms of continuous enzyme supply at acathode, e.g. in a continuous-flow system. In this respect, filamen-tous fungi are more difficult to handle since they grow in agglom-erates of fungal cells that loose viability over time (Kavanagh,2011).

Regarding further optimization, it should be considered thatpurified CueO shows an onset potential of oxygen reduction atpH 5.0 of around 0.649 V vs. NHE (0.450 V vs. Ag/AgCl (Miuraet al., 2009)). This potential is about 100 mV higher than the OCPrecorded in the present setup using the supernatant of E. coliCueO

at pH 5.0. To decrease the overpotential of CueO, the investigationof other electrode materials or culture media is thus a promisingroute. Furthermore, the results suggest that the electro-catalyticperformance of the supernatants of E. coliCueO at pH 7.4 or ofY. lipolytica YL4 at pH 5.0 can be improved by achieving a higherenzyme activity e.g. with optimized cultivation parameters.

4. Conclusions

The results show that without further purification, crude cul-ture supernatant from the planktonic microorganisms Y. lipolyticaYL4 and E. coliCueO can be used to supply the enzymes laccase andcopper efflux oxidase to a biofuel cell cathode, respectively. In con-trast to Y. lipolytica, the supernatant of E. coliCueO yields 10-foldhigher current densities at 0.400 V vs. NHE, and exhibits an elec-tro-catalytic activity also at pH 7.4. This opens the opportunityfor simpler and less expensive enzymatic cathodes, compatiblewith neutral pH microbial anodes.

Acknowledgements

We would like to thank Olivera Francetic for generous supportregarding the construction of a CueO exporting strain. Financialsupport by the German Research Foundation (DFG) through thePhD program ‘‘Micro Energy Harvesting’’ (GRK 1322) and by theState Graduate Funding is gratefully acknowledged. Furthermore,the authors want to thank Elena Kipf for designing the adaptedtwo-chamber MFC setup used in this work.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2014.02.038.

References

Barrière, F., Kavanagh, P., Leech, D., 2006. A laccase–glucose oxidase biofuel cellprototype operating in a physiological buffer. Electrochim. Acta 51, 5187–5192.

Bullen, R.A., Arnot, T.C., Lakeman, J.B., Walsh, F.C., 2006. Biofuel cells and theirdevelopment. Biosens. Bioelectron. 21, 2015–2045.

238 S. Sané et al. / Bioresource Technology 158 (2014) 231–238

Christenson, A., Dimcheva, N., Ferapontova, E.E., Gorton, L., Ruzgas, T., Stoica, L.,Shleev, S., Yaropolov, A.L., Haltrich, D., Thorneley, R.N.F., Aust, S.D., 2004. Directelectron transfer between ligninolytic redox enzymes and electrodes.Electronanalysis 16, 1074–1092.

Cinquin, P., Gondran, C., Giroud, F., Mazabrard, S., Pellissier, A., Boucher, F., Alcaraz,J.P., Gorgy, K., Lenouvel, F., Mathé, S., Porcu, P., Cosnier, S., 2010. A glucoseBioFuel cell implanted in rats. PLoS One 5, e10476.

Francetic, O., Pugsley, A.P., 2005. Towards the identification of type II secretionsignals in a nonacylated variant of pullulanase from Klebsiella oxytoca. J.Bacteriol. 187, 7045–7055.

Gibson, D.G., Young, L., Chuang, R.Y., Venter, J.C., Hutchison, C.A., Smith, H.O., 2009.Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat.Methods 6, 343–345.

Grass, G., Rensing, C., 2001. CueO Is a multi-copper oxidase that confers coppertolerance in Escherichia coli. Biochem. Biophys. Res. Commun. 286, 902–908.

Harnisch, F., Schröder, U., 2010. From MFC to MXC: chemical and biologicalcathodes and their potential for microbial bioelectrochemical systems. Chem.Soc. Rev. 39, 4433–4448.

Hussein, L., Urban, G., Krüger, M., 2011. Fabrication and characterization ofbuckypaper-based nanostructured electrodes as a novel material for biofuelcell applications. Phys. Chem. Chem. Phys. 13, 5831–5839.

Ieropoulos, I.A., Greenman, J., Melhuish, C., Horsfield, I., 2012. Microbial fuel cells forrobotics: energy autonomy through artificial symbiosis. Chemsuschem 5, 1020–1026.

Ivnitski, D., Atanassov, P., 2007. Electrochemical studies of intramolecular electrontransfer in laccase from Trametes versicolor. Electronanalysis 19, 2307–2313.

Jolivalt, C., Madzak, C., Brault, A., Caminade, E., Malosse, C., Mougin, C., 2005.Expression of laccase IIIb from the white-rot fungus Trametes versicolor in theyeast Yarrowia lipolytica for environmental applications. Appl. Microbiol.Biotechnol. 66, 450–456.

Kavanagh, K., 2011. Fungi: Biology and Applications. John Wiley & Sons, Chichester,UK.

Kerzenmacher, S., Ducrée, J., Zengerle, R., von Stetten, F., 2008. Energy harvesting byimplantable abiotically catalyzed glucose fuel cells. J. Power Sources 182, 1–17.

Kloke, A., Rubenwolf, S., Bücking, C., Gescher, J., Kerzenmacher, S., Zengerle, R., vonStetten, F., 2010. A versatile miniature bioreactor and its application tobioelectrochemistry. Biosens. Bioelectron. 25, 2559–2565.

Lapinsonniére, L., Picot, M., Barriére, F., 2012. Enzymatic versus microbial bio-catalyzed electrodes in bio-electrochemical systems. Chemsuschem 5, 995–1005.

Liu, X.W., Li, W.W., Yu, H.Q., 2014. Cathodic catalysts in bioelectrochemical systemsfor energy recovery from wastewater. Chem. Soc. Rev.. http://dx.doi.org/10.1039/c3cs60130g.

Logan, B.E., 2005. Simultaneous wastewater treatment and biological electricitygeneration. Water Sci. Technol. 52, 31–37.

Logan, B.E., 2008. Microbial Fuel Cells. John Wiley & Sons Inc, Hoboken, New Jersey.Lutz, R., Bujard, H., 1997. Independent and tight regulation of transcriptional units

in Escherichia coli via the LacR/O, the TetR/O and AraC/I-1-I-2 regulatoryelements. Nucleic Acids Res. 25, 1203–1210.

Madzak, C., Gaillardin, C., Beckerich, J.M., 2004. Heterologous protein expressionand secretion in the non-conventional yeast Yarrowia lipolytica: a review. J.Biotechnol. 109, 63–81.

Madzak, C., Mimmi, M.C., Caminade, E., Brault, A., Baumberger, S., Briozzo, P.,Mougin, C., Jolivalt, C., 2006. Shifting the optimal pH of activity for a laccasefrom the fungus Trametes versicolor by structure-based mutagenesis. ProteinEng. Des. Sel. 19, 77–84.

Mano, N., Fernandez, J.L., Kim, Y., Shin, W., Bard, A.J., Heller, A., 2003. Oxygen iselectroreduced to water on a ‘‘wired’’ enzyme electrode at a lesser overpotentialthan on platinum. J. Am. Chem. Soc. 125, 15290–15291.

Miura, Y., Tsujimura, S., Kamitaka, Y., Kurose, S., Kataoka, K., Sakurai, T., Kano, K.,2007. Bioelectrocatalytic reduction of O2 catalyzed by CueO from Escherichiacoli adsorbed on a highly oriented pyrolytic graphite electrode. Chem. Lett. 36,132–133.

Miura, Y., Tsujimura, S., Kurose, S., Kamitaka, Y., Kataoka, K., Sakurai, T., Kano, K.,2009. Direct electrochemistry of CueO and its mutants at residues to and neartype I Cu for oxygenreducing biocathode. Fuel Cells 9, 70–78.

Osman, M.H., Shah, A.A., Walsh, F.C., 2010. Recent progress and continuingchallenges in bio-fuel cells. Part II: Microbial. Biosens. Bioelectron. 26,953–963.

Osman, M.H., Shah, A.A., Walsh, F.C., 2011. Recent progress and continuingchallenges in bio-fuel cells. Part I: Enzymatic cells. Biosens. Bioelectron. 26,3087–3102.

Rosenbaum, M., Aulenta, F., Villano, M., Angenent, L.T., 2011. Cathodes as electrondonors for microbial metabolism: which extracellular electron transfermechanisms are involved? Bioresour. Technol. 102, 324–333.

Rozendal, R.A., Hamelers, H.V.M., Buisman, C.J.N., 2006. Effects of membrane cationtransport on pH and microbial fuel cell performance. Environ. Sci. Technol. 40,5206–5211.

Rubenwolf, S., Kerzenmacher, S., Zengerle, R., von Stetten, F., 2011. Strategiesto extend the lifetime of bioelectrochemical enzyme electrodes forbiosensing and biofuel cell applications. Appl. Microbiol. Biotechnol. 89,1315–1322.

Sané, S., Jolivalt, C., Mittler, G., Nielsen, P.J., Rubenwolf, S., Zengerle, R.,Kerzenmacher, S., 2013. Overcoming bottlenecks of enzymatic biofuel cellcathodes: crude fungal culture supernatant can help to extend lifetime andreduce cost. Chemsuschem 6, 1209–1215.

Schaetzle, O., Barriere, F., Schroder, U., 2009. An improved microbial fuel cell withlaccase as the oxygen reduction catalyst. Energy Environ. Sci. 2, 96–99.

Schuetz, B., Schicklberger, M., Kuermann, J., Spormann, A.M., Gescher, J., 2009.Periplasmic electron transfer via the c-type cytochromes MtrA and FccA ofShewanella oneidensis MR-1. Appl. Environ. Microbiol. 75, 7789–7796.

Soukharev, V., Mano, N., Heller, A., 2004. A four-electron O-2-electroreductionbiocatalyst superior to platinum and a biofuel cell operating at 0.88 V. J. Am.Chem. Soc. 126, 8368–8369.

Tsujimura, S., Miura, Y., Kano, K., 2008. CueO-immobilized porous carbon electrodeexhibiting improved performance of electrochemical reduction of dioxygen towater. Electrochim. Acta 53, 5716–5720.