Effective Photocurrent Enhancement in Nanostructured CuO by

Feb 6, 2018 - Effective Photocurrent Enhancement in Nanostructured CuO by Organic Dye Sensitization: Studies on Charge Transfer Kinetics. Priyanka Mar...
1 downloads 9 Views 8MB Size
Article Cite This: J. Phys. Chem. C XXXX, XXX, XXX−XXX

pubs.acs.org/JPCC

Effective Photocurrent Enhancement in Nanostructured CuO by Organic Dye Sensitization: Studies on Charge Transfer Kinetics Priyanka Marathey, Ranjan K. Pati, Indrajit Mukhopadhyay, and Abhijit Ray* Solar Research and Development Center, Department of Solar Energy, Pandit Deendayal Petroleum University, Raisan, Gandhinagar, Gujarat 382-007, India S Supporting Information *

ABSTRACT: Mercurochrome-sensitized nanostructured CuO grown directly on Cu has been used as efficient photocathode in photoelectrochemical cell. The photocurrent density of the sensitized electrode is found to enhance from −1 mA/cm2 (unsensitized) to −2.2 mA/cm2 (dye sensitized for 24 h) at 0 V vs RHE under AM1.5G artificial solar illumination in aqueous 0.5 M Na2SO4 solution. The photoluminescence spectra demonstrated a strong absorption at 465 nm by the mercurochrome. Subsequent transfer of the photoexcited electrons to CuO conduction band enhances the photocurrent density. An incident photon-to-current conversion efficiency (IPCE) of 8% has been observed in the visible region. This is so far the highest reported value in p-type CuO based organic dyesensitized photocathode. The dye sensitized CuO has thus the potential of photoreduction with higher Faradaic efficiency for various redox species due to additional carrier injection.

1. INTRODUCTION The photoelectrochemical (PEC) mode of solar energy conversion in the form of electrical power,1 solar fuel via water splitting2 and photocatalytic reduction of carbon dioxide3 is one of the renewable energy conversions which have attracted much attention globally. A large number of semiconducting compounds have been investigated in the costeffective PEC based solar energy conversation (both solar to electricity and solar to chemical fuel, like H2 or methanol etc.) such as TiO2, ZnO, α-Fe2O3, CdS, SnS2, SrTiO3, MoS2, CuO, Cu2O, CaFe2O4, etc for their elemental abundance on the earth crust. Wide band gap semiconductors, such as ZnO, TiO2, SrTiO3 etc which are sensitive only to UV region of the solar spectrum, have shown to produce visible light induced excitation by sensitization through doping4−7 or dyes.8 On the other hand, many narrow band gap materials, such as CdS, CuO, Cu2O, and Cu2ZnSnS4 etc have slightly negative conduction band position with respect to the H2/H+ redox level of water. Hence, these semiconductors require a catalyst, such as Pt or RuO2 having Fermi level at a more positive energy relative to their Fermi level in order to match the pseudo-Fermi level of H2/H+.9 Dye sensitization of the narrow band gap semiconductors by organic dyes has been found to be effective in oxides, NiO,10,11 CuO,12,13 Cu2O14 as well as chalcogenide, Cu2XSnS4 (X = Zn, Ni, Co, and Mn)15 semiconductors where anodic (cathodic) sensitization injects extra electrons (holes) into the conduction (valence) band. A handful of organic dye sensitizers has been demonstrated in NiO and CuO based dye sensitized solar cells (DSSCs), including porphyrinoids, such as © XXXX American Chemical Society

phthalocyanines of zinc (ZnPc), P1 (DN-FP01), DPP-NDI, YF1 etc as cost-effective alternative to the expensive Ru-based dye “N3” [chemically, Ru(4,4′-dicarboxy-2,2′-bipyridine) 2 (NCS) 2 ], commonly used in dye sensitized solar cells.16,17 However, many of these dyes show absorption peak in visible to infrared photon energies, which may overlap with that of the above-mentioned low-band gap semiconductors leading to a hindrance to photogeneration in them. On the other hand, the 9-phenylxanthene dyes (PXD), such as flurecin, eosin-Y, rhodamine B, and mercurochrome (merbromine) are demonstrated to be efficient visible light sensitizers where the absorption peaks appear at higher energies than that of the band gap energies of CuO or NiO. Among the abovementioned PXDs, Mercurochrome (structure as presented in Figure 2d) is a low cost dye that can be adsorbed on various oxide semiconductor surface with a stable carboxylate linkage. Among the narrow band gap semiconductors, NiO, CuO, and Cu 2O are promising candidates for their natural abundance, work function close to that of Pt (∼5.3 eV), and easy processing ability in various forms such as polycrystalline films and variety of nanostructures. These materials have been demonstrated as photocathodes in photoelectrochemical cells18,19 as well as in DSSCs,12,13,20,21 and their nanoparticles as photocatalysts.22,23 However, these compounds are having rather poor electrical conductivity causing slow electron Received: October 10, 2017 Revised: January 12, 2018 Published: February 6, 2018 A

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 1. Mircrostructure of the unsensitized as-prepared (a) and 250 °C annealed (b) CuO grown on Cu. X-ray diffraction spectra (c) of the CuO nanostructure before and after annealing at 150 and 250 °C.

not only for the DSSCs with better stability but also for various other photoredox applications, such as water splitting and photocatalysis which need neutral pH media. Moreover, the possibility of electron injection from a dye into the conduction band of a p-type semiconductor and its kinetics has not yet been studied in general. In this work, nanostructured CuO was grown directly over Cu foil and its photoelectrochemical performance was investigated in a three electrode configuration. The organic chromophore sensitization was found to enhance the photocurrent density in the presence of neutral electrolyte which is beneficial for both dye-sensitized solar cells and photoredox applications. A combined effect of photogeneration in CuO and carrier injection in its conduction band is proposed to become responsible for the photocurrent enhancement.

transfer kinetics with the electrolytes. In this respect, photosensitizers like dyes active in visible solar spectrum can help pumping extra electrons to the conduction band of the absorber material. Usually, when the charge injection kinetics from the excited state of the dye is faster than the recombination one, the electron−hole recombination can be substantially reduced. On the other hand, earlier studies on CuO based photoelectrochemical cells were demonstrated in the presence of either iodine (I3−/I−) or cobalt (cobalt bipyridine complex) based mediator electrolytes and these studies have established the CuO as DSSC cathode only replacing noble Pt.12,13 To the best of our knowledge, there are no reports available on dye-sensitized CuO based photocathodes with the enhancement of photocurrent density beyond 1 mA/cm2 in neutral pH electrolyte, which is beneficial B

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 2. UV−visible absorption (a) and photoluminescence emission (b) spectroscopy of the free dye mercurochrome (molecular structure as in part d) and the dye sensitized CuO dispersed in ethanol. The intersection of the absorption and emission spectra (c) of the dye correspond to the HOMO−LUMO energy difference of the dye.

2. EXPERIMENTAL SECTION 2.1. Fabrication of Dye-Sensitized Nanostructured CuO. Nanostructured CuO was grown by chemical etching of Cu-foil having a dimension of 1 cm ×1 cm in hot alkali bath as described in details elsewhere.24 However, the growth also occurs at room temperature and the nanostructure in the present case was grown at this temperature. Briefly, Cu-foil of thickness 25 μm (from Alfa Aesar, 99.8%) ultrasonically cleaned sequentially in acetone, dilute HCl, ethanol, and thoroughly rinsed by deionized (DI) water was etched in a bath containing a mixture of 3 M NaOH (from Merck) and 0.2 M (NH4)2S2O8 (from Sigma-Aldrich) at room temperature for 20 min. This step was followed by rinsing with DI water and ethanol and drying in an oven at 60 °C for 2 h. The obtained nanostructures on Cu are primarily Cu (OH)2 as confirmed by X-ray diffraction. This nanostructure was converted primarily into CuO by annealing at 150 °C for 4 h in air. The dye sensitization was done by immersing the nanostructured CuO in 0.5 mM ethanol solution of mercurochrome for various durations of 12−72 h at room temperature. In order to test the comparative performance with other organic dyes, Eosyn-Y and Rhodamine-B was selected with similar concentration. 2.2. Structural and Optical Characterization. The structural characterization of the films was carried out by Xray diffraction (PANalytical X’Pert Pro) using Cu Kα radiation, (1.540598 Å), with a step size of 0.05 s/step in the range of 2θ = 10−80°. The microstructure was analyzed by field emission scanning electron microscope (FESEM) (Zeiss, Ultra 55) with an accelerating voltage of 5 kV. The transmission electron microscopy (TEM) was conducted in a high resolution-TEM

system (JEOL, JEM 2100). Optical characterization was carried out by a UV−visible diffuse reflectance spectrophotometer (Shimadzu, UV-2600) by recording the transmission and absorbance spectra of the films in the range of 320−1400 nm. The photoluminescence (PL) spectrum was recorded at room temperature using a xenon lamp as an excitation source (PL spectrometer, Horiba Fluorolog 3-21). The excitation wavelength was 515 nm for recording the emission spectra. The absorption spectra of the dyes and the dye-adsorbed CuO dispersion in ethanol was recorded with the emission monochromator set at 550 nm. 2.3. Photoelectrochemical, Electrochemical Impedance, and Quantum Efficiency Measurements. The photoelectrochemical (PEC) measurements were carried out in three electrode configuration attached to a potentiostatgalvanostat (CH Instruments, 660D). The edge and back surface sealed CuO electrode was used as working, Pt-gauge as counter electrode (4 cm2), an Ag/AgCl as reference electrode, and aqueous solution containing 0.5 M Na2SO4 (pH 6.8 at 25 °C) was used as electrolyte. The electrochemical impedance spectroscopy (EIS) was carried out between 0.1 Hz−1 MHz at an AC amplitude of 10 mV and various applied DC bias between 0 and 1 V. Mott−Schottky analysis for the CuO photocathode was performed in three electrode configuration to determine its conduction and valence band positions, by applying an AC amplitude of 5 mV and frequency of 1 kHz in the applied potential range of +0.2 to −0.8 V vs Ag/AgCl in an electrolyte containing a mixture of 0.1 M K4[Fe(CN)6], 0.01 M K3[Fe(CN)6], and 1.3 mM KCl (measured pH of 9.2 at 25 °C). A solar simulator (Photo Emission Tech Inc., model SS80) was C

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 3. Linear sweep voltammetry (LSV) scan of mercurochrome dye sensitized CuO electrodes for different time of sensitizations: 24 (a), 48 (b), and 72 h (c) under dark and illuminated conditions.

used to produce AM1.5G (1000 W/m2) radiation. The light intensity at the electrode surface was calibrated by a certified standard silicon reference cell. The PEC cell was deaerated with nitrogen having a flow rate of 50 sccm for 10 min prior to measurement in order to remove any dissolved oxygen. Linear sweep voltammetry was carried out at a cathodic scan rate of 10 mV/s. Internal Photo-Conversion Efficiency (IPCE) of the fabricated PEC cells was characterized using an Optosolar SR 300 spectral response measurement set up. The spectra were acquired using a lock-in amplifier (Stanford Research, SR 830) with a current preamplifier under short-circuit conditions. The cells were illuminated with approximate spot size of 2 mm × 2.5 mm by monochromatic light produced by a double grating monochromator allowing a xenon lamp spectrum. Calibrated silicon photodiode with known spectral response was used for reference spectrum.

microstructures entangle locally with each other as shown in Figure 1b (representing CuO nanostructure at 250 °C) probably due to the dehydration from their bulk. A progressive change in the morphology of the nanostructure above 150 °C can be clearly visible in the FE-SEM images (Figure S2, Supporting Information). The conversion of crystalline Cu(OH)2 into CuO upon annealing was confirmed by the transmission electron microscopy (TEM), where the lattice fringes as well as the selected area electron diffraction (SAED) patterns are shown in figure S3 (Supporting Information). As shown therein, the as-prepared sample was highly crystalline with d-spacing of 2.25 Å, corresponding to the (130) orientation of Cu(OH)2 (inset shows its corresponding spot SAED pattern). However, after annealing at 250 °C, Cu(OH)2 was converted into polycrystalline CuO as indicated by the SAED ring-pattern. The major plane of growth in polycrystalline CuO are found to be (−111) and (111) and confirmed by the XRD studies. The bending of CuO may be attributed to the rotational symmetry among various evolved planes due to the dehydration process. The sharp peaks corresponding to Cu(OH)2 in the as-prepared sample suggest that polycrystalline Cu(OH)2 forms largely beneath the nanorods of CuO. Broadening of the two peaks, (−111) and (111) occurs due to the nanostructures of CuO. The morphologies suggest that the high aspect ratio is favorable to dye-adsorption better than that in polycrystalline form of CuO. 3.2. Dye Sensitization and Optical Characterizations. Figure 2a shows the absorption spectra of the CuO film, sensitized by using ethanol solution of mercurochrome and that of the unsensitized CuO. The broad peak of CuO was centered at around 465 nm. In the presence of mercurochrome solution

3. RESULTS AND DISCUSSION 3.1. Structure and Microstructure of the Nano-CuO. The formation mechanism of the CuO nanostructures through a room temperature alkaline chemical etching on the Cu foil has been well demonstrated in the literatures.24,25 Morphology of the as prepared structure on the Cu foil is randomly oriented nanorods as shown in Figure 1a. The XRD spectra (Figure 1c) show that this structure in predominantly monoclinic CuO with space group C2/c [PDF Card No. 00-005-0661] along with the presence of Cu(OH)2. By annealing the sample in air between 150 and 250 °C the as-prepared nanostructure was converted to pure CuO. However, at 250 °C, the nanostructures start peeling off from the Cu surface and therefore lower annealing temperature was preferred. After annealing above 150 °C, the D

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 4. Proposed schemes of light absorption and electron transfer via carrier-multiplication through dye injection and semiconductor band-edge absorption in the case of (a) photocurrent enhancement by charge collection at Cu and (b) photogenerated electrons reducing a species in electrolyte (such as a proton in the case of water splitting). Here, the band gap of CuO may be referred from the Kubelka−Munk plots as shown in the Supporting Information, Figure S5.

copy (EIS) in a conventional three electrode system. Figure 3 shows the LSV polarization curves of the mercurochrome sensitized CuO nanostructures, recorded in dark as well as illuminated conditions, for various length of time: 0 (unsensitized), 24, 48, and 72 h (denoted as C0, C24, C48 and C72, hereafter). The photocathodes exhibited an increase in photocurrent density by dye sensitization at applied potential of 0 V vs RHE and above. The highest enhancement in photocurrent density from −1.1 mA/cm2 for unsensitized to −2.4 mA/cm2 by the dye sensitization was observed for the photocathode C24 at 0 V vs RHE. The polarization curves indicate that, the dark current rises for dye treatment above 48 h as seen for C72. The increase in the dark current is associated with the development of greater amount of defects in the nanostructure of CuO as supported by the SEM image (Figure S1, Supporting Information). A longer duration of dye sensitization had an adverse effect on the CuO nanostructure morphologies. The CuO nanowires started curling and entangling with each other due to the prolonged dwelling of mercurochrome molecules adsorbed on their surface, possible due to a surface tension force acting between the dye molecules and the nanowires. Because of the loss of straightness of these nanowires, the diffusion length of charge carriers should decrease which eventually increased the dark current. The packing density of CuO nanostructures is reduced for dye sensitization of 48 h or above. The photocurrent enhancement as observed above may be attributed to a carrier multiplication event through anodic dye sensitization in the conduction band of CuO as shown in Figure 4. The photoexcited mercurochrome (D*) injects electrons into the conduction band of CuO to yield the dye cation D+ in the proposed mechanism as shown in eqs(1-3). A back electron transfer to the dye cation may occur if the interfacial charge transfer resistance between CuO and the dye becomes large. Under certain condition of dye adsorption, however, this resistance can be minimized as discussed in the subsequent section. The oxidized dye is further reduced by accepting electron from the electrolyte as provided by the counter electrode. A rapid acceptance of electron from the CuO/electrolyte interfacial Helmoholtz layer by D+ suppresses

an increase in the absorption peak intensity was observed. The absorption band was prominent in the range of 450−550 nm. A Gaussian deconvolution of the absorption spectra in this region produces two peaks, one at 480 nm and the other at λexmax= 517 nm. The latter peak agrees well with other 9-PXD, such as Eosin-Y26and mercurochrome reported earlier.16,17 The former peak, which indicates that an absorption by the dimer to be more intense than the latter.27 A small third peak was observed at 380 nm for both free as well as CuO dispersed dye solutions. The luminescence characteristic of the mercurochrome is demonstrated in the emission spectrum as shown in Figure 2b. The emission maximum is located at λemmax = 587 nm. Figure 2c shows the intersection of the absorption and emission spectra. The [0−0] transition energy of the mercurochrome dye can be estimated from the point of intersection at 2.25 eV (550 nm). The photoluminescence characteristics may be explored in the mercurochrome as well as in CuO by their typical absorption and emission characteristics. In the mercurochrome, the absorption as well as emission are directly related to its excitation and de-excitation characteristics across the HOMO and LUMO levels. On the other hand, CuO is intrinsically a p-type semiconductor by the presence of Cuvacancies (VCu). Therefore, the luminescent characteristics of CuO is determined by the radiative recombination between donor and acceptor level defects including the surface states, as well. The luminescence is however found to be strongly quenched by the presence of CuO nanostructure. This accompanied a Stokes shift of 70 nm (λemmax − λexmax), which is a common phenomenon in other 9-phenylxanthene dyes.16,17,28 The large Stokes shift is sometime advantageous due to the fact that, a smaller Stokes shift (few tens of nanometers) observed in typical fluorophore dyes usually exhibit larger probability of reabsorption of the emitted photon leading to undesired background interference.29 The emission quenching thus provides a clear indication of electron injection from mercurochrome into the conduction band of CuO. 3.3. Photoelectrochemical Responses. The photoelectrochemical properties of the CuO treated by the mercurochrome dye were determined by cathodic linear sweep voltammetry (LSV) and electrochemical impedance spectrosE

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 5. Internal photon to charge conversion efficiency (IPCE) of the dye-sensitized CuO electrodes. Inset shows the maxima of normalized photocurrent in the electrodes corresponding to that contributed by the semiconductor band edge absorption.

Table 1. Internal Photo-Conversion Efficiency (IPCE) of Reported Dye-Sensitized p-Type Photocathodes photocathode

dye

IPCE, % (at given wavelength, nm)

p-NiO

tetrakis(4-carboxyphenyl)porphyrin (TPPC) erythrosin B coumarin dyes, C343 coumarin dyes, N3 mercurochrome

0.24 (550) 3.5 (550) 0.24 (425) 7 (400) 8 (465)

p-CuO

(1)

D* → D+ − e−(1)

(2)

e−(1) → conduction band CuO

(3)

hγ → e−(2) + h+

(4)

10 11 this study

[e−(1) + e−(2)] conduction band CuO → as reducing agent

the recombination of photogenerated holes eq 4. The dyeinjected additional electron along with the photogenerated one in the conduction band of CuO traverse through the bulk of CuO-nanostructure to either reach the Cu-electrode (Figure 4a) or the CuO/electrolyte interface and reduce a redox species, such as protons (Figure 4b) depending upon thermodynamic feasibility. The photogenerated hole eventually oxidizes a redox species whose energy level is more negative with respect to the valence band of CuO to make it thermodynamically feasible. In the present situation, this hole is unable to oxidize water under zero external bias due to the thermodynamically unfavorable CuO valence band position relative to the H2O ↔ O2 redox energy level. hγ + D → D*

ref

(5)

2e− + 2H+ → H 2

(6)

The above set of reaction mechanisms thus suggest that the two electrons, e−(1) and e−(2), can act as reducing agent for two protons eq 6 in the presence of hydrogen evolution reaction (HER) catalyst as the CuO conduction band lies more negative with respect to 0 V RHE. In the absence of HER catalysts, however, hydrogen gas evolution is not feasible, rather the photocurrent enhancement is the only observable phenomena when external biasing arrangement is made. 3.4. Quantum Efficiency, Charge Transfer Kinetics, and Photocurrent. The IPCE action spectra for mercurochrome sensitized nanostructured CuO based PEC cells (using C0, C24, C48, and C72 electrodes) of approximate dimension of 25 μm thickness and 1 cm2 geometric area is shown in Figure 5. A substantial enhancement of photon-toelectron conversion process was noticed for the C24 based PEC cells in the visible region from 400 to 480 nm with a F

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 6. (a, b) Electrochemical impedance spectroscopy of various electrodes: bare Cu, unsesitizedCuO (C0) and CuO dye sensitized for various durations (24, 48 and 72 h, as C24, C48, and C72, respectively). The high frequency region (b) of the overall spectra (a) indicates the kinetics of dye/semiconductor interface. The smallest semicircular branch (and hence a time constant) in 24 h dye-sensitized electrode corresponds to a highest rate of electron injection. (c) Randle’s equivalent circuit representing the transmission line equivalence of various interfaces present.

Figure 5) of unsensitized as well as all dye sensitized photocathodes show peak at 465 nm (2.7 eV), where the corresponding absorption spectra shows a steep rise in the absorbance from 1.5 to about 4.0. As the photon energy corresponding to the photocurrent maximum is independent of dye sensitization, it may be inferred that band edge absorption occurs in the CuO at this point. The effect of the dye is however clearly seen in the relative change in the magnitude of IPCE, which may be attributed to the large extinction coefficient of the dye around 460 nm as described in following paragraphs. The IPCE for a system accompanying three important steps of light harvesting as a function of photon absorption capability (or light harvesting efficiency, LHE in short), charge injection efficiency (or the injection quantum yield, ηinj) and the charge collecting efficiency at the back contact (ηC) may be represented by following equation:16,17

maximum observed IPCE of 8% at 465 nm. It is the highest reported IPCE to date in organic dye sensitized p-type CuO photocathode to the best of our knowledge. The cell without dye sensitization (C0) and those sensitized for longer duration (C48 and C72) produced less photocurrent where IPCE ranges between 2 and 5%. A dye sensitization time longer than 24 h leads to the dye aggregation on the nanostructure surface.30 This was earlier reported in the case of sensitization of TiO2 nanotubes by Ru-complex dye.31 A brief comparison of the IPCE reported in other organic dye sensitized p-type photocathodes reported in the literature is given in Table 1. The IPCE in the longer wavelengths (above 550 nm) indicates the photocurrent generated due to the photon absorption in deep inside the nanostructures. However, a negligibly small magnitude of IPCE in this region demonstrates that the photoconversion takes place very close to the surface of the nanostructures. The onset of the action spectrum of IPCE (lower inset of Figure 5) occurs at 605 nm for the 24 h dyesensitized electrode, indicating the energy gap between the HOMO and LUMO of mercurochrome is approximately 2.0 eV. However, the 0−0 energy gap of mercurochrome in ethanol is estimated to be 2.25 eV as discussed in previous section. This energy difference was previously reported in mercurochrome sensitized TiO2 and it has been attributed to an interaction between the dye molecule and semiconductor surface.16,17 At moderately sensitized CuO (for 24 h), the formation of dimer and/or aggregate at the surface causes injection of electrons from mercurochrome LUMO level into the CuO conduction band easily causing an early onset of IPCE action spectrum. The normalized photocurrent action spectra (upper inset of

IPCE = LHE × ηinjηC

(7)

Here, the LHE may be represented by the following equation:32 LHE = 1 − T = 1 − 10 A

(8)

where, T is the transmittance and A is the absorbance expressed in terms of the logarithm of the incident to transmitted light intensity ratio. As per the reported theory, fraction of the total number of photons entering the bulk of materials undergo absorption to produce exciton (in the case of semiconductor) or causing excitement of the dye, and remaining photons suffer transmission losses.32 As shown in Figure S4 (Supporting Information), the absorbance of the free mercurochrome in G

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C Table 2. Results of the fitting of Nyquist plots of electrodes used in electrochemical impedance spectroscopy fit parameters charge transfer CPE electrode ID

equivalent circuit

series resistance (R0 )/ Ω

resistance (Ω)

exponent, n

admittance, Y (μmhO)

ωmax (Hz)

true capacitance Ceq (μF)

time constant (s)

Cu-bare C0 C24

(Rct,0CPE-0) (R ct,1-CPE-1) (R ct,1-CPE-1) (R ct,2-CPE-2) (R ct,1-CPE-1) (R ct,2-CPE-2) (R ct,1-CPE-1) (R ct,2-CPE-2)

10.9 76.0 108.3

18.5 k 7.9 k 0.1 M 0.9 k 0.1 M 0.95 k 0.1 M 3.0 k

0.73 0.3 0.305 0.908 0.34 0.99 0.312 0.803

90.0 960 530 40 490 34 810 46

− 10.0 10.0 4.64 × 105 11.75 2.1 × 104 10.0 1.43 × 104

− 214.0 107.0 12.0 96.0 30.0 166.0 7.0

− 1.69 10.74 0.01 9.67 0.03 16.6 0.02

C48 C72

40.0 63.2

frequency. In the present case, the charge depletion across the nanostructure interface with the electrolyte produces this CPE element, CPE-1 as depicted in the equivalent circuit in Figure 6c. In the case of the electrode C24, the smaller semicircle originated at high frequency region was due to the charge depletion across the dye−semiconductor interface, whereas, the larger one was due to that across the dyeelectrolyte interface. The later mentioned two depressed semicircles correspond to two CPE elements, CPE-2 and CPE-3 in the equivalent circuit as seen in Figure 6c. The resistors in parallel to the above CPEs represent the charge transfer resistances, Rct1, Rct2, and Rct3 across the abovementioned interface. The resistor in series, Rb represents the bulk resistance of the nanorods. The equivalent circuits fit parameters are listed in Table 2. The estimated parameters of the fits are extremely useful in calculating the charge transfer kinetics across the specific interfaces, which in turn determines the collection efficiency at the back contact and the charge injection efficiency from the dye into the semiconductor provided the time constant of their equivalent R-C network can be found. Here, we refer the time constant of the respective networks as overall time constant, without extracting different contributions from electron diffusion, trapping in the bulk and recombination at the surface states.35 For the case of a “classically” depressed semicircle (CPE in parallel with a resistance) in the equivalent circuit, Hsu and Mansfeld36 have given a method for calculating the “true” capacitance, Ceq, in which the frequency at which the imaginary component reaches a maximum is identified as ωmax and the Ceq can be expressed as

ethanol at a concentration of 0.5 mM in the wavelength range of 460 to 550 nm is around 4.1, making the LHE at a very high value of 99.9%. The molar extinction coefficient of the dye may be calculated from Lambert−Beer’s principle using the relationship:

A = ϵLc

(9)

Here, ϵ, L, and c are the molar extinction coefficient of the dye, length of the light path (in test cuvette) and molar concentration of the dye, respectively. The molar extinction coefficient of mercurochrome thus found to be ca. 8.2 × 103 M−1 cm−1 for L = 1 cm and c = 0.5 mM. In order to estimate the kinetics of charge transfer across various interfaces of the dye sensitized nanostructure of CuO over the Cu-back contact which is immersed in the electrolyte, and subsequently to estimate the injection efficiency (ηinj) and the collection efficiency (ηC), EIS measurement was carried out. Figure 6a shows the Nyquist plots (Z″ vs Z′, where they represent the imaginary and real components of the impedance, respectively) for the bare Cu, unsensitized and dye sensitized nanostructured CuO (C0, and C24) electrodes at zero applied bias in the dark. The bare Cu produces a characteristic similar to the part of a depressed semicircle, indicating charge depletion across the interface with the electrolyte having some associated inhomogeneities on its surface. The plots were consistent with Randles equivalent circuits as shown in Figure 6c.33 The series resistance R0 originates from the metal electrode and the parallel RC develops due to the interface charge layer in the electrolyte. The high frequency limit (Z″ → 0) gives R0 = 10.9 Ω-cm2. The Nyquist plot corresponding to the electrode C0 and C24/48/72, composed of one and two depressed semicircle(s), respectively. The larger branch originates at lower frequency and the smaller one originates at higher frequency as shown in Figure 6, parts a and b, respectively for all the electrodes except the bare Cu. The depression of the semicircle in Z″ vs Z′ is attributed to the existence of constant phase element(s) (CPE). A constant phase element (CPE) represents nonideal capacitor due to the defects or inhomogeneities at the interfaces.34 The impedance of a CPE can be represented by the equation ZCPE =

1 Y (jω)n

Ceq = Y(ωmax )n − 1

(11)

ωmax,Ceq and the time constants, τ1, τ2, and τ3 corresponding to CPE-1, -2, and -3 are listed in Table2. The injection efficiency, ηinj may be determined from the time constants, τ1 (without dye) and the effective time constant τ23 = τ2τ3/(τ2 + τ3) (in the presence of the dye) using the following equation: ⎛ τ ⎞ ηinj = ⎜1 − 23 ⎟ τ1 ⎠ ⎝

(12)

As seen, the time constant for charge transfer from dye to the electrolyte (τ3) was three orders higher than that of dye to semiconductor (τ2). Accordingly, the injection efficiency was estimated ca. 99.73%. Similar procedure may be followed to estimate the time constants and thus the injection efficiencies of other electrodes, viz. C48 and C72. It was found that, in all cases the injection efficiency remains above 99%.

(10)

where n (0 ≤ n ≤ 1) is a CPE exponent, which determines the degree of nonideality of the capacitance (n = 0 and 1 correspond to resistor and capacitor, respectively), Y is a frequency-independent admittance, and ω is the angular H

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C



The integrated photocurrent density under short circuit condition as shown in the LSV scan (Figure 3) is determined by integrating the product of incident photon flux density, F(λ), and the IPCE(λ) at respective wavelength, λ may be expressed as Jsc =

∫ qF(λ) × [1 − r(λ)] × IPCE(λ) dλ

*(A.R.) E-mail: [email protected]. Fax: +91-07923275030. ORCID

Indrajit Mukhopadhyay: 0000-0003-3756-6131 Abhijit Ray: 0000-0002-4285-2717

(13)

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS P.M. gratefully acknowledges financial supports from Council of Scientific and Industrial Research (CSIR), New Delhi via Grant No. 9/1074 (0001)/ 2017-EMR-1 in the form of a Senior Research Fellowship. Dr. D. N. Srivastava is being gratefully acknowledged for facilitating the TEM measurements. Prof. Shinji Kawasaki and Dr. Ishii Yosuke at Nagoya Institute of Technology are thankfully acknowledged for providing the photocurrent measurements at the initial stage of the study.

(14)



where, IQESC is the internal quantum yield of the semiconductor that may be obtained from the IPCE spectra of the unsensitized CuO electrode as shown in Figure 5. As evident from this figure, the IQESC remains below 0.5% between 400 and 700 nm. Therefore, a major contribution to photocurrent density improvement lies in the light harvesting and carrier injection process of the dye. As both the LHE and the ηinj are already found to be sufficiently large, a further improvement in the IPCE and hence in the photocurrent density is possible through an enhancement in ηC which relies on the reduction of bulk resistance in the nanorod.

REFERENCES

(1) Butler, M.; Ginley, D. Principles of photoelectrochemical, solar energy conversion. J. Mater. Sci. 1980, 15, 1−19. (2) Joya, K. S.; Joya, Y. F.; Ocakoglu, K.; van de Krol, R. Watersplitting catalysis and solar fuel devices: Artificial leaves on the move. Angew. Chem., Int. Ed. 2013, 52, 10426−10437. (3) Morris, A. J.; Meyer, G. J.; Fujita, E. Molecular approaches to the photocatalytic reduction of carbon dioxide for solar fuels. Acc. Chem. Res. 2009, 42, 1983−1994. (4) Rehman, S.; Ullah, R.; Butt, A.; Gohar, N. Strategies of making TiO2 and ZnO visible light active. J. Hazard. Mater. 2009, 170, 560− 569. (5) Park, J. H.; Kim, S.; Bard, A. J. Novel carbon-doped TiO2 nanotube arrays with high aspect ratios for efficient solar water splitting. Nano Lett. 2006, 6, 24−28. (6) Miyauchi, M.; Takashio, M.; Tobimatsu, H. Photocatalytic activity of SrTiO3 codoped with nitrogen and lanthanum under visible light illumination. Langmuir 2004, 20, 232−236. (7) Su, W.; Zhang, Y.; Li, Z.; Wu, L.; Wang, X.; Li, J.; Fu, X. Multivalency iodine doped TiO2: preparation, characterization, theoretical studies, and visible-light photocatalysis. Langmuir 2008, 24, 3422−3428. (8) Youngblood, W. J.; Lee, S.-H. A.; Maeda, K.; Mallouk, T. E. Visible light water splitting using dye-sensitized oxide semiconductors. Acc. Chem. Res. 2009, 42, 1966−1973. (9) Kawai, T.; Sakata, T. Photocatalytic hydrogen production from liquid methanol and water. J. Chem. Soc., Chem. Commun. 1980, 15, 694−695. (10) He, J.; Lindström, H.; Hagfeldt, A.; Lindquist, S.-E. Dyesensitized nanostructured p-type nickel oxide film as a photocathode for a solar cell. J. Phys. Chem. B 1999, 103, 8940−8943. (11) Qin, P.; Zhu, H.; Edvinsson, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. Design of an organic chromophore for p-type dye-sensitized solar cells. J. Am. Chem. Soc. 2008, 130, 8570−8571. (12) Langmar, O.; Ganivet, C. R.; Lennert, A.; Costa, R. D.; De La Torre, G.; Torres, T.; Guldi, D. M. Combining Electron-Accepting Phthalocyanines and Nanorod-like CuO Electrodes for p-Type DyeSensitized Solar Cells. Angew. Chem. 2015, 127, 7798−7802. (13) Jiang, T.; Bujoli-Doeuff, M.; Farré, Y.; Pellegrin, Y.; Gautron, E.; Boujtita, M.; Cario, L.; Jobic, S.; Odobel, F. CuO nanomaterials for ptype dye-sensitized solar cells. RSC Adv. 2016, 6, 112765−112770. (14) Tennakone, K.; Kumarasinghe, A.; Sirimanne, P. Dye sensitization of low-bandgap semiconductor electrodes: cuprous oxide photocathode sensitized with methyl violet. Semicond. Sci. Technol. 1993, 8, 1557.

4. CONCLUSIONS Mercurochrome, one of the organic (chromophore) dyes explored for the first time to sensitize CuO, exhibits better efficiency in a typical PEC application through a combined effect of carrier injection by the dye and photogeneration in the semiconductor. The photocurrent density is found to enhance from −1 mA/cm2 (unsensitized) to −2.2 mA/cm2 (dye sensitized for 24 h) at 0 V vs RHE under AM1.5G simulated solar spectrum in a PEC cell of configuration: CuO (1 cm2) | aqueous 0.5 M Na2SO4(pH 6.8) electrolyte| Pt-mesh (4 cm2). A strong light harvesting capability of the dye molecules (LHE > 99.9%) and a fast electron injection kinetics from the dye to semiconductor (injection efficiency >99% and a time constant of 10 ms) were confirmed by the photoluminesence and impedance spectroscopic analyses, respectively. The photoluminescence spectra demonstrated a strong absorption at 465 nm by the mercurochrome, where an IPCE of 8% was recorded. This is so far, to the best of our knowledge, is the highest reported quantum yield from p-type CuO based photocathode.



AUTHOR INFORMATION

Corresponding Author

where q is the elementary charge, and r(λ) is the optical loss factor due to reflection. In the case of conventional dye sensitized solar cells using wide band gap semiconductors such as TiO2, the visible light absorption occurs mainly at the photosensitive dye molecules, which in turn governs the IPCE as defined by eq 7. However, when a low band gap semiconductor is used, simultaneous absorption of visible photon occurs at the semiconductor bulk. This brings a separate contribution to the IPCE and eq 7 should be modified as, IPCE = LHE × ηinjηC × IQESC

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.7b10024. FE-SEM characterization showing morphological variation in nanostructured CuO by mercurochrome dye sensitization time, the absorbance spectrum of mercurochrome, and Kulbelka−Munk plots (PDF) I

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

constant phase element. J. Electroanal. Chem. Interfacial Electrochem. 1984, 176, 275−295. (35) Bisquert, J.; Vikhrenko, V. S. Interpretation of the time constants measured by kinetic techniques in nanostructured semiconductor electrodes and dye-sensitized solar cells. J. Phys. Chem. B 2004, 108, 2313−2322. (36) Hsu, C.; Mansfeld, F. Concerning the conversion of the constant phase element parameter Y0 into a capacitance. Corrosion 2001, 57, 747−748.

(15) Gonce, M. K.; Aslan, E.; Ozel, F.; Hatay Patir, I. Dye-Sensitized Cu2XSnS4 (X= Zn, Ni, Fe, Co, and Mn) Nanofibers for Efficient Photocatalytic Hydrogen Evolution. ChemSusChem 2016, 9, 600−605. (16) Kim, K. E.; Jang, S.-R.; Park, J.; Vittal, R.; Kim, K.-J. Enhancement in the performance of dye-sensitized solar cells containing ZnO-covered TiO2 electrodes prepared by thermal chemical vapor deposition. Sol. Energy Mater. Sol. Cells 2007, 91, 366−370. (17) Hara, K.; Horiguchi, T.; Kinoshita, T.; Sayama, K.; Sugihara, H.; Arakawa, H. Highly efficient photon-to-electron conversion of mercurochrome-sensitized nanoporous ZnO solar cells. Chem. Lett. 2000, 29, 316−317. (18) Zhang, Z.; Wang, P. Highly stable copper oxide composite as an effective photocathode for water splitting via a facile electrochemical synthesis strategy. J. Mater. Chem. 2012, 22, 2456−2464. (19) Barreca, D.; Fornasiero, P.; Gasparotto, A.; Gombac, V.; Maccato, C.; Montini, T.; Tondello, E. The potential of supported Cu2O and CuO nanosystems in photocatalytic H2 production. ChemSusChem 2009, 2, 230−233. (20) Anandan, S.; Wen, X.; Yang, S. Room temperature growth of CuO nanorod arrays on copper and their application as a cathode in dye-sensitized solar cells. Mater. Chem. Phys. 2005, 93, 35−40. (21) Sumikura, S.; Mori, S.; Shimizu, S.; Usami, H.; Suzuki, E. Syntheses of NiO nanoporous films using nonionic triblock copolymer templates and their application to photo-cathodes of p-type dye-sensitized solar cells. J. Photochem. Photobiol., A 2008, 199, 1−7. (22) Wei, A.; Xiong, L.; Sun, L.; Liu, Y.-J.; Li, W.-W. CuO nanoparticle modified ZnO nanorods with improved photocatalytic activity. Chin. Phys. Lett. 2013, 30, 046202. (23) Liu, S.; Tian, J.; Wang, L.; Luo, Y.; Sun, X. One-pot synthesis of CuO nanoflower-decorated reduced graphene oxide and its application to photocatalytic degradation of dyes. Catal. Sci. Technol. 2012, 2, 339−344. (24) Ray, A.; Mukhopadhyay, I.; Pati, R.; Hattori, Y.; Prakash, U.; Ishii, Y.; Kawasaki, S. Optimization of photoelectrochemical performance in chemical bath deposited nanostructured CuO. J. Alloys Compd. 2017, 695, 3655−3665. (25) Liu, J.; Huang, X.; Li, Y.; Sulieman, K.; He, X.; Sun, F. Hierarchical nanostructures of cupric oxide on a copper substrate: controllable morphology and wettability. J. Mater. Chem. 2006, 16, 4427−4434. (26) Moser, J.; Graetzel, M. Photosensitized electron injection in colloidal semiconductors. J. Am. Chem. Soc. 1984, 106, 6557−6564. (27) Rohatgi, K.; Singhal, G. Nature of bonding in dye aggregates. J. Phys. Chem. 1966, 70, 1695−1701. (28) Hilgendorff, M.; Sundström, V. Ultrafast electron injection and recombination dynamics of dye sensitised TiO2 particles. Chem. Phys. Lett. 1998, 287, 709−713. (29) Gao, Z.; Hao, Y.; Zheng, M.; Chen, Y. A fluorescent dye with large Stokes shift and high stability: synthesis and application to live cell imaging. RSC Adv. 2017, 7, 7604−7609. (30) Nelson, R. Minority Carrier Trapping and Dye Sensitization. J. Phys. Chem. 1965, 69, 714−718. (31) Ghicov, A.; Albu, S. P.; Hahn, R.; Kim, D.; Stergiopoulos, T.; Kunze, J.; Schiller, C. A.; Falaras, P.; Schmuki, P. TiO2 Nanotubes in Dye-Sensitized Solar Cells: Critical Factors for the Conversion Efficiency. Chem. - Asian J. 2009, 4, 520−525. (32) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphry-Baker, R.; Müller, E.; Liska, P.; Vlachopoulos, N.; Grätzel, M. Conversion of light to electricity by cis-X2bis (2, 2′-bipyridyl-4, 4′-dicarboxylate) ruthenium (II) charge-transfer sensitizers (X= Cl−, Br−, I−, CN−, and SCN−) on nanocrystalline titanium dioxide electrodes. J. Am. Chem. Soc. 1993, 115, 6382−6390. (33) Macdonald, J. R.; Johnson, W. B. In Impedance spectroscopy emphasizing solid materials and systems; MacDonald, J. R., Ed.; Wiley Interscience: New York, 1987. (34) Brug, G.; Van Den Eeden, A.; Sluyters-Rehbach, M.; Sluyters, J. The analysis of electrode impedances complicated by the presence of a J

DOI: 10.1021/acs.jpcc.7b10024 J. Phys. Chem. C XXXX, XXX, XXX−XXX