Long-term stability of Si-organic hybrid solar cells with a thermally tunable graphene oxide platform

Abstract

We demonstrated that a reduced graphene oxide (rGO) layer inserted between a poly(3,4-etylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) and Si interface improves the stability of Si-organic hybrid solar cells. Thermal reduction of a graphene oxide layer at different temperatures results in a change of water contact angle that is designed to eliminate the surface oxidation of the Si substrate. Higher temperatures generate a smaller aromatic sp² domain size that changes the GO surface from hydrophilic to hydrophobic. Additionally, two-dimensional GO layers can serve as a barrier for both liquid and vapor permeation. Consequently, the PEDOT:PSS/Si device with a rGO passivation layer shows a significantly enhanced air-stability from the J–V characteristic under illumination during a one-month storage period. The application of rGO is a promising technique to extend the stability of Si-organic solar cells without an encapsulation process.

---

1. Introduction

Inorganic–organic solar cells based on silicon (Si) are currently under intensive investigation due to the unique advantages they offer compared with the traditional high temperature process p–n junction cells, such as easy fabrication, simple device structure, and potentially low-temperature process.^1,2 The Si-organic solar cells combining the advantageous characteristics of crystalline Si and transparent conductive organic materials, such as poly(3,4-ethylene dioxythiophene):poly-styrenesulfonate (PEDOT:PSS), have emerged as promising candidates for cost-effective photovoltaic devices. The main functions of the organic layer are the following: (i) formation of a heterojunction with Si, (ii) a hole/electron transporting path, (iii) an optical window, and (iv) an antireflection coating.^3–5 Over the past few years, considerable progress has been achieved in the study of Si-organic solar cells, and power conversion efficiencies (PCE) of over 10% have been reported by many groups.^6–8 However, as the surface defects (i.e. microvoids) of PEDOT:PSS were always generated during the spin-coating process, the oxygen and moisture diffused through the PEDOT:PSS layer to form an insufficient oxide layer on Si substrate.⁹ As a result, the PCE of Si-organic solar cell gradually degrades during operation. To enhance the stability of Si-organic solar cell, previous studies have reported the usage of SiO_x, MoO₃, SiN_x, and Al₂O₃ layers.^8,10–12

Recently, for the photovoltaic applications, graphene oxide (GO) has been adopted in organic solar cells as the hole transport layer (HTL).^13–27 The GO sheets are exfoliated from graphite and dispersed in water to make the GO aqueous dispersion that has many advantages such as low cost, high yield, and processability. However, the PCE of organic solar cells with GO is highly sensitive to the thickness of GO layer because of its insulating properties. For the enhanced conductivity of GO, an insulating GO layer is easily tunable by chemical and thermal reduction processes to remove its oxygen functional groups, and the resulting layer is called reduced GO (rGO).²⁸ Yang et al. demonstrated that treated GO with oxygen plasma was efficient in blocking electrons and collecting holes in organic solar cells.¹⁶ The work function of GO with the oxygen plasma treatment was higher than that of GO layer, which results in the increase of the hole mobility and the enhancement of hole extraction. Liu et al. indicated that the rGO reduced by thermal annealing (130 and 230 °C) enhanced the electrical properties in organic solar cells.²³ Paci et al. mentioned that organic photovoltaic devices with the partially reduced GO (160 °C of annealing temperature) used as the HTL exhibited photovoltaic characteristics similar to the PEDOT:PSS-based devices but with a significant stability enhancement.²⁴ So far, almost all reports have paid attention to improving the performance of Si-organic solar cell with the tunable physical properties of rGO.^25,26 However, there is limited research regarding the effect of thermally annealed rGO on the long-term stability of Si-organic solar cells.

In this paper, we report the fabrication of package-free Si-organic solar cell with the use of rGO passivation layer. The coated GO layers on a Si substrate were changed to rGO layer by a thermal annealing process at temperatures of 700, 800, and 900 °C. Then, we observe the surface change of individual rGO layer by the water contact angle. The PEDOT:PSS/rGO/Si solar cell show long-term stability by current density–voltage and current–voltage characteristics in air ambient during a one-month storage period. We find that the rGO layer helps to prevent water molecules interaction in PEDOT:PSS/rGO/Si solar cell.

2. Experimental

2.1 GO synthesis

The GO was synthesized by using the modified Hummer's method.³⁰ 5 g of enlarged graphite and 500 mL of concentrated H₂SO₄ were mixed in a bottle to make a suspension. Then, 30 g of KMnO₄ was slowly added and the resulting mixture was stirred for 2 h at 800 rpm. The mixture was subsequently diluted with 6 L of de-ionized water and thermally stabilized below 70 °C before slowly adding 15 mL of H₂O₂ (30 wt%) until active bubbles appeared, and the color of the suspension changed from dark brownish to yellow. The suspension was filtered and washed four times with 5 L of 10% HCl solution, followed by six washes with 8 L of deionized (DI) water to create GO with neutral pH. The resulting GO dispersion had a concentration of 0.025 g L⁻¹ in DI water.

2.2 Device fabrication and characterization

The PEDOT:PSS solution (Clevios, PH1000), which was purchased from Heraeus, was first mixed with 5 wt% dimethyl sulfoxide (DMSO) and 1 wt% surfactant (Zonyl-FSH).³¹ N⁺-type Si wafers (100, phosphorous doped, ∼2 × 10¹⁵ cm⁻³) have a thickness of ∼500 μm and resistivity of 1–10 Ω cm. For the fabrication of Au/PEDOT:PSS/rGO/Si/GaIn/Ag solar cell, the Si substrates (1 × 1 cm²) were cleaned in acetone, ethanol, and DI water for 15 min each in sonication. Then, the Si substrate was immersed in a buffered oxide etchant (BOE) for 1 min to remove native oxide, rinsed with DI water, and blown dry with N₂. After the cleaning process, the GO dispersion was sprayed onto the Si substrate using an airbrush system with N₂ as a carrier gas. The airbrush inlet pressure of nitrogen, substrate temperature, and spraying rate were 2 bars, 150 °C, and 1 mL min⁻¹, respectively. For preparing the rGO layer, the next step was the thermal annealing process at 700, 800, and 900 °C with hydrogen ambient in a metal–organic chemical vapor deposition system. Subsequently, the rGO/Si substrate was spin-coated with PEDOT:PSS aqueous solution at 3000 rpm. Next, the PEDOT:PSS layer was baked at a temperature of 170 °C in air ambient. The thickness of PEDOT:PSS film was determined to be about 60 nm using a scanning electron microscope.³² Then, solar cell shaped finger-grid pattern Au (100 nm) electrode with 2 mm width was deposited on the PEDOT:PSS film through a finger-grid patterned metal mask by electron beam evaporator at a pressure of 3 × 10⁻⁶ torr. The size of the device was 1 × 1 cm². The schematic image of the hybrid device is shown in Fig. 1.

Fig. 1 Schematic illustration of the inorganic–organic solar cell structure with GO layer, the bottom electrode GaIn, interlayer materials n-Si with GO (or thermally reduced GO, rGO), and PEDOT:PSS (hole acceptor) and top electrode Au. (b) The schematic cross section of the n-type Si/GO or rGO/PEDOT:PSS solar cell. (c) Top image of the proposed device.

Atomic force microscopy (AFM, Digital Instruments, Nanoscope IV A) in tapping mode was used to observe the thickness and surface morphologies of GO and rGO layers annealed at 700, 800, and 900 °C. The chemical bonds of the samples were analyzed using X-ray photoelectron spectroscopy (XPS, AXIS-Nova, Kratos) with an Al Kα radiation source (1486.8 eV) in an ultrahigh vacuum chamber (7 × 10⁻⁹ torr). For the wettability of the surface, the contact angle was measured using an SEO Phoenix contact angle equipped with a dispensing needle (Surface Electro Optics Co., Korea). The optical transmittance of prepared samples was measured using an ultraviolet-visible-near-infra-red spectrophotometer. The sheet resistance of rGO layers was measured by a four-point probe technique to confirm the enhanced electrical properties. The crystalline quality of all samples was evaluated by Raman spectroscopy using a 633 nm laser. To examine the solar cell performance of the PEDOT:PSS/rGO/Si hybrid devices, the photocurrent was measured by recording the photocurrent density versus voltage (J–V) curve under illumination from a solar light simulator (AM 1.5 with a power density of 100 mW cm⁻², EXIL-0.5850KSVI). Electric current versus voltage (I–V) characteristics were measured to determine the Schottky barrier height (SBH) of devices using a probe station system connected to a parameter analyzer (Semiconductor Test and Analyzer, EL420) in the dark. Finally, the long-term reliability characteristics of PEDOT:PSS/rGO/Si hybrid solar cells were measured by the I–V and J–V characteristics during air exposures of up to 30 days.

3. Results and discussion

We have analyzed the thickness and surface morphologies of the rGO layer with various annealing temperatures (700, 800, and 900 °C) by AFM measurement in Fig. 2. The samples are designated as rGO-700, rGO-800, and rGO-900 based on the annealing temperature. Fig. 2a–d are the AFM images of the surface of as-sprayed GO and rGO samples with each annealing temperature and Fig. 2e shows a graph of the surface roughness (R_q) value by AFM analysis. As shown in Fig. 2a–d, the surface roughness of the as-sprayed GO (Fig. 2a) was very different from those of the rGO-700, rGO-800, and rGO-900 layers (Fig. 2b–d). The R_q value of the as-sprayed GO is 1.81 nm, and that of rGO-700, rGO-800, and rGO-900 are approximately 0.69, 0.6, and 0.52 nm, respectively. This indicates that the surface of the rGO layers with high annealing temperature were slightly smoother than that of the GO layer. Additionally, the thickness was measured to be ∼7.6 nm for GO layer, 1.4 nm for rGO-700, 1.3 nm for rGO-800, and 0.9 nm for rGO-900 by AFM. It is noted that the various oxygen functional groups are eliminated from GO sheets during thermal annealing.

Fig. 2 AFM images of (a) as-coated GO layer, and thermally rGO by (b) annealed 700 °C, (c) annealed 800 °C, and (d) annealed 900 °C. (e) The graph of RMS values depend on an annealed temperature of GO.

The chemical compositions of all samples were analyzed by conducting XPS measurements, and the results are shown in ESI, Fig. S1.† XPS results show a C1s peak at 284.8 eV. The C1s spectrum of GO are composed of the five kinds of components that are assigned to C–O or C–OH (hydroxyl, 285.5 eV), C=O (epoxide, 287.2 eV), O=C–OH (carboxyl, 288.2 eV), and O–C=O (carbonyl, 289.3 eV).^33,34 The peak intensity corresponding to the oxygen functional groups in rGO layer gradually decreased with the increase in annealing temperature. The C/O ratio is 0.89 for GO, 2.13 for rGO-700, 2.34 for rGO-800, and 2.71 for rGO-900. It appears that the C/O ratio increases with an increase in annealing temperature, indicating that the oxygen functional groups diminished with thermal annealing. The optical transmittance spectra of GO and rGO layers were shown in ESI, Fig. S2.† We observed that the transmittances of rGO layers are slightly lower than that of GO layer although the thickness of GO layer is higher than that of rGO layers. To further confirm the enhanced electrical property of rGO layers, sheet resistances of rGO-700, rGO-800, and rGO-900 were also measured by a four-point probe technique. The sheet resistances of rGO-700, rGO-800, and rGO-900 are 10.6 × 10³, 6.8 × 10³, and 2.5 × 10³ Ω □⁻¹, respectively. These findings indicate that the sheet resistance of rGO layer decreases with an increase in the C/O ratio. It should be noted that the PCE and short-circuit current (J_SC) of the Si-organic solar cell with rGO layer are strongly dependent on both the transmittance and sheet resistance of rGO layers.

As the wettability of a surface is strongly influenced by its chemical composition, we performed the water contact angle measurements by placing a droplet of DI water on the various rGO surfaces. A 1 μL water droplet was generated using the automatic dispenser of the system. The measurements were performed before and after the GO and rGO layer were stored in air ambient under room temperature for one-month. Fig. 3 shows the water contact angles of GO and rGO-700, rGO-800, and rGO-900 layers. For the surface of the GO layer, we find that the water contact angle is ∼22.06° on the first day and ∼11.72° on the 30th day. The GO layer is expected to have many existing oxygen functional groups attached to the sheet edges, which makes it display hydrophilic properties.^35,36 On the other hand, the water contact angle of the rGO layer is about ∼62.3° for rGO-700, ∼66.3° for rGO-800, and ∼71.4° for rGO-900 on the first day. The C/O ratio of the rGO layer increased as the annealing temperature increased. In the results, it is demonstrated that the rGO layer with a high annealing temperature has fewer surface functional groups, resulting in the water contact angle of rGO-900 being larger than that of others. After 30 days, the water contact angle of the rGO layer is about ∼57.9° for rGO-700, ∼62.5° for rGO-800, and ∼67.6° for rGO-900, after one month. The GO layer wettability increased more than the rGO layers after one month. The results show the effectiveness of rGO layer with various annealing temperatures as an anti-corrosion barrier with wetting stability. It implied that rGO layer could act as an oxygen/moisture barrier that can prevent oxygen/water from contacting the Si substrate, results in eliminating surface oxidation at the Si-organic interface.^24,29 Consequently, the rGO layer serves as a passivation layer between PEDOT:PSS and Si interface and facilitates enhancement of the lifetime of the Si-organic solar cell.

Fig. 3 (a) The water contact angle of the GO, rGO-700, rGO-800, and rGO-900 layer, respectively. A black body, an oblique line and a check pattern means the measurement dates as initial day, and after 7 and 30 days. (b) The photographs of water drops on the GO, rGO-700, rGO-800, and rGO-900 layer as a function of measurement days.

To investigate the effect of the rGO passivation layer on the Si-organic solar cells performance, our proposed devices were fabricated using PEDOT:PSS as the organic layer and tested under identical conditions. Fig. 4a shows the current density–voltage (J–V) curves of all the devices measured in AM 1.5 illuminated at 100 mW cm⁻². The relative short-circuit current (J_SC), open-circuit voltage (V_OC), fill factor (FF), and PCE of our proposed solar cells are summarized in Table 1. Fig. 4b displays the current–voltage (I–V) curves of our proposed devices, which are PEDOT:PSS/Si solar cells with the GO and rGO layers, measured in dark condition. The PEDOT:PSS/Si without GO layer shows PCE of 5.4%, J_SC of 22.3 mA cm⁻², V_OC of 0.56 V, and FF of 0.49. However, when the GO layer is inserted between PEDOT:PSS and Si, the PCE, J_SC, V_OC, and FF parameters of solar cell devices are 5.1%, 15.6 mA cm⁻², 0.56 V, and 0.59, respectively. We believe that the poor performance is attributable to the many oxygen functional groups of GO. It means that the oxygen functional groups of GO layer (i.e. vacancies and interstitial sites) cause the high charge trap density, resulting in the lower J_SC in PEDOT:PSS/Si with GO layer. The PEDOT:PSS/Si devices without and with GO layer exhibit rectification behavior in Fig. 3b. The Schottky barrier formed at the Si-organic interface is responsible for the rectification of major carrier (electron) flows. This property should persist after the insertion of GO passivation layer. The leakage current and saturation current (I₀) of the device are lower in a PEDOT:PSS/Si solar cell with an inserted GO layer than that of a PEDOT:PSS/Si solar cell without the GO layer. It is well known that the dark current in a Schottky diode is dominated by majority carrier (electron) injection. Considering that the bandgap of GO increases with the increase of the oxygen functional groups,³⁷ it is possible that the GO layer inserted between the PEDOT:PSS and Si interface blocks the electrons. Then, the electron current is reduced to a point where the hole current starts to dominate the saturation current density, leading to the reduction of leakage current. In addition, we propose that J_SC of PEDOT:PSS/Si with GO passivation layer is reduced by the presence of a hole barrier. In other words, in the device with a GO layer, the valence band (E_V) of GO could be more aligned with the E_V edge of Si, as compared to the PEDOT:PSS/Si without GO layer.²

Fig. 4 (a) The J–V characteristics of Si/PEDOT:PSS solar cells with the inserted GO, rGO-700, rGO-800, and rGO-900 barrier layers, respectively, under illumination of AM 1.5 G, 100 mW cm⁻² simulated solar light. (b) A typical I–V characteristics on the logarithmic current scale of corresponding solar cells under dark condition. A estimated energy band diagrams of Si/PEDOT:PSS devices (c) without GO and (d) with GO or rGO barrier layer by annealed 700, 800, and 900 °C.

Table 1 J–V characteristics of Si/PEDOT:PSS solar cell without GO, and with GO, rGO-700, rGO-800, and rGO-900 barrier layer

Device	JSC (mA/cm^2)	VOC (V)	FF	PCE (%)	SBH (eV)
w/o GO	22.3	0.56	0.43	5.4	0.81
With GO	15.6	0.56	0.59	5.1	0.91
With rGO-700	15.6	0.44	0.60	4.1	0.74
With rGO-800	16.0	0.42	0.63	4.2	0.75
With rGO-900	12.3	0.26	0.46	1.3	0.69

---

For the devices with the inserted rGO passivation layer, the J–V curves show a significant decrease in performance. The J_SC, V_OC, and PCE of PEDOT:PSS/Si with rGO-700 and rGO-800 layers are 15.6 and 16.0 mA cm⁻², 0.44 and 0.42 V, and 4.1 and 4.2%, respectively. The lower J_SC and V_OC of PEDOT:PSS/Si with rGO-700 and rGO-800 are attributed not only to the high sheet resistance of the rGO layer between PEDOT:PSS and Si but also to the defects in the rGO layer, as compared to a device without the rGO layer. Fig. 4b shows that a rectifying behavior was observed for the PEDOT:PSS/Si with the rGO-700 and rGO-800 layers. The leakage current of PEDOT:PSS/Si device with an rGO-700 and rGO-800 is higher than that of PEDOT:PSS/Si solar cell, which produces a smaller V_OC. Considering that the bandgap of the rGO layer decreases with the increase of C/O ratio due to the removal of oxygen functional groups, its electron barrier is lower than that of the GO layer and leads to a higher I₀ of the devices with rGO-700 and rGO-800 layers. Besides, we suggest that the E_V of rGO-700 and rGO-800 should be higher than that of Si. Hence, the inserted rGO-700 and rGO-800 layers should decrease the J_SC of the solar cell compared to the device without the rGO layer. Interestingly, the J_SC, V_OC, and FF of the PEDOT:PSS/Si solar cell with the rGO-900 passivation layer are 12.3 mA cm⁻², 0.26 V, and 1.3%, respectively. The poorer performance of PEDOT:PSS/rGO-900/Si device is a consequence of the ohmic behavior in dark I–V curves, as shown in Fig. 3b. It was noted that the rGO-900 layer should affect electrical properties such as the electron and hole barrier height at the interface between PEDOT:PSS and Si. We proposed that the PEDOT:PSS/Si with rGO-900 passivation layer has a higher reverse current and saturation current (I₀) than the others as the rGO-900 layer has a lower band gap and E_V caused by higher annealing temperature.

To further verify the decrease in J_SC of PEDOT:PSS/Si devices with the inserted GO and rGO passivation layers, the relative parameters (i.e. Schottky barrier height (SBH), work function) were calculated from the corresponding dark I–V curves. The diode characteristics of Schottky junction are described through the thermionic emission theory:³⁸

andwhere J_s is the saturation current density, e is the electronic charge, n is the ideality factor, k is the Boltzman constant, T is the absolute temperature, and A* is the Richardson constant of Si (∼112 A cm⁻² K⁻²). By using eqn (1) and (2), the SBH was calculated as 0.91, 0.74, 0.75, and 0.69 eV for PEDOT:PSS/Si with GO, rGO-700, rGO-800, and rGO-900 passivation layers, respectively, as compared to ∼0.81 eV for that of PEDOT:PSS/Si without GO layer. The SBH of PEDOT:PSS/Si with the inserted GO layer is larger than that of PEDOT:PSS/Si without GO layer. Besides, the SBH of our proposed devices decreased as the annealing temperature of rGO layer increased. Even though the rGO-700 and rGO-800 layer have different C/O ratios because of the annealing conditions, the calculated SBH for the two devices were found to be similar. Moreover, it is possible that the PEDOT:PSS/rGO-900/Si device shows ohmic behavior because the current flows through the low SBH of the device. Theoretically, the built-in electric field in the junction becomes stronger when the SBH of the device increases. The larger SBH is beneficial for widening the absorption region as well as for separating the photo-carriers.^39,40 These results indicate that the PEDOT:PSS/GO/Si solar cell shows a higher performance than the PEDOT:PSS/Si with the inserted rGO passivation layer. In addition, the SBH values of solar cells depend on the work functions of rGO layers and the built-in potential of the devices. The SBH is expressed as

Φ_SBH = W_rGO − χ_Si (3)

andwhere W_rGO is the work function of rGO, χ_Si is the electron affinity of n-type Si (4.05 eV), V_bi is the built-in potential, E_c is the conduction band level of Si, and E_F is the Fermi energy level of Si. By using eqn (3), the work function values of GO, rGO-700, rGO-800, and rGO-900 are approximately 4.96, 4.79, 4.8, and 4.74 eV, respectively, as compared to ∼4.86 eV for PEDOT:PSS. The calculated work function of PEDOT:PSS layer is lower than its reported value (∼5.1 eV).⁴¹ The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) of PEDOT:PSS are at 5.1 and 3.5 eV, respectively. The obtained work function of GO layer is higher than that of the rGO layers. The obtained work function of rGO-700 is similar to that of the rGO-800 layer, and the calculated work function of rGO-900 is lower than that of rGO-700 and rGO-800. This finding indicates that the increasing SBH of the devices is caused by the increasing work function of materials. The work functions of rGO-700, rGO-800, and rGO-900 layers were obtained from the UPS spectra (ESI, Fig. S3†). The work function is defined as a minimum energy to liberate an electron from a substance and are determined from the equation, W = hν − |E_cutoff − E_F|, where hν is the photon energy of an incident light set at 21.22 eV, E_cutoff is the inelastic low-kinetic-energy cutoff, and E_F is the Fermi level edge. The E_cutoff and E_F are determined by using a linear extrapolation of low and high kinetic-energy cutoff region of the UPS spectra. The individual E_cutoff and E_F values are obtained from UPS spectra. However, the E_F of GO layer is lower than that of rGO layers because the charge shift caused due to oxygen functional groups of GO sample. As a result, the work function of the GO layer was ∼6.2 eV, which is similar to the reported value by Khoa et al.⁴² However, Liu et al. reported that the work function of GO and rGO were ∼4.7 and 4.5 eV (ref. 43) while it was found to be 4.9 eV by Li et al.⁴⁴ It means that work function of GO and rGO strongly depend on synthesis and reduction method which causes fluctuation in aromatic size, defect level, and the number of functionalized groups of the GO sheets.^33,45 In this study, the work function of rGO was found to be 4.97, 4.94, and 4.83 eV for rGO-700, rGO-800 and rGO-900 layer, respectively, which are comparable with other reports,^33,46 and they tend to be similar to the work function calculated by eqn (3). The change in rGO work function arises from the formation of dipoles between carbon and oxygen atoms with different electronegativity. Different oxygen-functional groups such as carbonyl, epoxy, and hydroxyl that are located outside the graphene basal plane have different C–O bonds types with different chemical composition. As a consequence, the strength of the out-of-plane dipole moment in rGO can be tailored by changing both the overall oxygen content and the concentration of different functional groups. Thus, the decrease in the work function of the rGO layers with the increasing thermal annealing temperature attributed to the deoxygenation of GO with heat treatment. According to previous reports, the work function value of rGO layer decreases if a large amount of oxygen function groups are removed in the GO during annealing.²⁹ However, it is noteworthy that the work function of rGO materials is still not well-defined and is under controversy because reaction mechanism for deoxygenation of GO differs with different reduction processes. Fig. 4c and d show a proposed energy band gap diagram with the obtained work function of rGO annealed at 700, 800, and 900 °C. The higher E_V offset between Si and PEDOT:PSS is attributed to the higher work function of GO layer from Fig. 4d, as compared to PEDOT:PSS/Si device. In other words, it demonstrates that few holes could be injected into the E_V level of GO layer after the electron–hole pairs are generated in Si upon irradiation, and then collected by the Au anode. The higher E_V energy barrier prevents the hole transport, resulting in a large SBH that could suppress the electron through the GO layer. Thus, the PEDOT:PSS/Si with GO passivation layer reduced the J_SC of the devices under illumination and the enhanced V_OC under the dark condition is obtained, as shown Fig. 4a and b. On the other hand, the inserted rGO-700, rGO-800, and rGO-900 layers also allow the transfer of hole from the Si semiconductor in accordance with Fig. 4d. The E_V offset at the interface between rGO layers and PEDOT:PSS could block the holes and electrons after electron–hole pairs are generated by collecting them at the Au anode. It is likely that the hole barrier prevents the photocurrent at PEDOT:PSS/rGO interface, and a smaller electron barrier leads to a higher I₀, as compared to the PEDOT:PSS/Si device. Consequently, we believe that the GO and rGO layers can not only act as the hole transport layer but also serve as an electron blocking layer for reducing the carrier recombination at the anode. Further increase in J_SC would need a tunable energy band gap engineering, and enhanced V_OC would require a reduction in the hole injection.

Furthermore, the parameter V_OC is related to the short-circuit current (I_SC), and saturation current (I₀). By using eqn (1), the V_OC is expressed as:²

The calculated V_OC is estimated as 0.76 eV for PEDOT:PSS/Si without rGO, 0.65 eV for PEDOT:PSS/Si with rGO-700, 0.67 eV for PEDOT:PSS/Si with rGO-800, and 0.58 eV for PEDOT:PSS/Si with rGO-900. These obtained values tend to show similar V_OC for all the devices in the J–V curves, as shown in Fig. 3a. As mentioned before, the defects in GO or rGO affect the recombination process of the solar cell.⁴⁷ The crystalline qualities of GO, rGO-700, rGO-800, and rGO-900 layers determines the defects of GO and rGO layers, and they were examined by Raman spectroscopy. The Raman spectra of the GO, rGO-700, rGO-800, and rGO-900 layers are compared (ESI, Fig. S4†). The Raman spectra of GO consist of two kinds of peaks, which are designated as the G peak (∼1584 cm⁻¹) and the D peak (∼1335 cm⁻¹).⁴⁸ Peaks for the D and G bands of GO, rGO-700, rGO-800, and rGO-900 layers are clearly resolved (ESI, Fig. S4a†). Peak intensity ratios between the D and G bands, i.e. I_D/I_G, were calculated (ESI, Fig. S4b†). The I_D/I_G values for GO, rGO-700, rGO-800, and rGO-900 layers are ∼1.03, 0.79, 0.82, and 0.96, respectively. This indicates an apparent decrease in the I_D/I_G ratio upon thermal annealing, which suggests that thermal annealing eliminates the oxygen functional groups from GO sheets.⁴⁹ However, the I_D/I_G ratio increases when thermal annealing temperature increases to 900 °C. It means that very high temperature leads to further removal of oxygen functional groups, results in high density of disordered carbons in rGO-900 layer.⁵⁰ Moreover, the I_D/I_G ratio is related to the average size of in-plane crystallites (L_a), which corresponds to a mean sp² GO domain size (or inter-defect distance).⁵¹ The L_a values of GO, rGO-700, rGO-800, and rGO-900 layers are ∼4.31, 5.37, 5.32, and 4.57 nm, respectively. An introduction of a large amount of oxygen functional groups in GO limits the sp² aromatic size. Thus, the sp² aromatic size increases with the removal of oxygen functional groups (graphitized) when the samples were annealed at 700 and 800 °C. However, further increase of annealing temperature generates the defects, resulting in the decrease of the sp² domain size. In other words, the larger inter-defect distance of rGO-700 and rGO-800 tends to enhance the recombination process during the generation and separation of photo-carriers. For this reason, a device with rGO-700 and rGO-800 should have lower leakage current at reverse-bias in dark ambient than a device with rGO-900. As a result, the V_OC of solar cells should decrease with the increase in the annealing temperature of the rGO layer, as shown in Fig. 4a. In our proposed devices, we inferred that the rGO layer was optimized by annealing at 700 and 800 °C.

To observe the effect of the hybrid solar cells with GO or rGO passivation layer, the stability of our hybrid solar cells was characterized in air ambient under room temperature. This can be observed from the J–V and I–V characteristics of our proposed devices measured after one-month duration (ESI, Fig. S5 and S6†). Fig. 5a–c show the normalized PCE, J_SC, and V_OC values over 30 days. As we can see from Fig. 5a, the PEDOT:PSS/Si solar cells with and without GO passivation layers exhibit rapid degradation compared with the others. After 5 days of storage, the PCE of PEDOT:PSS/Si without GO layer reduces by nearly 30% from its initial value. It is contributed by the decrease in J_SC of devices and the reduction of V_OC of the device in Fig. 5b and c. As some previous reports, the PEDOT:PSS/Si solar cells exhibit poor air stability because organic material is sensitive to oxygen and moisture.^9,52,53 As a result, the conductivity of PEDOT:PSS layer dramatically decreases during operation, leading to a sharp decrease of the J_SC in solar cells. To confirm the conductivity of PEDOT:PSS layer, Fig. 5d displays the change of series resistance (R_S) in solar cells as a function of the time duration. The I–V curve of all devices was also measured in dark condition over 30 days, and the R_S of all devices were calculated by:

Fig. 5 (a) Normalized PCE, (b) J_SC, and (c) V_OC of Si/PEDOT:PSS solar cell without GO, and with GO, rGO-700, rGO-800, and rGO-900 barrier layer, respectively. (d) The series resistances (R_S) of Si/GO or rGO/PEDOT:PSS solar cell as a function of measurement days.

The estimated R_S of PEDOT:PSS/Si solar cells increases from ∼5.5 Ω for the first day to ∼700 Ω for 30 days, as shown in Fig. 5d. It is confirmed that the J_SC of PEDOT:PSS/Si solar cell decreases with the increase of R_S in devices. The V_OC of PEDOT:PSS/Si solar cell also decreases with the decline of J_SC and the low I₀ value as a function of measuring time is given by eqn (5). In addition, the PCE of the PEDOT:PSS/Si with GO passivation layer decreases very fast at the beginning and showed about ∼30% degradation after 3 days. A similar tendency was observed in the increase in R_S of PEDOT:PSS/Si with GO passivation layer. These results led to the decrease of J_SC and V_OC of the PEDOT:PSS/GO/Si device. It appears that the GO layer helps to supply the moisture and oxygen between PEDOT:PSS and Si interface because of higher oxygen functional groups caused by lower C/O ratio, as shown in Fig. 3. By contrast, the PEDOT:PSS/Si with the inserted rGO-700, rGO-800, and rGO-900 passivation layers show a slight degradation in PCE in 31 days. Interestingly, the J_SC values of PEDOT:PSS/Si with the rGO passivation layer are stable after one-month, and the V_OC values remain at about 90–80% when all devices are kept in air ambient. When the R_S of PEDOT:PSS/Si with rGO passivation layer are calculated from the I–V characteristics on dark condition by eqn (6), the obtained values exhibit a marginal variation over 30 days, as shown in Fig. 5d. The R_S values of PEDOT:PSS/rGO/Si devices increase by a smaller amount than those of PEDOT:PSS/Si device. This is because of the stable performance of the Si-organic solar cell with the inserted rGO passivation layer.^23,25,26 We also confirmed that the surface of rGO layer shows a small variation in the water contact angle as a function of storage time in air ambient, which is depicted in Fig. 3. Considering that the oxygen or moisture diffuse through the PEDOT:PSS layer, the inserted rGO passivation layer prevents the transport of oxygen or moisture on Si substrates, as compared to the GO layer. It can be assumed that the rGO layer helps to interrupt oxidation between PEDOT:PSS and Si interface by oxygen and moisture.²⁹ Moreover, it is possible that the rGO layer can serve as barriers for both liquid and vapor permeants because the typical distance between successive rGO layers is ∼0.3 nm, when the oxygen and moisture pass through the defects in rGO passivation layer stacked between PEDOT:PSS and Si.^54–57 Zhang et al. reported that the methyl group termination (–CH₃) suppressed the formation of SiO_x on Si surface in Si-organic solar cells, and the performance of organic-Si solar cell over time remained unchanged owing to the CH₃-termination.⁵⁸ In this study, the oxidation of Si substrate could be inhibited by use of rGO layer, results in extending lifetime of device with rGO passivation layer. It is well established that the PEDOT:PSS/Si solar cell with rGO layer enhances the air stability of the package-free device as the rGO layer with various annealing temperature play a critical role as a passivation layer in the PEDOT:PSS/Si interface.

4. Conclusions

We investigated the effect of the annealing temperature an rGO layer inserted in PEDOT:PSS/Si hybrid solar cell. The stability of PEDOT:PSS/GO/Si solar cell rapidly deteriorated in air ambient during one-month. The PEDOT:PSS/Si device with rGO-700, 800, and 900 layers shows lower J_SC, V_OC, and PCE than PEDOT:PSS/Si solar cell without rGO layer. On the other hand, the PCE of PEDOT:PSS/rGO/Si solar cell is more stable than that of PEDOT:PSS/Si device, as observed from the J–V and I–V characteristics in the air during one-month. The J_SC values of PEDOT:PSS/rGO/Si solar cell only slightly decreased after one-month, and the R_S slightly increased. Additionally, we observed that the rGO layer with high-temperature thermal annealing showed a very small variation in water contact angle over one-month. Consequently, the rGO layer provides not only a surprisingly high level of surface passivation but also prevents corrosion between PEDOT:PSS layer and Si substrate. Our study shows that the thermally annealed rGO layers are suitable as a surface oxidation suppression layers without being packaged, and they provide low-cost production and long-term stability for inorganic–organic hybrid solar cells.

Acknowledgements

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (Grant No. 2013R1A1A2013044 and 2015R1A6A1A04020421).

References

1. T. Mishima, M. Taguchi, H. Sakata and E. Maruyama, Sol. Energy Mater. Sol. Cells, 2011, 95, 18–21.
2. S. Avasthi, S. Lee, Y.-L. Loo and J. C. Sturm, Adv. Mater., 2011, 23, 5762–5766.
3. A. Kumar, S. Sista and Y. Yang, J. Appl. Phys., 2009, 105, 094512.
4. J. Schmidt, V. Titova and D. Zielke, Appl. Phys. Lett., 2013, 103, 183901.
5. L. He, C. Jiang, H. Wang, D. Lai and Rusli, Appl. Phys. Lett., 2012, 100, 073503.
6. W.-R. Wei, M.-L. Tsai, S.-E. Ho, S.-H. Tai, C.-R. Ho, S.-H. Tsai, C.-W. Liu, R.-J. Chung and H. He Jr., Nano Lett., 2013, 13, 3658–3663.
7. F. Zhang, T. Song and B. Sun, Nanotechnology, 2012, 23, 194006.
8. R. Liu, S.-T. Lee and B. Sun, Adv. Mater., 2014, 26, 6007–6012.
9. K. Norman, M. V. Madsen, S. A. Gevorgyan and F. C. Krebs, J. Am. Chem. Soc., 2010, 132, 16883–16892.
10. J. Sheng, K. Fan, D. Wang, C. Han, J. Fang, P. Gao and J. Ye, ACS Appl. Mater. Interfaces, 2014, 6, 16027–16034.
11. J. Chen, E. Cornagliotti, X. Loozen, E. Simoen, J. Vanhellemont, J. Lauwaert, H. Vrielinck and J. Poortmans, J. Appl. Phys., 2011, 110, 126101.
12. J. Schmidt, A. Merkle, R. Brendel, B. Hoex, M. C. M. van de Sanden and W. M. M. Kessels, Prog. Photovoltaics, 2008, 16, 461–466.
13. W.-T. Hsu, Z.-S. Tsai, L.-C. Chen, G.-Y. Chen, C.-C. Lin, M.-H. Chen, J.-M. Song and C.-H. Lin, Nanoscale Res. Lett., 2014, 9, 696.
14. Q. Xu, T. Song, W. Cui, Y. Liu, W. Xu, S.-T. Lee and B. Sun, ACS Appl. Mater. Interfaces, 2015, 7, 3272–3279.
15. D. Konios, G. Kakavelakis, C. Petridis, K. Savva, E. Stratakes and E. Kymakis, J. Mater. Chem. A, 2016, 4, 1612–1623.
16. D. Yang, L. Zhou, L. Chen, B. Zhao, J. Zhang and C. Li, Chem. Commun., 2012, 48, 8078–8080.
17. M. M. stylianakis, D. Konios, G. Kakavelakis, G. Charalambidis, E. Stratakis, A. G. Coutsolelos, E. Kymakis and S. H. Anastasiadis, Nanoscale, 2015, 7, 17827–17835.
18. J. Liu, Y. Xue, Y. Gao, D. Yu, M. Durstock and L. Dai, Adv. Mater., 2012, 24, 2228–2233.
19. A. Agresti, S. Pescetelli, L. Cinà, D. Konios, G. Kakavelakis, E. Kymakis and A. D. Carlo, Adv. Funct. Mater., 2016, 26, 2686–2694.
20. N. Vaenas, D. Konios, T. Stergiopoulos and E. Kymakis, RSC Adv., 2015, 5, 107771–107776.
21. M. Sygletou, P. Tzourmpakis, C. Petridis, D. Konios, C. Fotakis, E. Kymakis and E. Stratakis, J. Mater. Chem. A, 2016, 4, 1020–1027.
22. D. Konios, C. Petridis, G. Kakavelakis, M. Sygletou, K. Savva, E. Stratakis and E. Kymakis, Adv. Funct. Mater., 2015, 25, 2213–2221.
23. X. Liu, H. Kim and L. J. Guo, Org. Electron., 2013, 14, 591–598.
24. B. Paci, G. Kakavelakis, A. Generosi, V. R. Alvertini, J. P. Wright, C. Ferrero, D. Konios, E. Strataks and E. Kymakis, RSC Adv., 2015, 5, 106930–106940.
25. J.-M. Yun, J.-S. Yeo, J. Kim, H.-G. Jeong, D.-Y. Kim, Y.-J. Noh, S.-S. Kim, B.-C. Ku and S.-I. Na, Adv. Mater., 2011, 23, 4923–4928.
26. Y.-J. Jeon, J.-M. Yun, D.-Y. Kim, S.-I. Na and S.-S. Kim, Sol. Energy Mater. Sol. Cells, 2012, 105, 96–102.
27. H. P. Kim, A. R. B. M. Yusoff and J. Jang, Sol. Energy Mater. Sol. Cells, 2013, 110, 87–93.
28. Y.-J. Noh, S.-C. Park, Y.-T. Hwang, J.-H. Choi, S.-S. Kim, C.-H. Jung and S.-I. Na, Carbon, 2014, 79, 321–329.
29. X. Gao, J. Jang and S. Nagase, J. Phys. Chem. C, 2010, 114, 832–842.
30. W. S. Hummers Jr. and R. E. Offemen, J. Am. Chem. Soc., 1958, 80, 1339.
31. S. Jeong, E. Garnett, S. Wang, Z. Yu, S. Fan, M. Brongersma, M. McGehee and Y. Cui, Nano Lett., 2012, 12, 2971–2976.
32. J.-H. Hyung, G.-S. Kim, J.-T. Jeong, I.-H. Choi, C.-H. Hong, S. Yun, J.-H. Koh and S.-K. Lee, J. Nanoelectron. Optoelectron., 2015, 10, 229–233.
33. A. Misra, H. Kalira and A. Kottantharayil, ACS Appl. Mater. Interfaces, 2014, 6, 786–794.
34. B. D. Ryu, J.-H. Hyung, M. Han, K. B. Ko, Y. J. Park, T. V. Cuong, J. Cho and C.-H. Hong, Mater. Sci. Semicond. Process., 2016, 44, 1–7.
35. J. Rafiee, M. A. Rafiee, Z.-Z. Yu and N. Koratkar, Adv. Mater., 2010, 22, 2151–2154.
36. G.-T. Kim, S.-J. Gim, S.-M. Cho, N. Koratkar and I.-K. Oh, Adv. Mater., 2014, 26, 5166–5172.
37. M. Acik and Y. J. Chabal, J. Mater. Sci. Res., 2013, 2, 101–112.
38. Metal–semiconductor contacts, in Physics of Semi-conductor Devices, ed. S. M. Sze and K. K. Ng, John Wiley & Sons, Hoboken, 3rd edn, 2007.
39. X. Li, H. Zhu, K. Wang, A. Cao, J. Wei, C. Li, Y. Jia, Z. Li, X. Li and D. Wu, Adv. Mater., 2010, 22, 2743–2748.
40. E. Shi, H. Li, L. Yang, L. Zhang, Z. Li, P. Li, Y. Shang, S. Wu, X. Li, J. Wei, K. Wang, H. Zhu, D. Wu, Y. Fang and A. Cao, Nano Lett., 2013, 13, 1776–1781.
41. A. M. Nardes, M. Kemerink, M. M. De Kok, E. Binken, K. Maturova and R. A. J. Janssen, Org. Electron., 2008, 9, 727–734.
42. N. T. Khoa, S. W. Kim, D.-H. Yoo, E. J. Kim and S. H. Hahn, Appl. Catal., A, 2014, 469, 159–164.
43. J. Liu, M. Curstock and L. Dai, Energy Environ. Sci., 2014, 7, 1297–1306.
44. S.-S. Li, K.-H. Tu, C.-C. Lin, C.-W. Chen and M. Chhowalla, ACS Nano, 2010, 4, 3169–3174.
45. R. Garg, N. K. Kutta and N. R. Choudhury, Nanomaterials, 2014, 4, 267–300.
46. K. W. Kim, W. Song, M. W. Jung, M.-A. Kang, S. Y. Kwon, S. Myung, J. Lim, S. S. Lee and K.-S. An, Carbon, 2015, 82, 96–102.
47. M. Zhu, X. Li, Y. Guo, X. Li, P. Sun, X. Zang, K. Wang, M. Zhong, D. Wu and H. Zhu, Nanoscale, 2014, 6, 4909–4914.
48. A. C. Ferrari, Solid State Commun., 2007, 143, 47–57.
49. Z. Wang, D. Xu, Y. Huang, Z. Wu, L. Wang and X. Zhang, Chem. Commun., 2012, 48, 976–978.
50. Thermal reduction of graphene oxide, in Physics and Applications of Graphene-Experiments, Nanotechnology and Nanomaterials, ed. S. Mikhailov, 2011.
51. L. G. Cançado, K. Takai, T. Enoki, M. Endo, Y. A. Kim, H. Mizusaki, A. Jorio, L. N. Coelho, R. Magalhães-Paniago and M. A. Pimenta, Appl. Phys. Lett., 2006, 88, 163106.
52. M. G. Walter, X. Liu, L. E. O'Leary, B. S. Brunschwing and N. S. Lewis, J. Phys. Chem. C, 2013, 117, 14485–14492.
53. A. Guerrero, P. P. Boix, L. F. Marchesi, T. R.-Sanchis, E. C. Pereira and G. G.-Belmonte, Sol. Energy Mater. Sol. Cells, 2012, 100, 185–191.
54. O. C. Compton, S. Kim, C. Pierre, J. M. Torkelson and S. T. Nguyen, Adv. Mater., 2010, 22, 4759–4763.
55. R. R. Nair, H. A. Wu, P. N. Jayaram, I. V. Grigorieva and A. K. Geim, Science, 2012, 335, 442–444.
56. F. Guo, G. Silverberg, S. Bowers, S.-P. Kim, D. Datta, V. Shenoy and R. H. Hurt, Environ. Sci. Technol., 2012, 46, 7717–7724.
57. Z. Liu, J. Li and F. Yan, Adv. Mater., 2013, 25, 4296–4301.
58. F. Zhang, B. Sun, T. Song, X. Zhu and S. Lee, Chem. Mater., 2011, 23, 2084–2090.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra12441k

‡ B. D. Ryu and J.-H. Hyung contributed equally in this work.

---