N. Rajeswari Yogamalara,
K. Sadhanandama,
A. Chandra Boseb and
R. Jayavel*a
aCentre for Nanoscience and Technology, Anna University, Chennai 600025, India. E-mail: rjvel@annauniv.edu
bNanomaterials Laboratory, Department of Physics, National Institute of Technology, Tiruchirappalli 620015, India
First published on 26th January 2015
We employed a simple mechanism in tuning the electrical conductivity of graphene oxide (GO) by the inclusion of nano cadmium sulphide (CdS) and thereby, utilized it in a pn hetero-junction based solar cell. Two different kinds of hetero-junction solar cell devices are fabricated; one with CdS acting as a separate buffer layer in GO and in another, the GO was functionalized with quantum confined CdS particles. In both kinds, the n-type junction was interfaced with the p-type semiconductor Rose Bengal (RB). The nano CdS has the ability to harvest a useful spectral range of the solar spectrum and to retard the recombination rate between the junction interfaces and hence, promote the facile charge carrier transport within the device. The CdS functionalized GO exhibited a remarkable short circuit current density (Jsc) of 4 mA cm−2, and an open circuit voltage (Voc) of 685 mV, giving rise to an enhanced power conversion efficiency (η) of 1.97% in comparison to the planar GO hetero-junction. The enhancement in power conversion efficiency and the influence of nano CdS inclusion is analyzed and interpreted on the basis of the electrical studies performed with cyclic voltammetry (CV) and electrical impedance spectroscopy (EIS). From the electrical measurements, the charge storage capacitance of 138 F g−1 and the minimal resistance of 440 Ω are incurred in GO/CdS nanocomposites. The obtained results are further supported by the various physico-chemical characterizations such as X-ray diffraction (XRD), thermogravimetric analysis (TGA), Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), absorption and luminescence emission studies and demonstrate their significance in photovoltaic applications.
Currently, much work on photovoltaic application is carried out with high electrical conductive and transparent graphene sheet synthesized by physical synthesis technique. The certain limitations of graphene such as the poor solubility in water/organic solvent and the non-uniform film coating by chemical synthesis techniques established a new dimension to exert effort on the solution processable GO. The as-synthesized GO is an insulator with wide optical bandgap, however the switching in electrical conductivity can be achieved by the controlled deoxidation or partial reduction. In general, large numbers of complicated mechanism are adopted to tailor the electrical property of GO such as high energy irradiation,5 localized and non-localized defects creation,6 introducing large proportion of strong hazardous reducing agents like hydrazine7 and hydriodic acid,8 surface passivation by anchoring noble metals and metal oxide (Fe,9 Zn,10 Al,11 Sn,12 and SnO213), structure directing and chemically filling with metallic element or conducting polymers,14 and many more.
Although with poor electrical conductivity, the GO and the chemically reduced GO still finds an extensive application in the fabrication of photovoltaic device. Li et al.15 and Murray et al.16 reported the solution processed reduced graphene oxide (rGO) as an efficient hole transporter in the polymer and organic photovoltaic cell where the efficiency has been improved to 7.5%. An organic photovoltaic device fabricated with an ITO/PEDOT:PSS/P3HT-GO/Al structure rendered an increase in short-circuit current (Isc) and conductivity but a decrease in open circuit potential (Voc).17 The rise in the conductivity is evident from the fluorescence and electroluminescence quenching which promotes the easy charge transfer within the device. Similarly, the Cu2S/rGO sensitized with quantum dot CdSe nanoparticles shows a high fill factor (FF) of ∼75% and power conversion efficiency of 4.4%.18 The rGO/polypyrrole film is fabricated as a counter electrode for dye-sensitized solar cells and shows an improved power conversion efficiency of 8.14%.19 Marccus et. al. demonstrated the presence of GO at the enzyme/electrode interface diminishes the activation energy by decreasing the distance between the electrode surface and the enzyme cofactor thereby, it enable the GO to act as a biofuel cells and biosensors.20 Even though, with great achievements and introducing new concepts in the photovoltaic device modeling, the photovoltaic cell still lags in tuning high Isc and Voc simultaneously and the stability of the material is uncertain. To overcome the cited constraints such as poor conductivity in GO and an efficient photovoltaic device modeling, an approach is established to design a pn hetero-junction based solar cell with quantum confined CdS functionalizing the GO surface.
Many research articles published on GO/CdS nanocomposites are synthesized by utilizing an external driving force such as autoclave mediated solvothermal process and the usage of specific solvent medium (ethylene glycol) to enrich the homogeneous distribution of CdS nanoparticles on the surface of graphene.21,22 Moreover, the as-synthesized GO/CdS nanocomposites are effectively utilized as the fast photocatalyst where the environmental pollutant and waste material degradation are performed.23–25 In general, the graphene/semiconductor nanocomposites are recognized as active photocatalyst because the addition of semiconductor nanoparticles on graphene surface prevent the aggregation of graphene layers which in turn increases the surface area for the removal of organic pollutants from the aqueous solution.26 Gao et al. reported that the GO/CdS/Pt composite containing 0.5 at% of Pt exhibits the highest hydrogen evolution rate of 123 mL h−1 g−1 with strong photostability, which is about 2.5 times higher than that of GO/CdS composite and 10.3 times higher than that of pure CdS.27 To have a new insight, a simple chemical precipitation method is adopted towards the synthesis of GO/CdS nanocomposites and the same is implemented in the fabrication of hetero-junction based solar cell devices for the first time. Hence, the paper reports the dual performance of quantum confined CdS in enhancing the electrical conductivity of GO and further, improving the power conversion efficiency of pn hetero-junction based solar cells.
Here, a simple surface functionalization of GO is employed where the relative composition of oxygen to carbon atom is tuned to improve the electrical conductivity and for the same, the nanocrystalline CdS is selected to functionalize the GO surface. The optimum optical bandgap of 2.42 eV at 300 K, high absorption coefficient >104, and n-type electrical conductivity enable the CdS to extend the photon absorption over a wide spectral range of the solar spectrum (from visible to UV) and to act as a buffer layer and window absorber in the hetero-junction based solar cell.28,29 In addition, the oxygenated functional groups present in the GO surface easily bridge and link the CdS nanoparticles and render a stable chemical structure even at room temperature. Thus, the above features make CdS a promising semiconductor material for photovoltaic cells and open a framework to meet the global energy crisis.
To show the significance of the above synthesis method, the as-synthesized GO and CdS nanoparticles of equal weight ratio were blended together and the corresponding XRD and SEM micrographs were shown in Fig. S1.† The CdS nanoparticles were deposited and highly agglomerated on the surface of GO sheet rendering a random distribution. Due to the poor adherence, irregular dispersion and presence of large impurities as observed from the XRD pattern, the CdS blended GO nanocomposites was not considered for further property study.
To evaluate the role of nano CdS, three different combinations of hetero-junction based solar cell was designed. The first device designated as CFG (CdS functionalized GO) was composed of GO/CdS nanocomposites as n-type layer and RB as p-type layer. Both the layers were successively spin coated on the ITO coated glass substrate. The n-type layer was spin coated by dissolving 10 mg of GO/CdS nanocomposites in 10 mL of double distilled water and few drops of ethanol and citric acid were added to stabilize the coating solution. For hetero-junction, the p-type layer was achieved by dissolving 1 mg of RB in 10 mL of chloroform. The molecular structure, molecular formula, and the chemical name of RB31,32 were represented in the Scheme 2. The characteristic optical absorbance of RB peaked at 552 nm and 559 nm was shown in the Fig. S2.†
The second and third hetero-junction solar cells fabricated were designated as CSG1 (CdS sensitized GO 1) and CSG2 respectively. In CSG1, the CdS sensitizer acting as a buffer layer was deposited on the ITO electrode by SILAR technique using an ethanolic solution of cadmium nitrate and methanolic solution of thiourea. The buffer layer was successively spin coated by the GO layer and RB layer. Once after the layer-wise deposition of n-type and p-type material, the film was annealed at 200 °C for 40 min in temperature furnace under air atmospheric condition. The annealing temperature was ascertained from the heat transfer mechanism (refer Fig. 2(b)) revealing that the –S2− ions present in the GO matrix was retained and not transformed to its oxide form. Similarly, in CSG2, the CdS buffer layer and GO layer were deposited by interchanging the sequence of coating order. Finally, a proper ohmic contact was established on the n-type material and p-type material by the selective etching of ITO substrate and DC sputtering of Au electrode respectively. The schematic model of photovoltaic device architecture was illustrated in Scheme 3. The cross-sectional images and the approximate thickness were shown in the Fig. S3.† As the film layers were coated chemically, the individual layer was not distinguished clearly and the overall thickness was maintained in the range of 105 μm to 120 μm.
Fig. 1 (a) The XRD pattern of GO sheet, CdS nanoparticles and GO/CdS nanocomposites and (b) standard JCPDS data for cubic (C) and hexagonal phase (H) CdS. |
The XRD pattern of CdS nanoparticles synthesized by the chemical precipitation technique was represented in Fig. 1(a). The peaks were indexed and compared with the standard JCPDS data card no. 80-0006. The as-synthesized CdS nanoparticles exhibited hexagonal phase crystal structure with the lattice parameter a = b = 4.121 Å, c = 6.682 Å. The high intense reflection peak was observed at 2θ = 26.62° corresponding to the hexagonal (002) plane. The other peaks were indexed at 2θ = 25.06°, 28.24°, 36.82°, 43.94°, 48.06°, 52.16°, 67.19°, 71.26°, 75.97° and the planes were marked as (100), (101), (102), (110), (103), (112), (203), (211) and (105) respectively. The average crystallite size was estimated from the well-known Scherrer formula davg = (0.9λ/βcosθ). The davg corresponding to the (002) plane of CdS nanoparticles was calculated and found to be 130 nm.
The crystallinity of the GO/CdS nanocomposites was displayed in Fig. 1(a). As the CdS nanoparticles were incorporated into the GO matrix, the diffraction pattern was modified with few observations. Firstly, the strong reflection peak of GO at 2θ = 9.797° was shifted to lower scattering angle of 2θ = 8.54° and the peak was broadened with the full width half-maximum (FWHM) of 18.39°. The shift towards the lower angle implies that the interlayer spacing d was further increased to 12.6125 Å due to the inclusion of CdS nanoparticles. Moreover, the d value corresponding to the (002) plane of CdS nanoparticles was 3.3472 Å and for the (111) plane of GO/CdS nanocomposites the value was calculated to be 3.3895 Å. The difference arises due to the van der Waals force of interaction acting between the CdS nanoparticles and the GO sheet resulting in the GO sheet separation. The second observation in the XRD pattern of GO/CdS nanocomposite was that the reflection peaks corresponding to the CdS nanocrystal have undergone an incomplete phase transformation. The XRD pattern exhibits a mixed hexagonal phase and cubic phase of CdS nanocrystals as inferred and compared with the standard JCPDS data represented in Fig. 1(b). The peaks corresponding to the hexagonal phase and the cubic phase were denoted by the letter ‘H’ and ‘C’ respectively. The intensity of the ‘H’ peaks was weak and less pronounced than the ‘C’ peaks implying, under the given synthesis condition, the as-grown CdS particles with cubic phase were about to transform to stable hexagonal phase. Moreover, the davg corresponding to (111) plane of cubic CdS nanocrystal was 6 nm and was equivalent to the Bohr exciton radius of CdS. The reduced crystallite size and the presence of mixed phase imply the incomplete nucleation of cubic CdS crystals to stable hexagonal phase. Thus, the assigned synthesis reaction condition was insufficient to fasten the nucleation and the growth rate in GO/CdS nanocomposites. Finally, the nature of GO in GO/CdS nanocomposites can be revealed from the XRD study. The diffraction peak corresponding to the GO was still observed in the XRD pattern of GO/CdS nanocomposites implying that the GO was in the oxidized state. However, the intensity of the diffraction peak was reduced revealing that the GO was partially reduced as large number of oxygenated functional groups was substituted by the CdS nanoparticles. Thus, the XRD result clearly suggests that the CdS nanoparticles were incorporated and resulted in the formation of GO/CdS nanocomposites. The reaction mechanism was further supported from the TG/DTG study and the nature of GO can be inferred from the FTIR study and Raman analysis. The various parameters calculated from the XRD pattern of GO sheet, CdS nanoparticles, and GO/CdS nanocomposites were tabulated in Table 1.
Sample name | Average crystallite size davg (nm) | Intense peak position (°) | FWHM β (°) | Interplanar spacing d (Å) | |
---|---|---|---|---|---|
GO | 7 | 9.797 | 1.225 | 9.0213 | |
CdS | 130 | 26.61 | 0.658 | 3.3472 | |
GO/CdS | GO | 0.70 | 8.54 | 18.391 | 12.6125 |
CdS | 6 | 26.38 | 3.917 | 3.3895 |
The thermal behavior of CdS semiconductor nanoparticles was shown in Fig. 2(b). A gradual weight loss was observed in the as-synthesized CdS nanoparticles ranging from the room temperature to 600 °C and a sharp weight loss of about 22% was observed between 600 °C to 800 °C. The initial weight loss below 200 °C was attributed to the evaporation of physically absorbed water molecule and molecules with lower molecular mass. The second weight loss between 200 °C and 300 °C was the decomposition of S2− ions hence, the reaction temperature for the synthesis of CdS nanoparticles was fixed below 300 °C. The remaining change in the weight loss was attributed to the removal of SO42− group and NO32− group and the adhesion of O2− ions to the Cd2+ to form stable CdO material compound. When CdS was incorporated to the GO matrix to form GO/CdS nanocomposites, the same decomposition arises with the slight shift in the peak position as shown in Fig. 2(c). The weight loss appearing at 65 °C, 240 °C, 545 °C follows from the TG analysis of GO (Fig. 2(a)) and the loss at 393 °C, 678 °C, and 774 °C pursued from the CdS nanoparticles (Fig. 2(b)). In GO/CdS nanocomposite, the peak at 545 °C has high relative rate of weight loss in comparison to the GO peak at 585 °C implying the large fraction of oxygen removal from the GO surface due to the inclusion of CdS nanoparticles. This clearly suggests the formation of GO/CdS nanocomposites and the GO has undergone a partial redox reaction for the nucleation of CdS nanoparticles.
Fig. 3 (a) FTIR transmittance spectra of GO sheet, CdS nanoparticles and GO/CdS nanocomposites and (b) an enlarged view of the transmittance spectra obtained for GO sheet and GO/CdS nanocomposites. |
To define the characteristic adsorption peaks in GO and GO/CdS nanocomposites an enlarged view of FTIR spectra were shown in Fig. 3(b). The characteristic functional groups of GO were located at 3349 cm−1, 1724 cm−1, 1565 cm−1, and 1220 cm−1 and were assigned to the OH stretching of intercalated water molecule, CO stretching of COOH group, CC stretching vibration of alkyne, and C–O stretching vibration respectively. In comparison, the above assigned peaks were observed in the GO/CdS nanocomposites but, with reduced intensity or almost disappeared indicating that the oxygenated functional groups in GO were reduced partially. Moreover, the inclusion of CdS nanoparticles was confirmed from the presence of the characteristic Cd–S stretching vibrational peak at 671 cm−1 which was red shifted in comparison to the peak in pure CdS nanoparticles. Thus, the FTIR spectra conclude that the atomic composition of oxygen in GO/CdS nanocomposites was less than that in GO and the CdS nanoparticles were included in the GO matrix. The large variation in the transmittance intensity for GO and GO/CdS nanocomposites were highlighted in the Fig. 3(b).
The Fig. 4 represents the Raman scattering measured on GO and GO/CdS nanocomposites in the wavenumber ranging from 250 cm−1 to 2100 cm−1. The Raman spectrum of GO consists of two sharp peaks at 1342 cm−1 and 1588 cm−1 corresponding to the D-band and G-band of GO respectively. The intensity of the D band (ID) was related to the presence of defects and disorders of surface modified GO sheet and the IG band was associated with the vibration of sp2-hybridized carbon atoms in 2D hexagonal lattice of GO.36 The similar peaks appear for GO/CdS nanocomposites but, with the reduced intensity. The relative intensity ratio of D-band to G-band of GO and GO/CdS nanocomposites was 0.76 and 0.94 respectively. The enhancement in the relative intensity ratio ID/IG of GO/CdS nanocomposites reflects the relative fraction of sp2-hybridized carbon atom was lowered and the distorted structure of GO/CdS nanocomposites due to the redox reaction for the nucleation of CdS nanoparticles on the surface of GO. In addition to the D-band and G-band peaks, three optical vibrational Raman active modes of CdS nanoparticles were observed. The peaks at 298 cm−1, 584 cm−1 and 903 cm−1 were assigned to the fundamental longitudinal optical phonon mode 1LO, the first overtone mode 2LO, and the second overtone mode 3LO of CdS nanoparticles respectively.37 The presence of the above peaks gives an evident for the inclusion of CdS nanoparticles to GO.
From the heat transfer mechanism, FTIR study and Raman analysis, the probable redox reaction and the nature of interaction between the GO sheet and CdS nanoparticles was ascertained and illustrated in the Scheme 4. The cadmium nitrate and thiourea solution were sonicated individually for 1 h so as to detach the Cd2+ ions and S2− ions from the respective chemical precursor. The slow addition of Cd2+ ions and S2− ions to the dispersed GO solution render an easy solubility, uniform dispersion, and maximum substitution of the ions to the oxygenated functional groups adhered to the GO surface. The suitable reaction condition and the strong affinity favors the adsorbed ions to form CdS nanoparticles on the surface of GO.
The surface morphology of the as-synthesized samples was further examined by TEM. The multilayered 2D GO structures with nearly regular spacing (100 ± 9 nm) were clearly seen in Fig. 6(a). In high magnified TEM image displayed in Fig. 6(b), the oxygen functional group (denoted by dotted circle) adhered to the edge plane of graphite was noticeable which enable the GO to be stacked one over the other. Thus, the as-synthesized GO was in oxidized state as evident from the TEM micrographs. The selected area electron diffraction (SAED) pattern shown in Fig. 6(c) displays a ring pattern and the regular spacing in the GO layer was highly reflected. The CdS nanoparticles synthesized from the cadmium nitrate and thiourea precursors were rod shaped and the length of the rods were in the range of 220 nm to 370 nm and the width in the range of 90 nm to 110 nm (refer Fig. 6(d and e)). The SAED pattern represented in Fig. 6(f) was hexagonal dot pattern implying the 1D oriented growth of CdS nanoparticles along [002] direction as confirmed from the XRD study. The TEM images of GO/CdS nanocomposites were portrayed in Fig. 6(g and h). Here, the CdS nanoparticles were highly dispersed in the GO matrix and such distribution renders best material properties in hybrid hetero-structures. Moreover, the CdS nanoparticles nucleated on the surface of GO were very small in size and denoted by dotted circle. The SAED pattern of GO/CdS nanocomposites was shown in Fig. 6(i). The reflections originating from the GO were amorphous in nature whereas, the dotted reflections were from the CdS nanoparticles. The dotted reflections were random in nature and can be attributed to the presence of the mixed phase (cubic phase ‘C’ and hexagonal phase ‘H’) CdS nanoparticles as estimated from the XRD pattern (refer Fig. 1(a)).
However, to determine the individual particle size of CdS nanoparticles and to confirm the mixed phase of CdS in GO/CdS nanocomposites, HRTEM images along with the SAED were presented in Fig. 7(a–c). The particles were arranged lamellar-leaf like structure and the individual particle diameter range from 6.20 nm to 7.70 nm. In Fig. 7(b), the lattice spacing was clearly resolved and the d spacing was calculated to be 0.3346 nm revealing the cubic phase of CdS nanoparticles. Further, the formation of GO/CdS nanocomposite and the presence of mixed phase in CdS were confirmed from the SAED pattern as shown in Fig. 7(c). The high intense reflection indicated as spot 1 in Fig. 7(c) corresponds to the (002) plane of GO material. The remaining reflections originate from the mixed phase of CdS nanoparticles enwrapped in the GO matrix. In addition, the reflections from the trace material graphite were also observed in the SAED pattern. The d spacing values of various phases and the corresponding (hkl) planes were signified in the Fig. 7(d). Thus, the HRTEM images and SAED pattern of GO/CdS nanocomposites were in good agreement with the XRD pattern.
Fig. 7 (a and b) HRTEM images, (c) SAED pattern and (d) the corresponding d spacing values of GO/CdS nanocomposites. |
Fig. 8 (a) UV-visible DRS absorption spectra and (b) optical bandgap of GO sheet, CdS nanorods and GO/CdS nanocomposites. |
Further, the optical bandgap was estimated from the Kubleka–Munk function39 as shown in Fig. 8(b). The equation was represented as where F(R∞) was the K–M function or re-emission function, R∞ was the diffuse reflectance of an infinitely thick sample, K(λ) was the absorption coefficient, s(λ) was the scattering coefficient, and hν was the photon energy. The Eg was determined by extrapolating the linear portion (denoted by dotted line in the Fig. 8(b)) of the plot obtained between [(F(R∞)hν)1/2] versus hν. The optical bandgap of as-synthesized CdS nanorods and GO/CdS nanocomposites was 2.2 eV and 2.3 eV respectively.
The Fig. 9 shows the room temperature luminescence emission spectra of GO sheet, CdS nanoparticles and GO/CdS nanocomposites excited at 350 nm and the spectra were recorded in the wavelength of 400 nm to 800 nm. The GO shows flat band with no emission whereas, the CdS nanoparticles show a broad emission covering the maximum visible region of the spectrum (470 nm to 750 nm) with three distinct peaks centered at 540 nm, 570 nm and 650 nm. The near band edge emission peak at 540 nm corresponds to the recombination of free excitons of CdS nanoparticles. The other emission peaks were attributed to the presence of surface defects and vacancies in the as-synthesized CdS nanoparticles. The excess Cd atoms present in the CdS can lead to the formation of an acceptor level within the CdS optical band gap. Thus, the defect band at 570 nm was attributed to the transition of Cd-interstitial donors to the valence band. The shoulder peak at 650 nm was attributed to the recombination of an electron trapped in a sulfur vacancy with a hole in the valence band of CdS.40
In the spectrum of GO/CdS nanocomposites, the emission of GO/CdS nanocomposites was almost the same as that of pure CdS nanoparticles. However, there was a slight shift in the peak position and significant fluorescence quenching. The introduction of CdS nanoparticles of reduced crystallite size to the GO surface alters the van der Waals force of interaction between the GO layers and the electrostatic force of interaction between the CdS and GO. The above interactions result in the variation of the energy levels and hence, shift in the emission peak. The inclusion of CdS nanoparticles exhibit surface related trap centre which acts as a non-radiative electron–hole recombination centre or de-excitation centre for the photo-generated charge carriers.41 Thus, the fluorescence emission quenches, indicating the high surface interaction of CdS to GO sheet and an efficient electron transfer from the CdS nanoparticles to GO. The PL study clearly suggests the choice of CdS nanoparticle inclusion to GO will improve the photovoltaic performance.
The CV curve of GO and GO/CdS nanocomposites recorded at 100 mV s−1 scanning rate was displayed in Fig. 10(a). Here, the obtained voltammograms do not exhibit distinct anodic potential and cathodic potential in the applied potential range of 0.0 V to 0.5 V, implying the participation of capacitive current. From the average integral area of the CV curve, the specific capacitance Cs of the material can be calculated. The mathematical equation of Cs was Cs = (i)/(r × m) where, i the average current performed over the positive or negative half cycle, m the initial mass loaded to the electrolyte and r the potential sweep rate or scan rate (100 mV s−1). As the capacitance was proportional to the integral area, the GO nanosheet exhibits a higher Cs value of 220 F g−1 while that of GO/CdS nanocomposites was 138 F g−1. The Cs value of GO nanosheet was nearly 1.6 times greater than GO/CdS nanocomposites implying a higher ion storage capacity in the former. This significant feature was due to the maximum contribution of pseudo-capacitance from the oxygenated functional groups adhered on the surface of GO sheet.42 The CV curves of GO and GO/CdS nanocomposites at different scan rates (10 mV s−1, 50 mV s−1, and 100 mV s−1) were displayed in the Fig. S4† and their integral area of the curve were improved gradually with the increase in the scan rate.
Fig. 11 Current density and voltage characteristics of CdS functionalized GO (CFG) and CdS sensitized GO (CSG) solar cells under AM1.5 (100 mW cm−2) solar illumination. |
Photovoltaic devices | Voc (mV) | Jsc (mA cm−2) | FF (%) | η (%) |
---|---|---|---|---|
ITO/CSG1/RB/Au | 615 | 2.085 ± 0.063 | 73 | 0.94 ± 0.028 |
ITO/CSG2/RB/Au | 392 | 2.418 ± 0.073 | 66 | 0.48 ± 0.014 |
ITO/CFG/RB/Au | 685 | 4.071 ± 0.122 | 71 | 1.97 ± 0.059 |
From the table, it was found that the solar cell with the CFG/RB hetero-junction exhibited higher efficiency of 1.97% than CSG. The Voc, Jsc, and FF values of CFG were all tuned simultaneously and appreciable in comparison to the earlier reports on hetero-junction based solar cells.17,18 In CSG1 and CSG2, the oxygenated functional groups present in GO shields the charge transformation from the CdS nanoparticles (buffer layer) to GO and from the GO to CdS respectively. The hindrance offered in the charge transport results in insignificant η value and the value was lower for CSG2 than CSG1. The causes for such drastic variation were; (i) GO being an insulator with wide optical bandgap, the light trapping in the visible region of the solar spectrum was highly inhibited which in turn reduce the fraction of photo-generated charge carriers and the facile charge transfer from the GO to CdS in comparison to the electron transfer process from the photo-excited CdS quantum dots to the GO matrix, (ii) the deposition of GO between the metal electrode and semiconductor CdS was found to alter the work function of the electrode and decrease the injection of charge carriers into the PV device due to the high degree of oxygenated functional groups present on the GO surface and thus, lowers the Voc.
Moreover, to reveal the significance of CdS nanoparticles towards the functionalization of GO, the photovoltaic cell was designed with the CdS/RB np hetero-junction and the solar cell performance was evaluated and depicted in the Fig. S6.† The efficiency of CdS/RB was found to be 0.41% and the low efficiency clearly reveals that the n-type CdS nanoparticles alone cannot promote the easy transport of charge carriers between the hetero-junction and the collecting electrodes. The PL spectrum of CdS nanoparticles (refer Fig. 9) shows the presence of defect bands at 570 nm and 650 nm which were attributed as the electron trap centre and hence, prevent the facile transfer of electrons between the junction and thus, lowers the current density (1.4 mA cm−2) and solar efficiency.
But, for the photovoltaic device applications, the hetero-junction should promote facile charge transport, which, can be achieved by the complete reduction of GO to graphene. However, the graphene are no longer dispersible in water/organic solvent and the high transparency was lost which were essential for the uniform film coating and for the maximum adsorption of photon energy respectively. Based on the above crisis, the as-synthesized GO was reduced partially by the surface functionalization of CdS nanoparticles and thus, compensating both the characteristics. Thus, in CFG, the sp2-hybridized structure of GO was destructed by the inclusion of CdS nanoparticles which resides either at the edges or on the surface of GO sheet and therefore, leads to the formation of defects. The strained region of GO lattice exhibits higher chemical reactivity and consequently easy displacement of electron density from the upper plane of the sheet. This surface alteration and homogeneous distribution of quantum confined CdS nanoparticles on the GO surface render facile and rapid charge (electron and hole) transfer within the device. In addition, the size quantized CdS nanoparticles (davg = 6 nm) with low optical band gap of 2.3 eV was included to absorb wide range of the solar spectrum and the nanoscale mixing of CdS nanoparticles into the GO matrix increases the interfacial area and produce large number of excitons per incident photon. Further, the fluorescence quenching observed in the PL spectrum and the low interface resistance measured from the EIS facilitates the non-radiative recombination pathway for the electron–hole separation and transportation, thus, simultaneously controlling the Jsc, Voc, and η parameters of the CFG photovoltaic device.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra13061h |
This journal is © The Royal Society of Chemistry 2015 |