Applied Energy 120 (2014) 85–94

Contents lists available at ScienceDirect

Applied Energy

journal homepage: www.elsevier .com/locate /apenergy

Electrochemical reduction of carbon dioxide at low overpotentialon a polyaniline/Cu2O nanocomposite based electrode

http://dx.doi.org/10.1016/j.apenergy.2014.01.0220306-2619/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Address: Climate Change Technology Research Division,Korea Institute of Energy Research, 102 Gajeong-ro, Yuseong-gu, Daejeon 305-343,Republic of Korea. Tel.: +82 42 860 3623; fax: +82 42 860 3134.

E-mail address: jeongsk@kier.re.kr (S.K. Jeong).

Andrews Nirmala Grace a,b, Song Yi Choi a, Mari Vinoba a, Margandan Bhagiyalakshmi c, Dae Hyun Chu a,Yeoil Yoon a, Sung Chan Nam a, Soon Kwan Jeong a,⇑a Korea Institute of Energy Research, Daejeon 305-343, Republic of Koreab Centre for Nanotechnology Research, VIT University, Vellore 632014, Indiac Department of Chemistry, Central University of Kerala, Kasaragod 671314, India

h i g h l i g h t s

� Potentiostatic and current depositiontechnique was used to fabricateelectrodes.� Cu2O dispersed polyaniline electrodes

was fabricated.� CO2 was reduced by potential

electrolysis method at differentpotentials.� A faradaic efficiency of 30.4% and

63.0% HCOOH & CH3COOH wasobserved at �0.3 V.

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 9 September 2013Received in revised form 25 December 2013Accepted 7 January 2014Available online 14 February 2014

Keywords:Electrochemical CO2 reductionPolyanilineCu2OH-Type cellMembrane cellFaradaic efficiency

a b s t r a c t

The electrochemical reduction of CO2 using Cu2O nanoparticle decorated polyaniline matrix (PANI/Cu2O)in 0.1 M tetrabutylammonium perchlorate (TBAP) and methanol electrolyte was investigated under ambi-ent conditions. The experiment was carried out in a divided H-type two-compartment cell with a Nafionmembrane as diaphragm separating the cathodic and anodic compartments. The catalyst was synthesizedelectrochemically as a thin film by using cyclic voltammetry and constant current mode depositiontechnique. The as-fabricated electrode was analyzed with various techniques to probe the nature andcomposition of the nanoparticles deposited onto the polyaniline matrix, which confirmed the presenceof well-defined Cu (I) species in the film. The reduction of CO2 was carried out at various polarizationpotentials; the main products were formic and acetic acid with faradaic efficiencies of 30.4% and 63.0%at a polarization potential of �0.3 V vs. SCE (sat. KCl). A possible reduction pathway is through the forma-tion of Had atoms and subsequent transfer to CO2 through the polymer film to form the products. An appre-ciable efficiency was achieved in the formation of formic acid and acetic acid with the developed catalyst.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Among the contemporary energy objectives, the reduction ofcarbon dioxide emission is a major issue of immediate concern.

86 A.N. Grace et al. / Applied Energy 120 (2014) 85–94

Anthropogenic CO2 in the atmosphere is claimed to be the majorcontributor to the greenhouse effect, and the immediate conse-quence is the phenomenon of global warming [1–3]. The famousKeeling curve shows an annual rise of 1.9 ppm per year [4]. TheIntergovernmental Panel on Climate Change (IPCC) has predictedthat levels of CO2 could reach 730–1020 ppm by 2100, whichwould result in a mean global temperature rise of 1.4–1.9 K andocean acidification by 0.14–0.35 pH units [5]. The adverse impactsof global warming include rising sea levels, disappearance of someislands, extreme weather, and climate change [5,6]. Therefore,reducing CO2 emissions into the atmosphere has become a criticalissue. In addition to efforts to capture and store CO2 in the hope ofreducing the greenhouse effect, the conversion of CO2 to fuels andchemical feedstock is an attractive proposition that could providean alternative solution to both the current energy crisis and cli-mate issues [7–9]. Various methods have been developed to reduceCO2 emissions, such as chemical, thermochemical, photochemical,electrochemical, and biochemical techniques [10–12]. Of thesemethods, electrochemical reduction is simple and can be per-formed under ambient conditions, and is of particular interestsince it could both mitigate greenhouse gas emission and useCO2 as a carbon source to produce a variety of fuels such as formicacid and methanol [12–16]. The intermittency of renewable energysources requires practical energy storage solutions. This can beaccomplished by finding an efficient way to store energy in theform of chemical fuels, which would be particularly appealing ifcombined with CO2 capture [17,18]. Thus the electrochemicalreduction of CO2 using renewable sources of electricity is an alter-native approach to the production of fuels, akin to photosynthesis,which could potentially reduce the dependence on fossil fuels andmitigate CO2 emissions into the atmosphere [19]. The electrocata-lytic reduction of CO2 to fuels is thus a critical goal that would pos-itively impact the global carbon balance [20–22]. However, CO2 isan extremely stable molecule, so the conversion of CO2 to a usefulfuel on the same scale as its current production is beyond our pres-ent scientific and technological abilities [7].

In addition, the direct electrochemical reduction of CO2 needs atleast 1–2 V of overpotential. The conversions of CO2 to formic acid,methane, ethane, ethylene, propylene, oxalic acid, methanol, andethanol have been reported under various electrocatalytic condi-tions and in various solvents [23,24]. Of the useful chemicals intowhich CO2 can be electrochemically converted, formic acid appearsto offer the most promise for the practical development of techni-cal and economical viable processes. Traditional industries use for-mic acid in silage preservation, as an additive in animal feeds, intextile finishing, and as a chemical intermediate [25]. Moreover,formic acid offers easy transport and storage and is one of the mostpromising candidate fuels for low-temperature fuel cells [26].However, the manufacture of formic acid is relatively expensiveand has negative environmental impacts [27,28]. Thus moreattention has been paid to the electrochemical reduction of CO2

to formic acid in recent years [29–31]. The electrochemical reduc-tion of CO2 to formic acid is described by the following equation:

CO2 þ 2Hþ þ e� ! HCOOH ð1Þ

The theoretical potential for the electrochemical reduction ofCO2 to formic acid under standard conditions is �0.854 V (SCE,sat. KCl) [30]. As a result of the high overpotential, a voltage ofmore than 4 V is required to decrease the cathode potential ofany common metal electrode to below �1.545 V (SCE, sat. KCl)for effective CO2 reduction. Thus the development of new catalystsis crucial if we are to reduce this overpotential and create an effi-cient electroreduction process.

Many investigators have studied the electrochemical reductionof CO2 using various metal electrodes in organic solvents, giventhat organic aprotic solvents dissolve much more CO2 than water

[32–34]. Methanol is a better solvent of CO2 than water and thesolubility of CO2 in methanol is approximately 5 times that inwater, at ambient temperature, and 8–15 times that in water be-low 0 �C [35–37]. Further, if water is used as a solvent for thereduction of CO2, hydrogen evolution decreases the current effi-ciency significantly. It has also been shown that the faradaic effi-ciencies of products from CO2 in methanol are better than thoseobtained in water [38]. Many authors have reported the reductionof CO2 in methanol as the medium [39–43].

Mediated electrocatalysis represents the catalytic effect whichassumes an interaction between mediator-electrode and the sub-strate molecules i.e. CO2. The advantages of this method are a de-crease in the electrochemical potential at which electrochemicalreductions occur and an increase in the selectivity of the process.In general, metal electrodes need a high negative potential of lessthan �1.7 V (SCE, sat. KCl) for the electroreduction of CO2. To de-crease this potential, new electrode materials are required. In thisconnection, conducting polymers such as polyaniline, polypyrrole,and polythiophene have gained recognition over the last two dec-ades [43]. The reasons for this interest are their peculiar propertiesand the applications of these materials in various niche areas. Apolyaniline electrode laminated with Prussian blue (consisting ofa mediating inorganic conductor and polymer) has been used forCO2 reduction under high pressure by Ogura et al. [44]. With thismodified electrode, CO2 was reduced at �0.845 V (SCE, sat. KCl)and the reaction products were lactic acid, acetic acid, formic acid,methanol, and ethanol. In another study [45,46], bulk polyanilinewas used as the electrode material in the reduction of CO2 toorganics such as formaldehyde, formic acid, and acetic acid at�0.4 V (SCE, sat. KCl). Koleli et al. used polypyrrole electrodes to re-duce CO2 in a methanol/LiClO4 system, which produced formic andacetic acid at an electrolysis potential of �0.445 V (SCE, sat. KCl)[47,48]. All these reports suggest that the use of conductivepolymers enables the electrochemical reduction of CO2 at loweroverpotentials and produces formic acid with a good faradaic effi-ciency. Cu2O is yet another fascinating material, a p-type semicon-ductor with a direct band gap of 2.14 eV. Cu2O has receivedextensive attention in various applications because of its low costand robust chemical nature [49]. As this material is stable in mostorganic and aqueous solutions, it is a promising candidate for theelectrochemical reduction of CO2.

This study tested the use of a polyaniline matrix decorated withCu2O nanoparticles as electrode for the electrochemical reductionof CO2 to valuable fuels. We aimed to determine how to maintaina high current efficiency in this system and the concentration offormic acid that could be build up in the catholyte. A simple andnovel electrochemical process was used for the fabrication of theworking electrode. For CO2 electroreduction, a special H-type reac-tor was designed, and fabricated with Nafion membrane as aseparator.

2. Experimental

2.1. Reactor construction

The electroreduction of CO2 was carried out in an H-type two-compartment cell with an anode and a cathode. The two compart-ments were separated by a proton exchange membrane (Nafion117, 0.177 mm thickness) as the diaphragm. The cell was con-structed of glass, with two orifices and gas purging inlet in thecathode compartment, an orifice and a gas inlet port in the anodecompartment, and O-rings to hold the electrodes and membrane.The working and reference electrodes were housed in the cathodecompartment and the counter electrode was mounted in the anodecompartment. A schematic diagram of the electroreduction set-up

A.N. Grace et al. / Applied Energy 120 (2014) 85–94 87

is shown in Scheme 1. The apparatus and experimental conditionsfor the electrochemical reduction of CO2 are specified in Table 1.

2.2. CO2 electroreduction test and calculations

CO2 electroreduction was performed in 0.1 M TBAP/CH3OHsolution in the designed H-type cell and as a proton source 0.1 MH2SO4 was added. The role of acid is to only enhance the currentdensity, which doesn’t affect the product formation. The catholytewas saturated with pure CO2 for 1 h at 100 sccm and the electro-lyte was continuously stirred with a magnetic bar during electrol-ysis. The as-fabricated PANI/Cu2O and SCE (sat. KCl) electrodeswere placed in the cathode compartment and the platinized Ptwire was placed in the anode compartment. Cyclic voltammetryand the constant potential technique were performed to determinethe current density and faradaic efficiency. The electrolyte wasprepared from analytical grade methanol. The CO2 saturated solu-tion was then reduced electrochemically at cathodic polarization inthe range �0.1 to �0.4 V vs. SCE. The formed liquid products thatdissolved in the catholyte were analyzed with ion chromatogra-phy. Only the liquid product was analyzed with an emphasis onthe formation of formic acid. The dilution factor was taken into ac-count for all calculations. Further to identify if the product wasformed from CO2 and methanol, labeled experiments were doneusing 13CO2 and 13CH3OH and the products were analyzed withGC-MS and H NMR. All the experiments were done thrice andthe graph was plotted with error bars representing the standard

Scheme 1. A schematic of the H-type membrane based electrochemical cell.

Table 1Instrumentation and experimental conditions.

Potentiostat Biologic SP 240

Potential sweep �0.4 to 0.2 VWorking electrode 30% Pt loaded carbon paperCounter electrode Pt wire (30 � 20 mm2)Reference electrode SCE (sat. KCl)ElectrolyteCatholyte 1.1 M TBAP in methanolAnolyte 0.1 M TBAP in methanolCarbon dioxide 99.999% purePotential �0.4 to 0.2 V vs. SCE sat. KClTemperature 293 ± 5 KLiquid product analysis Ion chromatography (Metrohm 881 compact with

conductivity detector)Metrosep organic acids column with 1 mMHClO4/5% acetone as mobile phase and10 mM LiCl suppressor

deviation of three repetitive experiments. The faradaic efficienciesof the products were calculated from the number of electrons con-sumed in the electroreduction process by using the formula:

F � E½g�ð%Þ ¼ mnFR t

0 Idtð2Þ

where m is the number of moles of product harvested, n is the num-ber of electrons required for the formation of the product, F is theFaraday constant (96,485 C/mol of electrons), and I is the circuitcurrent.

2.3. Electrochemical deposition of polyaniline

The polyaniline film was grown on the working electrodepotentiostatically between �0.2 and 0.8 V vs. SCE in a solutionof 0.5 M H2SO4 at a scan rate of at 50 mV/s for 25 continuous cy-cles [50]. The corresponding cyclic voltammetric curve is shownin Fig. 1. As can be seen in the figure, the electrooxidation of ani-line commences at +0.1 V in the first cycle, which produces thefirst layer of polyaniline. Subsequent cycles result in the furthergrowth of the polymer film, as evident from the increase in theredox current. The cyclic voltammograms obtained during theelectropolymerization of polyaniline contains well-defined redoxpeaks corresponding to a series of redox transitions: the oxidationof the fully reduced insulating form (leucoemeraldine) to its rad-ical cation (polaron, emeraldine), followed by the oxidation of thedegradation products and/or intermediate species, and finally thetransition from the delocalized polaronic state to a localizedbipolaron or quinoid form (pernigraniline) [51]. The peaks labeledas x1 and y1 correspond to the oxidation of the polymer fromleucoemeraldine to the emeraldine redox states and from theemeraldine to the pernigraniline states. The peaks marked as x2

and y2 correspond to the reduction of the polymer from thepernigraniline to the emeraldine form and further to the leuco-emeraldine states, respectively. The whole mechanism is givenas Scheme S2. The mass of monomers electropolymerized ontothe substrate can be roughly estimated from the total faradaiccharges consumed in the electropolymerization and by assumingan average of 2.5 electrons per aniline monomer in the emeral-dine state [52]:

m ¼ CMm

2:5Fð3Þ

where m is the mass of PANI polymerized onto the electrode ingrams, C is the total faradaic charge consumed in the electropoly-merization, Mm is the molecular mass of the aniline monomer

Fig. 1. Cyclic voltammetry of the electropolymerization process in 0.1 Maniline + 0.5 M H2SO4; scan rate = 50 mV/s.

Fig. 2. X-ray diffraction spectrum of PANI/Cu2O film.

88 A.N. Grace et al. / Applied Energy 120 (2014) 85–94

(93.13 g/mol), and F is the Faraday constant (96485 C/mol). Hence,a 3.14 lg polyaniline film was obtained. The fabricated polyanilinefilm was then removed from solution, dried, and used as a probe forthe electrochemical deposition of Cu2O nanoparticles. The electro-deposition was performed with the constant current mode tech-nique. A constant current of �1.5 mA/cm2 was applied for 100 sand the electrolyte was 0.5 M CuSO4. As the current is applied tothe PANI electrode, Cu2+ in the solution is reduced and depositedas Cu2O nanoparticles on the polyaniline-coated electrode (PANI/Cu2O).

3. Results and discussion

3.1. Morphological, XRD, and optical characterization of the fabricatedelectrode

To determine the phase and crystallinity of the deposited Cu2Onanoparticles, XRD was recorded and the pattern of PANI/Cu2O isshown in Fig. 2. The peaks corresponding to Cu2O viz. (110),

Fig. 3. FE-SEM images of PANI/Cu2O

(111), (200), (211), and (220) are in good agreement with JCPDScard no. 77-1049, which indicates the presence of a cubic systemwith a lattice constant of 4.258 Å [53]. There were no other peaksdue to impurities and the sharp peaks suggest that the particles arenanocrystalline in nature. In addition to the Cu2O peaks, a broadand poorly defined peak is present at 2h = 24�, which is ascribedto the presence of polyaniline [54]. Thus the XRD analysis showsthat the deposited copper oxide nanoparticles are in the Cu2Ophase.

The surface morphology of the fabricated PANI/Cu2O electrodewas investigated with FE-SEM and AFM analysis. The FE-SEMimages of the PANI/Cu2O electrode are shown in Fig. 3. As canbe seen in this figure, well dispersed Cu2O nanoparticles of sizesvarying from 80 to 100 nm are dispersed onto the polyanilinematrix. Thus, this process is a good method for the electroreduc-tion of Cu2+ to Cu2O nanoparticles. To further investigate thesurface roughness and thickness of the film, atomic force micros-copy (AFM) images of the samples were obtained (Fig. 4). In a re-cent study of the use of Cu2O films in CO2 electroreduction, itwas found that thickness and roughness play important rolesin product formation [55]. AFM images of polyaniline thin filmswithout and with Cu2O are shown in Fig. 4a and b respectively;Cu2O nanoparticles that are well dispersed over the polyanilinematrix are evident. The 2-D micrograph image of the film revealsits dense, uniform, and homogeneous morphology (Fig. 4c). Thedepth profile graph shows that the film has structural dips upto 145 nm and peaks up to 105 nm, which are orders ofmagnitude smaller than the total thickness and hence the as-pre-pared film is continuous. The roughness of the PANI/Cu2O filmwas estimated from Fig. 4 to be 58.3 nm over an area of25 � 25 lm2. The thickness of the film was estimated to bearound 156 nm.

Optical absorption behavior is one of the very important funda-mental properties in revealing the energy structures of theprepared thin films. The optical band gap is obtained by the follow-ing equation, with the help of absorption spectra.

films at different magnifications.

Fig. 4. Atomic force micrographs of (a) polyaniline (b) PANI/Cu2O thin film (c) 3-D AFM profile and (d) AFM sectional analysis of the PANI/Cu2O film.

A.N. Grace et al. / Applied Energy 120 (2014) 85–94 89

aht ¼ Aðht� EgÞn=2 ð4Þ

To determine the energy band gap, (aht)2 vs. (ht) was plotted,where ‘a’ is the absorption coefficient, ‘ht’ is the photon energy,’A’ is a constant, ’Eg’ is the band gap and ’n’ is either ½ for an indi-rect transition or 2 for a direct transition. Thus, a plot of (aht)2 vs.(ht) is a straight line whose intercept on the energy axis gives theenergy gap. According to the above equation, based on the directtransition as shown in Fig. 5, the band gap of as-obtained compos-ite film is 1.92 eV. This is lesser than the original band gap of 2.3 eVfor bulk Cu2O, which may be due to the influence of polyaniline inthe film [56]. The decrease in the band gap by introducing PANI asa supporting layer overtly indicates the better electron transferfrom the valence band to the conduction band, possibly becausethe PANI/Cu2O increase the recombination rate of electron–holepairs [56]. This phenomenon in turn will enhance the rate of CO2

reduction.

Fig. 5. Optical band gap spectrum of PANI/Cu2O modified electrodes.

3.2. XPS analysis of the fabricated electrode

XPS photoelectron spectroscopy was used to further probe thechemical nature of the polyaniline matrix and the chemical com-position of the deposited Cu2O, as shown in Fig. 6. All the peakswere corrected with reference to C1s (284.6 eV). In the wide scanXPS spectrum, peaks due to C 1s, N 1s, Cu 2p, and O 1s are presentat 287, 402, 535, and 937 eV, respectively. The peak at 402 eV cor-responding to N 1s confirms the presence of polyaniline in thecomposite [57]. The corresponding deconvoluted spectrum isshown in Fig. 7. The three C 1s peaks at 284.6, 285.4, and288.6 eV are due to non-oxygenated C or graphitic carbon (CAC),C in ACAO bonds, and carboxylate carbon (OAC@O) respectively.The N 1s peak was deconvoluted into three Gaussian peaks, whichcorrespond to AN@ (398.2 eV), ANHA (399.9 eV), and AN+@(401.8 eV) in polyaniline [57]. There are two main peaks at 933.3and 953.1 eV in the deconvoluted Cu 2p spectrum, which areattributed to the Cu+ double peaks of Cu 2p3/2 and Cu 2p1/2respectively [58]. The satellite peak at 941.7 eV located at a higher

Fig. 6. Wide scan X-ray photoelectron spectroscopy analysis (XPS) of PANI/Cu2Othin film.

Fig. 7. Deconvoluted XPS spectrum of C1s, N1s, Cu 2p and O 1s of PANI/Cu2O thin film.

-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3

-1.9

-1.3

-0.6

0.0

0.6

1.3

1.9

2.5

3.2

a b c

Potential VSCE

Cur

rent

den

sity

j/m

Acm

-2

Cur

rent

den

sity

j/m

Acm

-2

Potential VSCE

-0.3 -0.2 -0.1 0.0 0.1 0.2

Fig. 8. Current-potential curves of CO2 reduction on PANI/Cu2O electrode in 0.1 MTBAP/methanol (a) CO2 saturated (b) blank and (c) in N2; scan rate = 50 mV/s.

90 A.N. Grace et al. / Applied Energy 120 (2014) 85–94

binding energy can be assigned to Cu2+ in CuO or possibly Cu(OH)2.The intensity of this peak is relatively low, which implies that onlya small amount of Cu(OH)2 or CuO is present on the surface ofCu2O. Thus the XPS measurements demonstrate that Cu2O hasformed on the polyaniline matrix. The O 1s peaks were fitted tothree peaks with the binding energies 530.8, 531.2, and 532.8 eV.The peaks at 530.8 and 532.8 eV are attributed to lattice oxygenO2- in the Cu2O and CuO phases respectively [59]. The other peakat 531.2 eV is attributed to adsorbed ‘O’ on the surface [60].

3.3. Electrochemical CO2 reduction

Large overpotentials are generally required in electrochemicalprocess, which means they are inefficient consumers of energyand restricts practical applications. To diminish overpotentials,considerable effort has recently been invested in the developmentof various homogeneous and heterogeneous electrocatalysts. Oneeffective means of reducing the overpotential is the use of an elec-tron mediator consisting of an inorganic conductor and a conduc-tive polymer film; it is has been claimed that the electronmediation effect is considerably enhanced by attaching a metalcomplex to the conducting polymer. The electrochemical reductionwas carried out in TBAP in methanol and the results are presentedin the following sections.

3.3.1. Voltammetric behavior of the fabricated electrodeCyclic voltammetry is a good technique for ascertaining the CO2

reduction potential and the onset. To initially test the feasibility ofCO2 reduction with the fabricated electrode, current-potentialcurves were recorded from �0.4 to 0.2 V vs. SCE in 0.1 MTBAP + methanol as the supporting electrolyte under ambient con-ditions. The potential was scanned at a sweep rate of 50 mV/s and atypical curve for PANI/Cu2O electrode saturated with CO2 is shownin Fig. 8. There is no sign of reduction in the absence of CO2

(Fig. 8b), and when CO2 is purged, a reduction peak was observedat �0.1 V vs. SCE (Fig. 8a). In the presence of CO2, there is an abrupt

onset of catalytic current at approximately �0.1 V vs. SCE. Thus theCV profile confirms the reduction of CO2 using the developed cat-alyst. Cyclic voltammetry was performed using a bare Pt loadedC electrode without catalyst in 0.1 M TBAP + methanol under thesame experimental conditions (see the inset in Fig. 8). No reductionpeaks were present, which shows that the bare electrode makes nocontribution to the CO2 reduction. In addition to this, as a controlthe reaction was carried out under N2 atmosphere using PANI/Cu2O electrode in 0.1 M TBAP + methanol under the same experi-mental conditions but without CO2 (Fig. 8c). Results showed noreduction and product formation. This indicated that the productswere exclusively from CO2 reduction and not from electrolytes orother contamination. Consequently the products were not formedfrom the decomposition of methanol and supporting salts at thecathode and hence were produced only from the electrochemical

A.N. Grace et al. / Applied Energy 120 (2014) 85–94 91

reduction of CO2. Once the onset potentials were determined fromthe voltammetry measurements, further experiments were at-tempted to investigate the electrochemical reduction of CO2 inCO2 saturated electrolyte at cathodic polarizations exceeding theonset potential.

3.3.2. Electroreduction of CO2-constant potential electrolysisElectrolyses were performed under potentiostatic conditions

while the current and product concentration were monitored asfunctions of time. Prior to electrolysis, the electrolyte solutionwas saturated with pure CO2 for 1 h. Controlled potential electrol-ysis was carried out under the same experimental conditions asused in the cyclic voltammetry experiment. We investigated themechanism of CO2 reduction by analyzing the faradaic efficienciesof the products: five different cathodic potentials of �0.4, �0.3,�0.25, �0.15, and �0.10 V were applied to the electrode with elec-trolysis duration of 120 min (Fig. 9). During the electrolysis, thecathodic current asymptotically closed in the steady state at�0.25, �0.15, and �0.10 V and the current densities were almostequivalent. From the results of the electrolysis measurementsshown in Fig. 9, a current potential curve was plotted (Fig. 10).The maximum current density was observed at �0.3 V vs. SCE.Note that the current density tends to decrease at potentials lessthan �0.3 V. At higher potentials, the efficiency is lower due tothe retardation of the electrochemical reduction of CO2. This resultindicates that the transport of CO2 to the electrode surface is con-trolled by diffusion, which is supported by the linear dependenceof current vs. sq.rt of scan rate in the CVs (Fig. S1). To further know

Fig. 9. Current density profile obtained during CO2 electroreduction at variouspotentiostatic electrolysis potentials for 120 min in 0.1 M TBAP/methanol.

Fig. 10. I–V plot of the CO2 reduction on PANI/Cu2O electrode in 0.1 M TBAP/methanol.

the faradaic efficiencies of the products at various potentials, thesample was withdrawn at 30 min during electrolysis and the sameis plotted in Fig. 11. As can be seen in this figure, the formationrates typically increased with the potential from �0.1 V (SCE),reached maxima near �0.3 V, and then started to decrease. Hence,�0.3 V was chosen for the electrochemical reduction of CO2 andsubsequent experiments were carried out at this potential.

To determine the formation rate of formic acid at �0.3 V, elec-trolysis was carried out for 120 min as shown in Fig. 9 and theproducts were examined at regular intervals. Fig. 12 shows the cor-responding geometric current density vs. time (reduction currentprofile) for the electroreduction of CO2 with the fabricated elec-trode at �0.3 V. A high initial current density is evident in the first10 min of electrolysis, which thereafter declines slowly, reaching0.05 mA/cm2 after 2 h of electrolysis. During this period of electrol-ysis, samples were periodically withdrawn and the products wereexamined with ion chromatography analysis.

The corresponding faradaic efficiency vs. time diagram for theelectroreduction of CO2 at �0.3 V over a period of 2 h is shown inFig. 12. As can be seen in the figure, acetic acid is the main productwith an efficiency of 63.0% initially and a value of 19.2% after 2 h.The formation efficiency for formic acid was initially 30.4% and de-clined to a value of 7.7% after 2 h. These results are in accordancewith those of previous studies [45]. These results are reasonablebecause this electrochemical reduction of CO2 occurred at a low

Fig. 11. Faradaic efficiencies (%) vs. various cathodic potentials on PANI/Cu2Oelectrode. Applied time is 30 min for each experiment.

Fig. 12. Current density profile and Faradaic efficiencies of CO2 reduction on PANI/Cu2O electrode in TBAP/methanol system (�0.3 V vs. SCE). (Error bars representstandard deviation of three repetitive experiments).

Table 2Faradaic efficiency obtained in electrochemical reduction of CO2 using PANI/Cu2Oelectrodes in 0.1 M TBAP/methanol at different time period.

Time (min) Charge (A s) HCOOH CH3COOH

mmol g (%) mmol g (%)

30 36.7 0.058 30.4 0.12 63.060 39.2 0.039 19.2 0.11 54.190 46.01 0.029 12.1 0.08 33.6

120 60.14 0.024 7.7 0.06 19.2

92 A.N. Grace et al. / Applied Energy 120 (2014) 85–94

overpotential of -0.3 V (vs. SCE). In addition, the reduction was car-ried out under ambient conditions without increased pressure orreduced temperature with a simple catalyst. The overall results ob-tained for the electrochemical reduction of CO2 at �0.3 V are givenin Table 2. Acetic acid is the main product with a current efficiencyof 63%. To suffice the ion chromatography results, samples electro-lyzed after CO2 electroreduction for 2 h at �0.3 V were analyzed byHPLC. Results showed the presence of formic and acetic acid with0.0284 and 0.0519 mmol, which matched with IC results. To fur-ther confirm the formation of formic and acetic acids, 1H NMRspectrum was recorded and the results are given in Fig. S2. As seenfrom the spectrum, formic and acetic acid could be seen at 8.27 and2.2 ppm and the other peaks are due to the supporting electrolyteand unreacted methanol in the solution [61–64]. Apart from thesestudies, to further prove that the reaction products were from CO2

and methanol, labeled experiments were done and in this view,two independent tests were carried out. In the first, 13CO2 was usedas a source gas with unlabeled CH3OH as electrolyte and 0.1 MTBAP as supporting electrolyte. The formed product was identifiedby GC-MS and H NMR spectroscopy. In the mass spectrum of aceticacid extracted from GC-MS analysis (Fig. S3a), results showed thatthe molecular ion peak appeared at an m/z value of 61, whereas forstandard unlabeled acetic acid, it is well known that the molecularion peak appear at 60 [65]. Hence this peak at m/z value of 61 cor-respond to CH3COOH+ with one carbon labeled. To further identifythe position of C13 in acetic acid, the peak at m/z 46 was lookedinto, as this corresponds to the fragmented COOH+. According tothe standard fragmentation pattern of acetic acid, instead of 45,the peak appeared at 46 corresponding to 13COOH+ [66]. This is aclear evidence that the labeled C has gone to ACOOH group of ace-tic acid. Peaks at 43 and 44 correspond to CH2

13CO+ and CH313CO+,

which clearly stated that only one carbon is labeled. As an add-onproof, H NMR was carried out and the corresponding spectrum isgiven in Fig. S4a. It could be observed that the HCOOH protonsshowed a doublet attributed to the protons bound to the 13C atomin H13COOH, which is an evidence that the carbon in HCOOH hasoriginated from CO2 gas (marked blue). In a recent report on label-ing experiment, similar doublet peak in H NMR spectrum was ob-served for H13COOH [67]. Acetic acid protons are at region 2.2 and11 ppm for ACH3 and ACOOH groups respectively [62]. As ob-served in Fig. S4a, the former was not affected whereas the lattergot splitted into two signals, which clearly showed that 13C hasgone to ACOOH group of acetic acid and the CH3 carbon must havecome from methanol as per the mechanism. These phenomenonare termed as carbon satellite peaks [68].

To further supplement the above mechanism, in the secondexperiment, both 13CO2 and 13CH3OH were used and the experi-ments were carried out at similar conditions. The product was ana-lyzed again using GC-MS (Fig. S3b) and results showed that the m/zvalue incremented to 62 corresponding to 13CH3

13COOH+. This di-rectly proved the presence of 13CH3

13COOH and also is an evidencethat the 13C in CH3 group has come from methanol. The peak at m/z46 due to 13COOH+ remain unchanged whereas peaks at m/z = 43and 44 observed in the first experiment increased to 44 and 45.These peaks corresponded to the 13CH2

13CO+ and 13CH313CO+

groups. This clearly shows that the labeled carbon of methanolhas gone to the ACH3 group of acetic acid. The correspondingNMR spectrum for the product obtained with 13CO2 and 13CH3OHis given in Fig.S4b. It can be seen that the peak at 2.2 ppm is nowobserved as a doublet (marked green). This region corresponds tothe ACH3 group of acetic acid and thus proving that the labeledC13 in methanol has gone to CH3 group of acetic acid. All these re-sults indicated the formation of acetic acid from the process withone C atom in the product originated from CO2 gas and the otherfrom methanol electrolyte. Thus both the above experiments haveclearly shown the formation of acetic acid from formic acid andmethanol electrolyte.

The general scenario for CO2 electroreduction is the formationof radical anionic species (COO��) as the initial step, which requiresa potential of �2.144 V (SCE, sat. KCl) in both protic and aproticmedia whereas in our system, the reduction potential is �0.3 V(vs. SCE, sat. KCl) and hence it is not feasible to generate CO��2 . Withreference to labeling experiments and the role of polyaniline film[45], the mechanism is explained as follows: During CO2 electrore-duction, it was found that the adsorption of CO2 on PANI via hydro-gen bonding is a preliminary step guiding the reduction process[45,46]. The ANH groups of polyaniline interacts with CO2 mole-cule forming hydrogen bonds. Initially, at �0.3 V, H adatoms(Had) are generated on the electrode surface. The as generatedHad atoms are then transferred to the CO2 molecules through thehydrogen bonding to form HCOOads. This reacts with further Had

to produce formic acid. The formic acid is attacked by the solvent(CH3OH) to form acetic acid. The whole mechanism can be de-scribed with the following equations [45]:

Hþ þ e� ! Had ð5Þ

CO2 þHad ! HCOOad ð6Þ

HCOOad þHad ! HCOOH ð7Þ

HCOOad þ CH3OH! CH3COOHþ OHad ð8Þ

OHad þHad ! H2O ð9Þ

The formation of Had has also been reported on reduced Prus-sian blue, which in turn favors CO2 reduction [69]. In another workstudy by Ogura et al., CO2 was reduced at �0.845 V (SCE, sat. KCl)by using a mediator electrode system with a polymer [44]. Thereduction potential achieved in our study using the developed cat-alyst is significantly lower than other reports. The high efficiencyand low reduction potential achieved with this new catalyst arealso due to the Cu2O dispersed on the polyaniline matrix. Cu2Ohas been utilized for the electrochemical reduction of CO2 [70–72]. There are good reasons to believe that the Cu+ ions in Cu2Oare the active sites that promote CO2 reduction. In particular, theelectronic properties of Cu2O, which is a p-type semiconductor,play an important role in the adsorption of CO2. The oxygen speciesin copper oxides increase the number of defect electrons so thatCO2 can easily be adsorbed onto the catalyst surface [70]. It hasalso been suggested by Freso et al. that CO2 is initially chemisorbedon the oxidized copper surface (p-Cu2O sites) to form COads andOads. The O2 anions in Cu2O are present in a body centered latticeand the Cu+ cations are present in a face centered lattice. The va-lence band states in Cu+ are sources for electrons for chemisorp-tion. As a result, the presence of Cu+ species on the surfacepromotes CO2 reduction reactions [71].

Park et al. have shown that the presence of copper oxides signif-icantly increases the carbon dioxide adsorption capacity [72]. It iswell known that CO2 molecules are soft acids. Copper oxide nano-particles have electron-donor features, which results in theenhancement of the adsorption capacity of CO2 molecules because

A.N. Grace et al. / Applied Energy 120 (2014) 85–94 93

of their electron-acceptor properties. It was also postulated thatthere can be repulsive forces between CO2 and transition metalssuch as Cu, which have acidic features (electron acceptors),whereas oxides such as Cu2O enhance the adsorption of CO2 be-cause they have basic features and are electron-acceptor friendly.Hence the basicity of Cu2O means that more CO2 molecules be-come attached to the electrode, where they are protonated and re-duced to formic acid with the assistance of the polyaniline matrix.Thus a synergistic effect between PANI and Cu2O enables CO2

reduction and the presence of Cu2O enhances CO2 adsorption ontothe polymer surface. On the polymeric surface, the adsorbed CO2 isconverted to formic acid through the mechanism mentionedabove. Although the efficiency is high, the exact role of the catalystmaterial needs to be further explored and additional work isunderway to further probe the role of such nanocomposites. Thisresearch can contribute to the large scale manufacturing of fuelsfrom inexpensive catalyst materials.

4. Conclusion

The electrochemical reduction of CO2 with a PANI/Cu2O-basedelectrode in an electrolyte consisting of TBAP and methanol hasbeen investigated. The catalyst consisted of Cu2O nanoparticlesdeposited onto a polyaniline matrix with an electrochemical meth-od; our results demonstrated the presence of Cu(I) species on theelectrode surface. The developed catalyst was tested in the electro-chemical reduction of CO2 by electrolysis at various potentials. Theproducts were formic and acetic acids at all the studied potentials.The faradaic efficiency for the formation of formic acid was 30.4%at a polarization potential of �0.3 V vs. SCE (sat. KCl). Thus, thepresent study offered a simple and effective way for electrochem-ical reduction of CO2 to formic acid with an appreciable faradaicefficiency. The synthesis of formic acid from readily available inex-pensive raw materials via the electrochemical reduction of CO2 isexpected to be useful in fuel production.

Acknowledgements

‘‘The work was supported through Korea CCS R&D centerfunded by the Ministry of Science, ICT & Future Planning of KoreanGovernment’’.

Appendix A. Supplementary material

Supplementary information associated with this article regard-ing materials, Fabrication, synthesis, electrochemical details,Instrumentation, polyaniline growth mechanism, GC-MS andNMR results can be found in the online version, at http://dx.doi.org/10.1016/j.apenergy.2014.01.022.

References

[1] Crowley TJ, Berner RA. CO2 and climate change. Science 2001;292:870–2.[2] Honisch B, Ridgwell A, Schmidt DN, Thomas E, Gibbs SJ, Sluijs A, et al. The

geological record of ocean acidification. Science 2012;335:1058–63.[3] Pacala S, Socolow R. Stabilization wedges: solving the climate problem for the

next 50 years with current technologies. Science 2004;305:968–72.[4] Aresta M, Dibenedetto A. Utilisation of CO2 as chemical feedstock:

oppurtunities and challenges. Dalton Trans 2007;28:2975–92.[5] Harris DC. Charles david keeling and the story of atmospheric CO2

measurements. Anal Chem 2010;82:7865–70.[6] Li B, Duan Y, Luebke D, Morreale B. Advances in CO2 capture technology: a

patent review. Appl Energy 2013;102:1439–47.[7] Bell AT. Basic research needs: catalysis for energy. Report from the U.S.

Department of Energy, Basic Energy Sciences, Workshop; 2007; August 6–8:69.

[8] Jiang B, Wang X, Gray ML, Duan Y, Luebke D, Li B. Development of amino acidand amino acid-complex based solid sorbents for CO2 capture. Appl Energy2013;109:112–8.

[9] Hori Y. Electrochemical CO2 reduction on metal electrodes. Mod AspElectrochem 2008;42:89–189.

[10] Inui T, Anpo M, Izui K, Yanagida S, Yamaguchi T. Advances in chemicalconversions for mitigating carbon dioxide, vol. 114. Elsevier Science; 1998. p.1–699.

[11] Zhang M, Guo Y. Rate based modeling of absorption and regeneration for CO2

capture by aqueous ammonia solution. Appl Energy 2013;111:142–52.[12] Song C, Gaffney A, Fujimoto K. CO2 Conversion and utilization. In: American

Chemical Society symposium series; 2002. p. 809.[13] Wang H, Leung DYC, Xuan J. Modeling of a microfluidic electrochemical cell for

CO2 utilization and fuel production. Appl Energy 2013;102:1057–62.[14] Angamuthu R, Byers P, Lutz M, Spek AL, Bouwman E. Electrocatalytic CO2

conversion to oxalate by a copper complex. Science 2010;327:313–5.[15] Halmann MM. Chemical fixation of carbon dioxide: methods for recycling CO2

into useful products. CRC Press; 1993. p. 1–192.[16] Sullivan BP. Electrochemical and electrocatalytic reduction of carbon

dioxide. Amsterdam, Netherlands: Elsevier Press; 1993.[17] Soussan L, Riess J, Erable B, Delia ML, Berge A. Electrochemical reduction of CO2

catalysed by Geobacter sulfurreducens grown on polarized stainless steelcathodes. Electrochem Commun 2013;28:27–30.

[18] Mignard D, Sahibzada M, Duthie JM, Whittington HW. Methanol synthesisfrom flue-gas CO2 and renewable electricity: a feasibility study. Int J HydrogenEnergy 2003;28:455–64.

[19] Peterson AA, Norskov JK. Activity descriptors for CO2 electroreduction tomethane on transition-metal catalysts. J Phys Chem Lett 2012;3:251–8.

[20] Wu J, Risalvato FG, Ke FS, Pellechia PJ, Zhou XD. Electrochemical reduction ofcarbon dioxide I. Effects of the electrolyte on the selectivity and activity withSn electrode. J Electrochem Soc 2012;159:F353–9.

[21] Hu B, Curtis G, Suib SL. Thermal, electrochemical and photochemicalconversion of CO2 to fuels and value added products. J CO2 Utiliz2013;1:18–27.

[22] Benson EE, Kubiak CP, Sathurum AJ, Smeja JM. Electrocatalytic andhomogeneous approaches to conversion of CO2 to liquid fuels. Chem Soc Rev2009;38:89–99.

[23] Kwon Y, Lee J. Formic acid from carbon dioxide on nonlayered electrocatalyst.Electrocatalysis 2010;1:108–15.

[24] Hussain ST, Rahman MH. Nano catalyst for CO2 conversion to hydrocarbons. JNanosyst Technol 2009;1:1–9.

[25] Kirk-Othmer. Encyclopedia of chemical technology. John Wiley & Sons Inc.;2004.

[26] Oloman C. Electrochemical processing of carbon dioxide. Chem Sus Chem2008;1:385–91.

[27] Agarwal AS, Zhai Y, Hill D, Sridhar N. The electrochemical reduction of carbondioxide to formate/formic acid: engineering and economic feasibility. ChemSus Chem 2011;4:1301–10.

[28] Li H, Oloman C. Development of a continuous reactor for the electroreductionof carbon dioxide to formate – Part 1: process variables. J Appl Electrochem2006;36:1105–15.

[29] Chen Y, Kanan MW. Tin oxide dependence of the CO2 reduction efficiency ontin electrodes and enhanced activity for tin/tin oxide thin-film catalysts. J AmChem Soc 2012;134:1986–9.

[30] Kwon Y, Lee J. Formic acid from carbon dioxide on nanolayeredelectrocatalysts. Electrocatalysis 2010;1:108–15.

[31] Akahori Y, Iwanaga N, Kato Y, Hamamoto O, Ishii M. New electrochemicalprocess for CO2 reduction to from formic acid from combustion flue gases.Electrochemistry 2004;72:266–70.

[32] Ito K, Ikeda S, Iida T, Nomura A. Denki Kagaku 1982;50:463.[33] Ito K, Ikeda S, Yamauti N, Iida T, Takagi T. Electrochemical reduction products

of carbon dioxide at some metallic electrodes in nonaqueous electrolytes. BullChem Soc Jpn 1985;58:3027–8.

[34] Ikeda S, Takagi T, Ito K. Selective formation of formic acid, oxalic acid, andcarbon monoxide by electrochemical reduction of carbon dioxide. Bull ChemSoc Jpn 1987;60:2517–22.

[35] Lide DR. Handbook of chemistry and physics. 72nd ed. Boca Raton: CRC Press;1991. p. 6, 4, 3–321.

[36] Chihara H. Kagaku binran-kiso (handbook of basic chemistry, in Japanese), 3rded., vol. II, Maruzen, Tokyo; 1984. p. 158.

[37] Kaneco S, Hiei NH, Xing Y, Katsumata H, Ohnishi H, Suzuki T, et al. ElectrochimActa 2002;48:51–5.

[38] Naitoh A, Ohta K, Mizuno T, Yoshida H, Sakai M, Noda H. Electrochemicalreduction of carbon dioxide in methanol at low temperature. Electrochim Acta1993;38:2177–9.

[39] Kaneco S, Iiba K, Ohta K, Mizuno T. Electrochemical CO2 reduction on a copperwire electrode in tetraethylammonium perchlorate methanol at extremelylow temperature. Energy Sources 1999;21:643–8.

[40] Kaneco S, Iwao R, Katsumata H, Suzuki T, Ohta K. Carbon dioxide sequestrationtechnology by electrochemical conversion at cadmium electrode in methanolunder mild conditions. Photo/Electrochem Photobiol Environ Energy Fuel2003;2:181–9.

[41] Koleli F, Ropke T, Hamann CH. Electrochemical impedance spectroscopicinvestigation of CO2 reduction on polyaniline in methanol. Electrochim Acta2003;48:1595–601.

[42] Schrebler R, Cury P, Suarez C, Munoz E, Gomez H, Cordova R. Study of theelectrochemical reduction of CO2 on a polypyrrole electrode modified byrhenium and copper–rhenium microalloy in methanol media. J ElectroanalChem 2002;533:167–75.

94 A.N. Grace et al. / Applied Energy 120 (2014) 85–94

[43] Li Y, Peng H, Li G, Chen K. Synthesis and electrochemical performance ofsandwich-like polyaniline/graphene composite nanosheets. Eur Polymer J2012;48:1406–12.

[44] Ogura K, Endo N, Nakayama M. Mechanistic studies of CO2 reduction on amediated electrode with conducting polymer and inorganic conductor films. JElectrochem Soc 1998;145:3801–9.

[45] Aydin R, Koleli F. Electrochemical reduction of CO2 on a polyaniline electrodeunder ambient conditions and at high pressure in methanol. J ElectroanalChem 2002;535:107–12.

[46] Koleli F, Ropke T, Hamann CH. The reduction of CO2 on polyaniline electrode ina membrane cell. Synth Met 2004;140:65–8.

[47] Aydin R, Koleli F. Electrocatalytic conversion of CO2 on a polypyrrole electrodeunder high pressure in methanol. Synth Met 2004;144:75–80.

[48] Aydin R, Dogan HO, Koleli F. Electrochemical reduction of carbon dioxide onpolypyrrole coated copper electro-catalyst under ambient and high pressure inmethanol. Appl Catal B 2013;140–141:478–82.

[49] Khan KA. Stability of a Cu2O photoelectrode in an electrochemical cell and theperformances of the photoelectrode coated with Au and SiO2 thin films. ApplEnergy 2000;65:59–66.

[50] Buron C, Lakard B, Monnin A, Moutarlier V, Lakard S. Elaboration andcharacterization of polyaniline films electrodeposited on tin oxides. SynthMet 2011;161:2162–9.

[51] Leon-Silva U, Nicho M, Hu H, Cruz-Silva R. Effect of modified ITO substrate onelectrochromic properties of polyaniline films. Sol Energy Mater Sol Cells2007;91:1444–8.

[52] Peng C, Hu D, Chen GZ. Theoretical specific capacitance based on chargestorage mechanisms of conducting polymers: comment on vertically orientedarrays of polyaniline nanorods and their super electrochemical properties.Chem Commun 2011;47:4105–7.

[53] Park JC, Kim J, Kwon H, So H. Gram-scale synthesis of Cu2O nanocubes andsubsequent oxidation to CuO hollow nanostructures for lithium-ion batteryanode materials. Adv Mater 2009;21:803–7.

[54] Wei H, Zhu J, Wu S, Wei S, Guo Z. Electrochromic polyaniline/graphite oxidenanocomposites with endured electrochemical energy storage. Polymer2013;54:1820–31.

[55] Li CW, Kanan MW. CO2 reduction at low overpotential on Cu electrodesresulting from the reduction of thick Cu2O films. J Am Chem Soc2012;134:7231–4.

[56] Xue T, Loo LS, Wang X, Kwak SK, Lee JM. Electrodeposition of mesoporousbilayers of polyaniline supported Cu2O semiconductor films from lyotropicliquid crystalline phase. Chem Eng Sci 2012;80:452–9.

[57] Gao Z, Yang W, Wang J, Yan H, Yao Y, Ma J, et al. Electrochemical synthesis oflayer-by-layer reduced graphene oxide sheets/polyaniline nanofibers

composite and its electrochemical performance. Electrochim Acta2013;91:185–94.

[58] Hua Q, Shang DL, Zhang WH, Chen K, Chang SJ, Ma YS, et al. Morphologicalevolution of Cu2O nanocrystals in an acid solution: stability of different crystalplanes. Langmuir 2011;27:665–71.

[59] Teo JJ, Chang Y, Zeng HC. Fabrications of hollow nanocubes of Cu2O and Cu viareductive self-assembly of CuO nanocrystals. Langmuir 2006;22:7369–77.

[60] Wang L, Wei G, Qi B, Zhou HL, Liu ZG, Song YH. Electrostatic assembly of Cu2Onanoparticles on DNA templates. Appl Surf Sci K 2006;252:2711–6.

[61] Berregi I, Campo GD, Caracena R, Miranda JI. Quantitative determination offormic acid in apple juices by 1H NMR spectrometry. Talanta 2007;72:1049–53.

[62] Birnie III DP, Bendzko NJ. 1H and 13C NMR observation of the reaction of aceticacid with titanium isopropoxide. Mater Chem Phys 1999;59:26–35.

[63] Poh BL, Teem CM. A 1H NMR study on the complexation oftetraalkylammonium cations, mono and diprotonated amines, and aminoacids with a derivatized cyclotetrachromotropylene in an aqueous solution.Tetrahedron 2005;61:5123–9.

[64] Fulmer GR, Gottlieb HE, Miller AJM, Nudelman A, Sherden NH, Stoltz BM, et al.NMR chemical shifts of trace impurities: common laboratory solvents,organics, and gases in deuterated solvents relevant to the organometallicchemist. Organometallics 2010;29:2176–9.

[65] Savary BJ, Nuriez A. Gas chromatography–mass spectrometry method fordetermining the methanol and acetic acid contents of pectin using headspacesolid-phase microextraction and stable isotope dilution. J Chromatogr A2003;1017:151–9.

[66] Leach S, Schwell M, Jochims HW, Baumgartel H. VUV photophysics of aceticacid: fragmentation, fluorescence and ionization in the 6–23 eV region. ChemPhys 2006;321:171–82.

[67] Maeda K, Sekizawa K, Ishitani O. A polymeric-semiconductor–metal-complexhybrid photocatalyst for visible-light CO2 reduction. Chem Comm2013;49:10127–9.

[68] Gunther H. NMR spectroscopy: basic principles, concepts, and applications inchemistry. 2nd ed. John Wiley & Sons; 1995. p. 214–219. ISBN:0-471-95201-X.

[69] Ogura K, Nakayama M, Kusumoto S. In situ fourier transform infraredspectroscopic studies on a metal complex-immobilized polyaniline/Prussianblue-modified electrode and the application to the electroreduction of CO2. JElectrochem Soc 1996;143:3606–15.

[70] Chang TY, Liang RM, Wu PW, Chen JY, Hsieh YC. Electrochemical reduction ofCO2 by Cu2O-catalyzed carbon clothes. Mater Lett 2009;63:1001–3.

[71] Frese KW. Electrochemical reduction of CO2 at intentionally oxidized copperelectrodes. J Electrochem Soc 1991;138:3338–44.

[72] Kim BJ, Cho KS, Park SJ. Copper oxide-decorated porous carbons for carbondioxide adsorption behaviors. J Colloid Interf Sci 2010;342:575–8.