Enhanced photoelectrochemical performance of novel p-type MoBiCuSe₄ thin films deposited by a simple surfactant-mediated solution route

Abstract

Low-dimensional nanostructures with reduced grain boundaries show superior charge transportation in a photoelectrochemical cell. Therefore, nanostructures of MoBiCuSe₄ thin films deposited using different surfactants are expected to be favorable for providing a direct pathway for smooth transport of photogenerated charge carriers across a reduced number of grain boundaries within the photoelectrode. In the present investigation, we have studied the effect of different surfactants, such as polyethylene glycol (PEG), sodium dodecylsulfate (SDS) and trioctylphosphine oxide (TOPO), on the opto-structural, morphological and photoelectrochemical (PEC) properties of MoBiCuSe₄ thin films. We have demonstrated a soft chemical route that facilitates the formation of a compact, homogeneous deposition with a large effective (photoactive) surface area, which could be suitable for PEC cells. The MoBiCuSe₄ thin films have been deposited using the arrested precipitation technique (APT) and their formation confirmed by energy dispersive X-ray spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS). The nanocrystalline nature of the MoBiCuSe₄ thin films and the mixed rhombohedral crystal structure with reduced number of grain boundaries were confirmed by the X-ray diffraction (XRD) pattern. The direct allowed type of transition in the material, with an average absorption coefficient above 10⁴ cm⁻¹, makes it suitable for PEC applications. The maximum light conversion efficiency achieved for MoBiCuSe₄ thin films deposited with surfactant is 0.18%. PEC analysis verifies that the synthesized nanostructures of the surfactant-assisted MoBiCuSe₄ photoelectrode material are suitable for PEC cells.

---

Introduction

The morphology and size-dependent properties of metal chalcogenides have attracted intense attention in materials science owing to their crucial role in significant applications.^1,2 The unique structures of semiconducting chalcogenide materials could lead to distinct properties and applications in optoelectronics, solar cells, gas sensing and as a photocatalytic material.^3–6 However, to have control over the morphology and properties of chalcogenides, surfactant-mediated chemical methods have elicited great interest due to their significant advantages, such as controlled size, low-temperature growth, cost effectiveness and less complicated nature.^7,8 Surfactants are a class of molecules that form thermodynamically stable aggregates of inherently nanoscale dimensions, both in solution and at interfaces. In solution, surfactant self-assembly has been investigated both theoretically⁹ and experimentally¹⁰ to some extent because of the significance of such ordered templates in the synthesis of nanometer to micron scale structures with controlled dimensions.^11,12 Recently, it has been reported that anionic, cationic and non-ionic surfactants can be used to assist the formation of different nanostructures of chalcogenides, oxides and inorganic nanoparticles by giving control over the morphology attained by surfactant assembly.^13–18 Surfactant templating has provided a breakthrough in methodology for the preparation of a number of semiconducting materials having a variety of structures with diverse tunable properties.^12,19

Earlier, we have reported the deposition of novel mixed metal chalcogenide thin films of MoBi_2−xCu_xSe₄ by a facile chemical method for photoelectrochemical (PEC) applications and we have reported the maximum light conversion efficiency of 0.074% for MoBiCuSe₄ thin films deposited without surfactant.²⁰ Herein, we report the deposition of MoBiCuSe₄ thin films by a simple surfactant-mediated solution route. The surface directing agents, such as PEG, SDS and TOPO, were used to obtain different nanostructures of MoBiCuSe₄. The films were obtained just by keeping the deposition baths at room temperature for several hours without stirring. The structural, morphological, optical and PEC properties of the MoBiCuSe₄ thin films were studied as a function of the different surfactants. We have demonstrated a soft chemical route that facilitated the formation of a compact, defect-free deposit with a large effective (photoactive) surface area to improve PEC cell performance.

Experimental

MoBiCuSe₄ thin films were deposited using three different surfactants viz. sodium dodecyl sulphate (SDS) [NaC₁₂H₂₅SO₄], polyethylene glycol (PEG) [C_2nH_4n+2O_n+1] and trioctylphosphine oxide (TOPO) [C₂₄H₅₁OP]. Ammonium molybdate tetrahydrate [(NH₄)₆Mo₇O₂₄·4H₂O], bismuth nitrate pentahydrate [Bi(NO₃)₃·5H₂O], copper acetate monohydrate [(CH₃COO)₂Cu·H₂O], selenium metal powder, sodium sulphite [Na₂SO₃], aqueous ammonia (NH₃) and triethanolamine (TEA) [N(CH₂–CH₂–OH)₃] were used as precursors and complexing agents. The glass substrates and fluorine-doped tin oxide (FTO)-coated glass substrates used for deposition were ultrasonically cleaned using detergent followed by chromic acid/methanol treatment and finally cleaned with acetone. Herein, the MoBiCuSe₄ thin films were deposited using a modified chemical bath deposition technique, i.e. the arrested precipitation technique (APT), using different surfactant media.

The optimized experimental conditions for deposition of surfactant-mediated MoBiCuSe₄ thin films were kept fixed as per our previous report.²⁰ Briefly, a bath containing 1 ml 0.05 M Mo–TEA complex solution, 5 ml 0.05 M Bi–TEA complex solution and 10 ml 0.05 M Cu–TEA complex solution was taken, as a source of Mo⁴⁺, Bi³⁺ and Cu⁺ ions, and the pH of the solution was adjusted to 10 by adding aqueous ammonia solution. Subsequently, 10 ml 0.25 M Na₂SeSO₃ solution was added with continuous stirring as a source of Se²⁻ ions. The total volume of the bath was made up to 40 ml by adding double-distilled water and designated as the 1^st bath, without surfactant. Preparation of a 0.25 M Na₂SeSO₃ solution is the same as described in our earlier report.²⁰ Furthermore, in order to deposit MoBiCuSe₄ thin films with different nanostructures and tuned surface morphologies, we have used PEG, SDS and TOPO as surfactants. In a typical experiment, 1 ml 0.05% solutions of PEG, SDS and TOPO (in pure ethanol) were added separately into three (40 ml) baths prepared same as that of the bath 1^st and they were designated as bath 2^nd, 3^rd and 4^th, respectively. For deposition, well-cleaned FTO-coated glass substrates were immersed in the four baths and kept at room temperature (300 K) for 8 h. After the desired deposition time, the films were rinsed with double-distilled water and dried at room temperature. In order to remove surfactants and to improve the crystallinity, the as-deposited samples were annealed at 130 °C for 3 h. The MoBiCuSe₄ thin film deposited without surfactant is denoted by MBCS and those deposited using different surfactant media, namely PEG, SDS and TOPO, are denoted by MBCS:PEG, MBCS:SDS and MBCS:TOPO, respectively, and used for further characterization.

The surface morphology and elemental analysis of all the samples were examined using a scanning electron microscopy (SEM) equipped with an energy dispersive spectroscopy (EDS) analyzer (JEOL-JSM-6360A). Inferences about the oxidation states of molybdenum, bismuth, copper and selenium elements in the chemically deposited MoBiCuSe₄ thin films were analyzed by X-ray photoelectron spectroscopy (XPS) (VG Multilab 2000-Thermo Scientific, USA, K-α) with a multi-channel detector, which can endure high photonic energies from 0.1 to 3.0 keV. The X-ray diffraction (XRD) patterns of the MoBiCuSe₄ thin films were recorded using an X-ray diffractometer (Bruker AXS, D8 Model) using Cu Kα (λ = 1.5418 Å) radiation. A UV-vis-NIR spectrophotometer (model: Shimadzu UV-1800) was used to measure the optical absorption of the thin films in the wavelength range 400–1100 nm. PEC performance was recorded using a semiconductor characterization system (Keithely, SCS-4200 Germany) with a two electrode configuration in the dark and under illumination (30 mW cm⁻²). A glass-FTO/MoBiCuSe₄/I⁻/I₃⁻/G cell configuration was used to record J–V plots. The MoBiCuSe₄ film and graphite rod (G) were employed as the working and counter electrodes, respectively. A 0.5 M aqueous solution of iodide–polyiodide (I⁻–I₃⁻) was used as the redox electrolyte. The charge recombination properties and electron transport resistance of the MoBiCuSe₄ thin films were investigated by electrochemical impedance measurements (model: AUTOLAB PGSTAT100 potentiostat) in the frequency range 0.1–10⁵ Hz.

Results and discussion

Growth mechanism and kinetics

Surfactants are used as morphology-directing agents to obtain low-dimensional nanostructures due to their ability to form thermodynamically stable aggregates of nanoscale dimensions.²¹ Different surfactants were found to have a specific mechanism involved in the synthesis of nanostructures.⁹ During the synthesis process, the surfactants were adsorbed by the growing crystal and depending on their properties, they moderate the growth rate of the crystal faces, which helps to control the size and shape of the crystal faces.^8,22 Surfactants do not directly take part in the reaction. They act as surface active agents. Fig. 1 presents the structures of the PEG, SDS and TOPO surfactant molecules.

Fig. 1 Chemical structures of (i) polyethylene glycol (PEG), (ii) sodium dodecyl sulfate (SDS) and (iii) trioctylphosphine oxide (TOPO).

PEG is a polymer compound of ethylene oxide with a molecular weight of 20 000 g mol⁻¹. SDS is an organosulphate compound consisting of a 12-carbon tail attached to the sulphate functional group. TOPO is an organophosphorus compound with uncharged alkyl chains. It also has polarity resulting from the dipolar phosphorus–oxygen (P–O) bond. The chemical reaction kinetics for the formation of MoBiCuSe₄ thin films has been briefly discussed in our previous report.²⁰ The possible growth mechanism of surfactant-mediated MoBiCuSe₄ thin film nanostructures using different surfactants is shown in Fig. 2.

Fig. 2 Possible growth mechanism of surfactant-mediated MoBiCuSe₄ thin film nanostructures using PEG, SDS and TOPO.

PEG does not ionize in water and has good stability, which leads to a better tuning of surface morphology. The non-ionic surfactant PEG controls the growth rate of different faces by joining selectively to certain crystal faces.²¹ Initially, during the chemical growth, seeds are formed which act as nucleation centers for the growth of nanoparticles. The hydrophilic groups in the polymeric chain of PEG may get adsorbed on the surface of the nanoparticles in order to reduce their surface energy, preventing quick aggregation of the particles.²³ As a result, anisotropic growth occurred, leading to nanostructured film formation via oriented attachment.⁸ PEG plays an important role as a template in optimizing the growth rate and orientation in the formation of crystallites.²⁴ The different low-dimensional morphologies evident from the morphological analysis indicate that the involvement of surface active agents in the synthesis process is a simple and effective way to control the nucleation and growth process of MoBiCuSe₄ nanostructures.

It is well known that SDS acts as a soft template.^25–27 The surfactant molecules self-assemble into aggregates in aqueous solution above a critical concentration, called the micelle concentration, which is the concentration above which surfactant molecules arrange themselves in a spherical form in aqueous solution. In the process of chemical deposition, a group of freestanding crystallites with irregular sizes in the non-equilibrium state were formed, which ultimately dissolve in the solution followed by nucleation growth to form nanoparticles. Under prolonged reaction conditions, nanoparticles self-assemble and the process of dissolution–recrystallization takes place, leading to the formation of different nanostructures depending on the reaction conditions.²⁸ The nanostructures formed through this approach grow mainly by the Ostwald ripening mechanism; nanostructures of larger size grow at the cost of smaller ones due to higher surface free energy.²⁹ Particle formation is a complex process, involving nucleation, growth, coagulation and flocculation, which may be influenced significantly by SDS additive assemblies. Therefore, SDS plays an important role in the preparation of chalcogenide nanoclusters.

The cationic TOPO has polarity resulting from the dipolar P–O bond. Also, the structure of TOPO is responsible for its excellent complexing properties due to the presence of the phosphine group (P=O). It acts as a surface active amphiphilic molecule containing a hydrophilic head and hydrophobic tail. It also contains a lone pair of electrons, which is used for the formation of complexes with metal ions. Thus, TOPO plays a dual role, as a complexing agent as well as a surfactant in the reaction bath. Due to this dual role performed by TOPO, the rate of reaction and deposition may lower, which results in ordered and nanostructured growth of thin films.^30,31 Notable differences in the morphology of the MoBiCuSe₄ were observed due to the different surfactants; hence, it can be predicted that the PEC performance could be enhanced significantly.

Morphological study

A detailed surface analysis of the deposited structures was carried out using SEM. Fig. 3 shows the SEM images of MoBiCuSe₄ thin films prepared with and without surfactants. Fig. 3a shows the lower magnification SEM image of the MoBiCuSe₄ thin film deposited without surfactant (MBCS). It displays a petal-like morphology with sizes varying from 50 to 90 nm, with a 16 nm average thickness (Fig. 3b). The SEM images of the surfactant-mediated MoBiCuSe₄ thin film samples MBCS:PEG, MBCS:SDS and MBCS:TOPO shown in Fig. 3c to h indicates the uniform coverage over the substrate surface. The lower magnification SEM image shown in Fig. 3c clearly reveals that a nanofibrous hollow-sphere-like morphology is formed for the MBCS:PEG sample. Numerous nanofibers (50 nm thick and 500 nm long) are spherically assembled in such a manner that a hollow cavity of 200 nm diameter was formed and is clearly observed in the high magnification SEM image shown in Fig. 3d. Dong et al.²⁴ reported the hollow nanospherical morphology of ZnS using PEG as a surfactant; they found that the presence of PEG is necessary for the generation of ZnS hollow nanospheres. Li et al.²³ also reported the formation of uniform flower-like ZnO microspheres of 12–25 nm, which are formed by assembled 2D nanosheets, using PEG as a surfactant. Randomly oriented, vertically aligned nano-platelets were observed for the SDS-mediated MoBiCuSe₄ thin films (MBCS:SDS) (Fig. 3e). The high magnification SEM image of the MBCS-SDS thin film (Fig. 3f) reveals that the platelets, with an average length of 875 nm and thickness of 85 nm are diffused to each other, which results in the formation of porous nanostructures. SDS is a peculiar surfactant, which acts like a template and catalyzes the self-assembled growth of lamellar nanostructures.³² A marigold flower-like morphology was exhibited by TOPO-mediated MoBiCuSe₄ thin films (MBCS:TOPO, Fig. 3g). Flower sizes of about 925 nm with 45 nm sized petals are clearly seen in the high magnification SEM image shown in Fig. 3h. This observation leads to a clear conclusion that the surface adsorption of surfactant molecules on the growing MoBiCuSe₄ surface significantly modifies crystal growth. Surfactants, viz. PEG, SDS and TOPO, act like a template during deposition, altering the way the MoBiCuSe₄ thin films grow on the substrate surface. The SEM images of all three samples reveal that the nanostructures grew compactly with a uniform layer of thin film. Such a growth of nanostructures will provide not only a larger effective surface area for light absorption but also a smooth path for the movement and transportation of charge carriers in-between the crystallites.

Fig. 3 SEM images of MoBiCuSe₄ thin films deposited without and with surfactant: (a and b) MBCS, (c and d) MBCS-PEG, (e and f) MBCS:SDS, and (g and h) MBCS:TOPO.

Compositional study

Quantitative and qualitative analyses of the prepared MoBiCuSe₄ thin films were carried out using EDS and XPS techniques to determine the atomic percentage of Mo, Bi, Cu and Se elements in the film as well as to confirm their valence states. In our earlier report²⁰ we have discussed the EDS analysis of MoBiCuSe₄ thin films. An increased atomic percentage of Bi and decreased atomic percentage of Se relative to the theoretically expected values were observed, which was explained with appropriate reasons. Indeed, the precise compositional control of multinary compounds remains challenging. The deviation of product stoichiometry from the desired metal ratio was widely observed for multinary chalcogenide thin films,^33,34 which was likely attributed to different reactivities of the metal ions.

From the XPS analysis data, the peaks of the XPS binding energies were deconvoluted with Gaussian peak shapes using the Origin software package. The typical XPS survey spectrum of the MoBiCuSe₄ thin film (MBCS) shown in Fig. 4a reveals that all the peaks are assigned to C, O, Mo, Bi, Cu and Se elements. The binding energy of the C 1s transition was used as a reference at 284.60 eV to standardize the binding energy of the other elements. It should be mentioned that the presence of trace amounts of C and O on the surface of the thin films is infrequent, and could be due to the adsorption of CO₂, O₂ or H₂O on the film surface.^35,36 An overlapping of peaks (Bi 4f and Se 3p) is observed, but the intensities of the Bi 4f peaks are higher than that of the Se 3p transitions, which indicated the deficiency of selenium in the film. Moreover, it is known that Se atoms escape easily during the ablation process, thus resulting in Bi-rich MoBiCuSe₄ thin films. The high resolution core spectrum for Mo 3d is shown in Fig. 4b. The two intense peaks of Mo 3d3/2 and Mo 3d5/2 at 230.82 and 228.23 eV binding energies indicate the +4 oxidation state of the Mo ions. The high resolution core spectrum for Bi 4f is presented in Fig. 4c and shows that peaks appear at 158.00 eV (Bi 4f7/2) and 163.27 eV (Bi 4f5/2) binding energies with a splitting of 5.27 eV, indicating the +3 oxidation state of the Bi ions, which is in good agreement with previous reports.^36,37 The core level spectrum of Cu 2p is shown in Fig. 4d; the two narrow peaks at 931.34 eV (Cu 2p3/2) and 951.26 eV (Cu 2p1/2) binding energies with a peak separation of 19.92 eV indicate the +1 oxidation state of the copper ions. Moreover, the characteristic peak of Cu²⁺ (942 eV) cannot be observed, which implies that only the valence state Cu(I) exists in the film.^38–41 The Se 3d spectrum (Fig. 4e) can be fitted perfectly using a single peak for 3d3/2 and 3d5/2 and a binding energy of about 53.31 eV. This means that the Se exists in the deposited films as the MoBiCuSe₄ compound. No evidence of the contribution of elemental Se atoms at a binding energy of about 54.1 eV was observed.⁴¹ A distinctive peak at around 58 eV binding energy appeared in the core level spectrum of Se 3d due to the exposure of the sample to the air.⁴²

Fig. 4 Typical XPS spectra of the MoBiCuSe₄ thin film (MBCS): (a) XPS survey spectrum, (b) high resolution core spectrum in the molybdenum region (Mo 3d), (c) high resolution core spectrum in the bismuth region (Bi 4f), (d) high resolution core spectrum in the copper region (Cu 2p) and (e) high resolution core spectrum in the selenium region (Se 3d).

Structural study

The structural characterization was carried out via X-ray diffraction (XRD) patterns, showing the phases present, the phase concentrations and the crystallite size^43,44 of the deposited material. The XRD patterns of the MoBiCuSe₄ thin films deposited without and with surfactant are shown in Fig. 5. The observed peaks corresponding to the (012), (110), (021), (205) and (0015) planes were indexed to the rhombohedral MoSe₂ phase, whereas the (110), (104), (107), (110), (120), (231), (119), (020), (220), (0021) and (125) planes were indexed to the rhombohedral Bi₂Se₃ phase, and the (220), (131), (002), (012) and (802) planes were indexed to the orthorhombic Cu₂Se phase (JCPDS file no. 20-0757, 33-0214 and 19-0401, respectively). The experimental d-values for the MoBiCuSe₄ thin films deposited with and without surfactant were calculated from the Bragg’s relation.⁴³ The observed interplaner distance ‘d’ was compared with the standard JCPDS ‘d’ values. The observed d-values are in good agreement with those of the standard values. No significant difference was observed between the XRD patterns of the MoBiCuSe₄ thin films deposited without surfactant and with different surfactants. Only the intensity of the diffraction peaks in the XRD patterns of the MoBiCuSe₄ thin films deposited with different surfactants was slightly increased. This indicates that the addition of surfactants does not affect the crystallographic orientation of the MoBiCuSe₄ thin films. All samples exhibit a prominent and intense diffraction peak at 2θ, 27.95° which corresponds to the (104) plane of the Bi₂Se₃ phase and the (006) plane of the MoSe₂ phase, indicating the predominant growth of crystallites in these directions.

Fig. 5 XRD patterns of MoBiCuSe₄ thin films deposited without and with surfactant.

The average crystallite size (D) was calculated from the broadening of the diffraction peak according to the well-known Scherrer equation;⁴⁴

where ‘K’ is a dimensionless constant (0.94). The constant K takes on different values depending on the conversion; in particular it is affected by the grain shape, the grain size distribution and how the peak width is defined. Typically it is given by the value

This value is appropriate for spherical crystals with cubic symmetry, where the peak width is defined using the FWHM. If one can use an integral breadth then the constant would be 0.89. Here, the peak width is defined using the FWHM, hence the value 0.94 is taken for K. In general, the constant K varies between 0.62 and 2.08. ‘β’ is the corrected broadening of the diffraction peak measured at half of its maximum (FWHM) intensity (taken in radians by multiplying a factor of π/180), ‘λ’ is the wavelength of X-rays used (1.5406 Å) and ‘θ’ is the diffraction angle. The crystallite size calculated for all MoBiCuSe₄ thin film samples deposited without and with surfactant is summarized in Table 1. As can be seen from the diffraction pattern, improved crystallinity is observed for samples MBCS:SDS, MBCS:PEG and MBCS:TOPO, which can be attributed to the use of different surfactant media, which leads to ordered growth of crystallites and the subsequent Ostwald ripening process.

Table 1 Variation of crystallite size (D), dislocation density (δ) and microstrain (ε) of MoBiCuSe₄ thin films deposited without and with surfactant

Sample code	Crystallite size (D) nm	Dislocation density (δ) × 10^−3 lines per nm^2	Microstrain (ε) × 10^−3
MBCS	89	0.13	1.84
MBCS:PEG	98	0.10	1.76
MBCS:SDS	91	0.12	1.81
MBCS:TOPO	95	0.11	1.80

---

From the calculated values of crystallite size, the dislocation density (δ) (lines per m²) and microstrain (ε) of all the samples were determined by using eqn (2) and (3) and listed in Table 1.

where D is the crystallite size, θ is Bragg’s diffraction angle, λ is the X-ray wavelength used and β is FWHM. As compared to sample MBCS, samples MBCS:SDS, MBCS:PEG and MBCS:TOPO show lower values of δ and ε. This is attributed to decreased defect levels and grain boundaries due to increased crystallite size. The decrease in microstrain and dislocation density indicates the presence of a lower number of lattice imperfections and the formation of higher quality films using different surfactants.

Optical absorption study

From the absorption spectra (Fig. 6a and b), the corresponding band gaps of the MBCS, MBCS:PEG, MBCS:SDS and MBCS:TOPO samples were determined by applying the Tauc’s law^45,46 given by,

αhν = A(hν − E_g)ⁿ (4)

where α is the absorption coefficient, which has been obtained directly from the absorbance against wavelength curves using the relation,

Fig. 6 (a) Plot of absorption (a. u.) vs. wavelength (λ, nm) of MoBiCuSe₄ thin films deposited without surfactant, with plot of (αhν)² vs. (hν) shown in inset. (b) Plots of absorption (a. u.) vs. wavelength (λ, nm) of MoBiCuSe₄ thin films deposited with surfactant, with plots of (αhν)² vs. (hν) shown in inset.

ν is the frequency of the incident beam (ω = 2πν), A is a constant, E_g is the optical band gap and n is an exponent, which can be assumed to have values of 1/2, 3/2, 2 and 3 depending on the nature of the electronic transition responsible for the absorption: n = 1/2 for allowed direct transition, n = 3/2 for forbidden direct transition, n = 2 for allowed indirect transition and n = 3 for forbidden indirect transition.⁴⁷ The optical transitions in transition metal chalcogenides are found to be the direct allowed type.^48,49 Hence, plots of (αhν)² vs. hν shown in the insets of Fig. 6a and b are drown and the optical band gap energy has been estimated by extrapolating the straight line portion to cut the energy axis. A prominent feature that can be observed from the value of the absorption coefficient is found in the range of 10⁴ cm⁻¹, supporting the direct band gap nature of the semiconductor material,^50,51 which is in good agreement with an earlier report.²⁰

The MoBiCuSe₄ thin film deposited without surfactant (MBCS) has a band gap energy of 1.26 eV, while the MBCS:PEG, MBCS:SDS and MBCS:TOPO samples exhibit energies of 1.47, 1.42 and 1.37 eV, respectively. A slight difference in the absorption edge observed for all the samples results in a small difference in their band gap. The difference in the absorption edge occurred due to strengthening of the quantum confinement in these nanostructures, where the Fermi level becomes discrete by eliminating its consecutive nature.^50,52,53 The optical absorption study results reveal that the morphologically different nanostructures formed using different surfactants have suitable band gaps for solar cell applications.⁵²

Photoelectrochemical (PEC) study

Fig. 7 shows the J–V curves for the MBCS, MBCS:PEG, MBCS:SDS and MBCS:TOPO samples. For all the samples, the cathode photocurrent was found to increase as the cathode potential shifted in the negative direction, which is a characteristic of p-type semiconductors.^20,53–56

Fig. 7 Photocurrent–voltage (J–V) curves of MoBiCuSe₄ thin films deposited without and with surfactant in 0.5 M I⁻/I₃⁻ redox electrolyte in the dark and under illumination: (a) MBCS, (b) MBCS:PEG, (c) MBCS:SDS and (d) MBCS:TOPO.

The output parameters of the PEC solar cell, i.e. the fill factor (FF) and conversion efficiency (η%), are calculated from the relations (6) and (7) respectively,⁵⁷

where P_in is the input light intensity, J_sc is the short circuit current density and V_oc is the open circuit voltage. J_max and V_max are the values of maximum current and maximum voltage, respectively, which can be extracted from an output characteristic of the PEC cell. The series resistance (R_s) and shunt resistance (R_sh) of all three samples were estimated from the inverse of the slope of the J–V curves using eqn (8) and (9) respectively,⁴⁶

The parameters of the PEC cell performance, namely J_sc, V_oc, J_max, V_max, FF, η, R_s and R_sh, when using different MoBiCuSe₄ photoanodes illuminated under light radiation are depicted in Table 2.

Table 2 Variation of short circuit current (J_sc), open circuit voltage (V_oc), maximum current (J_max), maximum voltage (V_max), series resistance (R_s), shunt resistance (R_sh), fill factor (FF) and photoconversion efficiency (η) of MoBiCuSe₄ thin films deposited without and with surfactant

Sample code	Jsc (mA cm^−2)	Voc (mV)	Jmax (mA cm^−2)	Vmax (mV)	Rs (Ω)	Rsh (Ω)	FF	η (%)
MBCS	0.145	356	0.103	216	917	2454	0.44	0.074
MBCS:PEG	0.282	420	0.216	147	308	6666	0.45	0.178
MBCS:SDS	0.175	318	0.130	180	622	6876	0.42	0.078
MBCS:TOPO	0.197	330	0.145	192	714	3425	0.43	0.093

---

The light conversion efficiency of PEC cells using a semiconductor photoelectrode is effectively inhibited by the factor of recombination of photogenerated electrons and holes within the thin film photoelectrode. Thus, a better charge transport path is advantageous to achieve efficient electron transfer to minimize the recombination rate. As a low-dimensional morphology with reduced grain boundaries shows superior charge transportation,⁵⁸ the nanostructured MoBiCuSe₄ thin films deposited using different surfactants are expected to be favorable to provide an effortless pathway for smooth transport of the photogenerated charge carriers across a reduced number of grain boundaries within the photoelectrode. On the other hand, the observed improvement in the photoconversion efficiency of the surfactant-mediated samples due to the oriented nanostructured morphology is still not as significant as expected. This consequence may be rationalized by taking into account that the morphology is not the sole factor influencing the photoelectronic transport properties and some other factors, such as the crystallinity of the primary nanocrystals and the connection between the neighboring nanocrystals, may also have significant effects. Therefore, the photoconversion efficiency in the thin films is further increased by optimizing all of these factors.

From Table 2 it is observed that the surfactant-mediated MoBiCuSe₄ thin films (MBCS:PEG, MBCS:SDS and MBCS:TOPO) show better performance as compared to the MoBiCuSe₄ thin film without surfactant (MBCS). As observed from the structural and morphological studies, sample MBCS has smaller crystallite size, higher dislocation density and a petal-like surface morphology which offer a larger surface area to the photoelectrode for light absorption, however it has a recombination problem due to the existence of considerable grain boundaries in the film. Therefore, the electron trapping at the surface and intergrain boundaries results in lower light conversion efficiency. Samples MBCS:PEG, MBCS:SDS and MBCS:TOPO have improved crystallite size and lower dislocation density, which leads to reduced grain boundaries. This significantly reduces the recombination losses of photogenerated charge carriers due to the decrease in the grain boundary resistance. Also, they have nanofibrous, hollow-sphere-like vertically aligned nano-platelets and a marigold flower-like surface morphology, respectively, which is advantageous in providing a larger effective surface area for light absorption as well as providing a direct pathway for electron transport through the compactly interconnected nanofibers, nano-platelets and flowers, much faster than with loosely connected nanopetals. Further, the current–voltage characteristics are largely dependent on R_s and R_sh. As compared to the MoBiCuSe₄ thin film without surfactant, the surfactant-mediated MoBiCuSe₄ thin films give lower R_s and higher R_sh values. A lower R_s means that higher current will flow through the device and high R_sh corresponds to fewer shorts or leaks in the device. The ideal cell would have R_s approaching zero and R_sh approaching infinity. Thus, the nanostructured MBCS:PEG, MBCS:SDS and MBCS:TOPO thin films deposited using PEG, SDS and TOPO surfactants showed enhanced photoconversion efficiency.

From the J–V measurements it is observed that a higher magnitude of J_sc (0.282 mA cm⁻²), V_oc (420 mV) and light conversion efficiency, η (0.18%) were obtained for the MoBiCuSe₄ thin film deposited using the PEG surfactant. This is a considerable enhancement compared to the MoBiCuSe₄ thin film (MBCS, η = 0.07%) deposited without surfactant. As observed from the morphological studies, sample MBCS:PEG has a nanofibrous hollow-sphere-like structure and it is believed that the multiple reflections of light within the interior cavity of the hollow microspheres are responsible for the enhanced photoconversion efficiency of the core shell structures. Also, the fibrous nature of the microspheres is responsible for the better interconnection between crystallites, as well as the involvement of surfactant during the deposition giving thin films with an improved degree of crystallinity, which provides a direct pathway for smooth transport of photogenerated charge carriers across a reduced number of grain boundaries and is indirectly responsible for the enhancement in PEC conversion efficiency. Hence, it is concluded that the increased light conversion efficiency of surfactant-mediated MoBiCuSe₄ thin films (MBCS:PEG, MBCS:SDS and MBCS:TOPO) may be due to the increased effective surface area for light absorption and lower density of grain boundaries in the electron transport.

Though the photoconversion efficiency of the MoBiCuSe₄ thin films is still very low compared to other chalcogenide PEC cells,⁵⁹ study and development of an efficient p-type MoBiCuSe₄ thin film photoelectrode can greatly help in the improvement of solar radiation conversion efficiency by forming p–n heterojunction PEC solar cells, as they offer more protection against photocorrosion because the p-type semiconductors are cathodically protected under illumination. Efforts are in progress to design a p–n heterojunction solar cell by considering MoBiCuSe₄ as a p-type material in contact with another suitable n-type material, which will completely cover the interfaces across the MoBiCuSe₄ nanostructures.

Electrochemical impendence spectroscopy (EIS) measurements

In order to understand the effect of the different nanostructures of the MoBicuSe₄ thin films (MBCS, MBCS:PEG, MBCS:SDS and MBCS:SDS) deposited without and with surfactants on the PEC performance, electrochemical impedance spectra were recorded. Fig. 8 shows Nyquist plots of the MoBicuSe₄ thin films under ac bias. The Nyquist plot is composed of a semicircle in the high to middle frequency region and a sloping line in the low frequency region.

Fig. 8 The Nyquist plots of EIS of MoBiCuSe₄ thin films deposited without (MBCS) and with surfactants (MBCS:PEG, MBCS:SDS and MBCS:TOPO).

Generally, the semicircle, i.e. the intermediate frequency region, represents the charge transfer resistance (R_ct) (resulting from the interfacial redox reaction resistance) connected in parallel with an electrical double layer capacitance (C_dl) at the electrolyte–electrode interface and the sloping line with a 45° angle, i.e. the low frequency region, signifies the line of the Warburg diffusion element (Z_w) which corresponds to the diffusion of charge carriers into the bulk of the electrode material and charging of the film. The first arc in the high frequency region was attributed to impedance related to the series resistance (R_s) of the system (i.e. electrolyte/substrate resistance). The magnitude of R_s, R_ct, C_dl and Z_w were calculated by data fitting into a Randles circuit model, presented in the inset of Fig. 8, and the values are listed in Table 3.

Table 3 EIS parameters of MoBicuSe₄ thin films (MBCS, MBCS:PEG, MBCS:SDS and MBCS:SDS) obtained by fitting the data to a Randles circuit

Sample code	Rs (Ω)	Rct (Ω)	Cdl (μF)	Zw (mMho)
MBCS	22.2	560	8.27	2.76
MBCS:PEG	24.1	303	7.36	2.15
MBCS:SDS	21.3	533	8.56	3.51
MBCS:TOPO	23.6	355	6.67	1.59

---

From Table 3 it is observed that the surfactant-mediated MoBiCuSe₄ thin films (MBCS:PEG, MBCS:SDS and MBCS:TOPO) show smaller values of R_ct as compared to the MoBiCuSe₄ thin film without surfactant (MBCS), which is attributed to morphological modifications. Further, it can be seen that the MBCS:PEG thin film exhibits a much smaller semicircle in the high frequency region and lower slope in the low frequency region, compared to those of the MBCS:SDS and MBCS:TOPO thin films. According to former reports,^60–62 a smaller semicircle means a lower charge-transfer resistance. Hence, it can be concluded that the MBCS:PEG thin film has a high charge-transfer rate. This information supports the PEC results of the MoBiCuSe₄ thin films.

Conclusions

In conclusion, nanostructured MoBiCuSe₄ thin films were successfully deposited by APT using different surfactant media, namely PEG, SDS and TOPO. The surfactants have played a vital role in tuning the surface morphological features, which are suitable for the proposed applications. Different nanostructures were formed for the surfactant-mediated MoBiCuSe₄ thin films. As the surfactants do not affect the crystallographic orientation, the XRD patterns of the MoBiCuSe₄ thin films deposited without and with surfactant show the same crystal structures. A difference in the optical band gap was observed because of the strengthening of quantum confinement in the nanostructures of MoBiCuSe₄ thin films deposited using different surfactants. A substantial enhancement in the PEC cell performance of MoBiCuSe₄ thin films deposited using surfactants was observed. The special nanostructures formed for the surfactant-mediated MoBiCuSe₄ thin films improve the light harvesting capacity, not only because of the increase in surface area, which enhances trapping of light, but also by extreme scattering of light that is not absorbed, providing an effective path for light to travel throughout the nanostructures. Higher magnitudes of J_sc (0.282 mA cm⁻²), V_oc (420 mV) and η (0.18%) have been obtained for the MoBiCuSe₄ thin film deposited using PEG as a surfactant. As compared to the MoBiCuSe₄ thin film (η = 0.07%) deposited without surfactant, a significant improvement in light conversion efficiency (0.18%) of the PEG-mediated MoBiCuSe₄ thin film (MBCS : PEG) was obtained. The above efficiency results were also supported by the EIS study which shows a decrease in resistance for the MBCS:PEG sample. It is concluded that the surfactant-mediated approach is a promising route to deposit novel p-type MoBiCuSe₄ thin films with modified morphologies.

Acknowledgements

The authors are very much thankful to the Department of Science and Technology (DST) and University Grant Commission (UGC), New Delhi for providing financial assistance. This work was supported by the Priority Research Centers Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (20090094055).

Notes and references

1. M. R. Gao, Y. F. Xu, J. Jiang and S. H. Yu, Chem. Soc. Rev., 2013, 42, 2986.
2. M. A. Lazar, J. K. Tadvani, W. S. Tung, L. Lopez and W. A. Daoud, IOP Conf. Ser.: Mater. Sci. Eng., 2010, 12, 012017.
3. Q. Zhang and G. Cao, Nano Today, 2011, 6, 91.
4. M. Vahidi, S. W. Lehner, P. R. Buseck and N. Newman, Acta Mater., 2013, 61, 7392.
5. N. Han, F. Wang and J. C. Ho, Nanomater. Energy, 2012, 1, 4.
6. P. A. Cormier and R. Snyders, Acta Mater., 2015, 96, 80.
7. A. J. Naylor, E. Koukharenko, I. S. Nandhakumar and N. M. White, Langmuir, 2012, 28, 8296.
8. S. E. Wark, C. H. Hsia, Z. Luo and D. H. Son, J. Mater. Chem., 2011, 21, 11618.
9. R. F. Tabor, J. Eastoe and P. J. Dowding, J. Colloid Interface Sci., 2010, 346, 424.
10. R. F. Tabor, J. Eastoe and P. Dowding, Langmuir, 2009, 25(17), 9785.
11. C. J. Brinker and D. R. Dunphy, Curr. Opin. Colloid Interface Sci., 2006, 1, 126.
12. Y. Liu, J. Goebl and Y. Yin, Chem. Soc. Rev., 2013, 42, 2610.
13. K. K. Khun, A. Mahajan and R. K. Bedi, Electron. Mater. Lett., 2011, 7, 303.
14. M. Saranya, C. Santhosh, R. Ramachandran, P. Kollu, P. Saravanan, M. Vinoba, S. K. Jeong and A. N. Grace, Powder Technol., 2014, 252, 25.
15. R. Sathyamoorthy, P. Sudhagar, A. Balerna, C. Balasubramanian, S. Bellucci, A. I. Popov and K. Asokan, J. Alloys Compd., 2010, 493, 240.
16. C. S. McNally, D. P. Turner, A. N. Kulak, F. C. Meldrum and G. Hyett, Chem. Commun., 2012, 48, 1490.
17. C. N. R. Rao, H. S. S. R. Matte, R. Voggu and A. Govindaraj, Dalton Trans., 2012, 41, 5089.
18. D. P. Dubal, G. S. Gund, R. Holze, H. S. Jadhav, C. D. Lokhande and C. J. Par, Dalton Trans., 2013, 42, 6459.
19. M. Kuppayee, G. K. Nachiyar and V. Ramasamy, Appl. Surf. Sci., 2011, 257, 6779.
20. S. D. Kharade, N. B. Pawar, S. S. Mali, C. K. Hong, P. S. Patil, M. G. Kang, J. H. Kim and P. N. Bhosale, J. Mater. Sci., 2013, 48, 7300.
21. A. Purkayastha, Q. Yan, M. S. Raghuveer, D. D. Gandhi, H. Li, Z. W. Liu, R. V. Ramanujan, T. B. Tasciuc and G. Ramanath, Adv. Mater., 2008, 20, 2679.
22. A. Bhirud, N. Chaudhari, L. Nikam, R. Sonawane, K. Patil, J. Baeg and B. Kale, Int. J. Hydrogen Energy, 2011, 36, 11628.
23. J. Li, G. Lu, Y. Wang and Y. Guo, J. Colloid Interface Sci., 2012, 377, 191.
24. L. Dong, Y. Chua, Y. Zhang, Y. Liua and F. Yang, J. Colloid Interface Sci., 2007, 308, 258.
25. Z. Tan, H. Abe, M. Naito and S. Ohara, J. Colloid Interface Sci., 2010, 348, 289.
26. P. K. Panigrahi and A. Pathak, J. Nanopart., 2013, 2013, 10.
27. T. H. Tran, K. A. Tran, T. H. Pham, T. V. Le and Q. M. Le, Adv. Nat. Sci.: Nanosci. Nanotechnol., 2012, 3, 035012.
28. Y. Li, J. Wang, Y. Zhang, M. N. Banis, J. Liu, D. Geng, R. Li and X. Sun, J. Colloid Interface Sci., 2012, 369, 123.
29. N. Uekawa, R. Yamashita, J. W. Yong and K. Kazuyuki, Phys. Chem. Chem. Phys., 2004, 6, 442.
30. N. B. Pawar, S. S. Mali, M. M. Salunkhe, R. M. Mane, P. S. Patil and P. N. Bhosale, New J. Chem., 2012, 36, 1807.
31. H. Sharma, S. N. Sharma, U. Kumar, V. N. Singh, B. R. Mehta, G. Singh, S. M. Shivaprasad and R. Kakkar, J. Mater. Sci.: Mater. Med., 2009, 20, 123.
32. S. A. Vanalakar, M. P. Suryawanshi, S. S. Mali, A. V. Moholkar, J. Y. Kim, P. S. Patil and J. H. Kim, Curr. Appl. Phys., 2014, 14, 1669.
33. J. Tang, S. Hinds, S. O. Kelley and E. H. Sargent, Chem. Mater., 2008, 20, 6906.
34. F. M. Courtel, R. W. Paynter, B. Marsan and M. Morin, Chem. Mater., 2009, 21, 3752.
35. X. Yang, X. Wang and Z. Zhang, J. Cryst. Growth, 2005, 276, 566.
36. J. R. Ota, P. Roy, S. K. Srivastava, R. Popovitz-Biro and R. Tenne, Nanotechnology, 2006, 17, 1700.
37. V. B. Nascimento, V. E. de Carvalho, R. Paniago, E. A. Soares, L. O. Ladeira and H. D. Pfannes, J. Electron Spectrosc. Relat. Phenom., 1999, 104, 99.
38. A. Mamun, A. B. M. O. Islam and A. H. Bhuiyan, J. Mater. Sci.: Mater. Electron., 2005, 16, 263.
39. Z. Su, K. Sun, Z. Han, F. Liu, Y. Lai, J. Li and Y. Liu, J. Mater. Chem., 2012, 22, 16346.
40. T. P. Vinod, X. Jin and J. Kim, Mater. Res. Bull., 2011, 46, 340.
41. T. Takahashi, T. Sagawa and H. Hamanaka, J. Non-Cryst. Solids, 1984, 65, 261.
42. S. C. Riha, D. C. Johnson and A. L. Prieto, J. Am. Chem. Soc., 2011, 133, 1383.
43. B. D. Cullity, The elements of X-ray Diffraction, Addison-Wesley Publishing Company Inc, United States of America, 1st edn, 1957.
44. T. J. S. Anand and S. Shariza, Electrochim. Acta, 2012, 81, 64.
45. J. Tauc and A. Menth, J. Non-cryst solids, 1972, 8–10, 569.
46. I. Sisman and M. Bicer, J. AlloysCompd., 2011, 509, 1538.
47. R. K. Mane, B. D. Ajalkar and P. N. Bhosale, Mater. Chem. Phys., 2004, 84, 247.
48. G. P. Smestad, Optoelectronics of Solar Cells, SPIE – The International Society for Optical Engineering, Bellingham, Washington, 2002, ISBN-13: 978–0819444400.
49. B. D. Ajalkar, R. K. Mane, B. D. Sarwade and P. N. Bhosale, Sol. Energy Mater. Sol. Cells, 2004, 81, 101.
50. G. Chen, J. Seo, C. Yang and P. N. Prasad, Chem. Soc. Rev., 2013, 42, 8304.
51. R. S. Selinsky, Q. Ding, M. S. Faber, J. C. Wright and S. Jin, Chem. Soc. Rev., 2013, 42, 2963.
52. A. Khare, A. W. Wills, L. M. Ammerman, D. J. Norrisz and E. S. Aydil, Chem. Commun., 2011, 47, 11721.
53. K. E. Andersen, C. Y. Fong and W. E. Pickett, J. Non-Cryst. Solids, 2002, 299, 1105.
54. H. P. Garg and J. Prakash, Solar Energy: Fundamentals and Applications, Tata McGraw-Hill Education, 7 West Patel Nagar, New Delhi, 2000, ISBN: 0-07-4630631-6.
55. C. Garza, S. Shaji, A. Arato, E. P. Tijerina, G. A. Castillo, T. K. Das Roy and B. Krishnan, Sol. Energy Mater. Sol. Cells, 2011, 95, 2001.
56. S. B. Ambade, R. S. Mane, S. S. Kale, S. H. Sonawane, A. V. Shaikh and S. H. Han, Appl. Surf. Sci., 2006, 253, 2123.
57. S. S. Mali, C. A. Betty, P. N. Bhosale and P. S. Patil, Electrochim. Acta, 2012, 59, 113.
58. N. Han, F. Wang and J. C. Ho, Nanomater. Energy, 2012, 1, 4.
59. M. Berruet, M. Valdes, S. Cere and M. Vazquez, J. Mater. Sci., 2012, 47, 2454.
60. A. Gupta, P. Srivastava, L. Bahadur, D. P. Amalnerkar and R. Chauhan, Appl. Nanosci., 2015, 5, 787.
61. V. V. Kondalkar, S. S. Mali, R. R. Kharade, K. V. Khot, P. B. Patil, R. M. Mane, S. Choudhury, P. S. Patil, C. K. Hong, J. H. Kime and P. N. Bhosale, Dalton Trans., 2015, 44, 2788.
62. P. B. Patil, S. S. Mali, V. V. Kondalkar, N. B. Pawar, K. V. Khot, C. K. Hong, P. S. Patil and P. N. Bhosale, RSC Adv., 2014, 4, 47278.

---