Enhanced photovoltaic performance in TiO2/P3HT hybrid solar cell by interface modification

    Corresponding author: Haizheng Tao, thz@whut.edu.cn
  • 1. State Key Laboratory of Silicate Materials for Architecture, Wuhan University of Technology, Wuhan 430070, China
  • 2. Faculty of Materials Science and Engineering, Hubei University, Wuhan 430062, China

Key words: TiO2solar cellinterface modification

Abstract: A TiO2/P3HT hybrid solar cell was fabricated by infiltrating P3HT into the pores of TiO2 nanorod arrays. To further enhance the photovoltaic performance, anthracene-9-carboxylic acid was employed to modify the interface of TiO2/P3HT before P3HT was coated. Results revealed that the interface treatment significantly enhances the photovoltaic performance of the cell. The efficiency of the hybrid solar cells reaches 0.28% after interface modification, which is three times higher compared with the un-modified one. We find that except for the increased exciton dissociation efficiency recognized by the previous reports, the suppressing of electron back recombination is another important factor leading to the enhanced photovoltaic performance.


1.   Introduction
  • Hybrid solar cells based on a conjugated conducting polymer and inorganic semiconductor have attracted much attention. It combines the advantages of high mobility, excellent chemical stability of metal oxides and the solution processable property of polymers[1, 2]. In a hybrid solar cell, the polymer serves as the donor and the inorganic semiconductor serves as the acceptor. When the cell is illuminated, excitons (bound electron-hole pairs) are generated in donor materials and dissociated into free electrons and holes at the interface between donor and acceptor, which lead to the photovoltaic effect[3, 4, 5, 6]. Up to now, polymers such as poly 3-hexylthiophene (P3HT), polyphenylene vinylenes (PPVs) and metal oxide nanostructures such as TiO$_{2}$, ZnO, and CdS have been utilized to fabricate hybrid solar cells[7, 8, 9, 10].

    However, the efficiency of HSCs is still far from satisfactory and cannot even compete with an organic solar cell or dye sensitized solar cell[11, 12]. In the past few years, great efforts have been devoted to improve the solar cell performances. Recently, it has been reported that the interfacial modification can improve the chemical compatibility between polymer and inorganic nanostructures and enhance the conversion efficiency of the hybrid solar cell[13, 14, 15]. For instance, Liu et al.[16] showed that the photovoltaic performance of a hybrid solar cell can be adjusted by modifying TiO$_{2}$ nanorods with different ligands. Liao and his co-workers[3] observed a three-fold efficiency enhancement when the TiO$_{2}$ nanostructure is modified by dye D149. As far as the mechanism for the improved photovoltaic performance is concerned, it is mainly ascribed to the increased exciton dissociation efficiency induced by the ligands.

    In this article, we report the improved efficiency of the TiO$_{2}$/P3HT hybrid solar cell through interface engineering by anthracene-9-carboxylic acid (ACA). The efficiency of the hybrid solar cell with an ACA interlayer was improved to 0.28 % from 0.06 %. Using time-resolved photoluminescence (TRPL) spectroscopy, electrochemical impedance spectroscopy (EIS) and open-circuit voltage decay (OCVD) measurements, we find that the suppressing of electron back recombination is another important factor leading to the enhanced photovoltaic performance, except for the increased exciton dissociation efficiency recognized by the previous report.

2.   Experimental details

    2.1.   Fabrication of TiO$_{2}$ nanorod arrays and devices

  • TiO$_{2}$ nanorod arrays were prepared by the hydrothermal method as reported in our previous work[4]. Briefly, FTO substrates, previously cleaned ultrasonically in a mixed solution of deionized water, acetone, and 2-propanol ($v$ : $v$ : $v$ $=$ 1 : 1 : 1) for 60 min, were immersed in a 0.15 M TiCl$_{4}$ aqueous solution for 30 min at 70 C, and subsequently annealed at 450 C for 30~min to prepare the TiO$_{2}$ seed layer. Then the FTO substrates were transferred into the precursor solution containing 24 mL of deionized water, 24 mL of concentrated hydrochloric acid (36 %-38 %) and 0.6 mL of titanium butoxide (TTIP, 97 %). The growth of the TiO$_{2}$ nanorod was conducted at 170~C for 6 h. ACA adsorption was carried out by immersing the TiO$_{2}$ nanorod arrays in a 0.5 M ACA ethanol solution for one day. P3HT was dissolved in 1, 2, 4-trichlorobenzene and infiltrated into the pores of the TiO$_{2}$ nanorod arrays by spin coating with a speed of 3000 rpm. The fabrication was performed in a glovebox. The Au counter electrode was deposited by magnetron sputtering with a shadow mask.

  • 2.2.   Characterization

  • Surface and cross-sectional morphologies of the hybrid solar cell were characterized by scanning electron microscopy (JEOL, JSM6510LV). The UV-vis diffuse reflection spectra were recorded on a Shimadzu UV-vis NIR spectrophotometer (UV-3600). TRPL spectroscopy was measured with a time correlated single photon counting spectrometer (Picoharp 300, Picoquant Inc.). A pulse laser (430 nm) with an average power of 1 mW operating at 40 MHz with a duration of 70 ps was used for excitation. Photocurrent-voltage measurements were performed on a Keithley 2400 source meter using a simulated AM 1.5 sunlight with an output power of 100 mW/cm$^{2}$ produced by a solar simulator (Newport 91192). The EIS were obtained using an AutoLab (IM6, Zahner). It was measured at the open circuit voltage of the cell under AM 1.5 sunlight illumination. The OCVD was measured by an electrochemical workstation (IM6) equipped with a short-interval (50-100 ms) sampling module. The cell was illuminated to a steady voltage and then the illumination was shut off and the open voltage was recorded.

3.   Results and discussion
  • Top and cross-sectional views of as-synthesized TiO$_{2}$ nanorod arrays are shown in Figure 1(a). TiO$_{2}$ nanorod arrays are densely vertically grown on the FTO substrate. The diameter is about 100 nm and the length is about 2.2 $\mu $m. Figure 1(b) is the SEM image of TiO$_{2}$ nanorod arrays after the spin coating of P3HT. We can see clearly that the pores of the TiO$_{2}$ nanorod arrays were fully infiltrated by P3HT, and a layer of P3HT was also formed on the top of the TiO$_{2}$ nanorod arrays.

    Figure 2 reveals the UV-vis diffuse reflection spectra of the sample. The pure TiO$_{2}$ nanorod arrays present an absorption edge at about 410 nm, corresponding to its band gap of 3.0~eV. After the modification by ACA, the light absorbance edge extends to 450 nm, which indicates that the ACA molecules are adsorbed on the TiO$_{2}$ nanorod successfully. The amount of ACA loaded on the surface of the nanorod was measured by the desorption method. In detail, we desorbed ACA on the nanorod by a NaOH solution, since it bonded to the nanorod by carboxylic, and then measured the absorption spectrum of the NaOH solution. According to Beer's law, the loading amount of ACA was determined to be 0.13 g/m$^{2}$. For the absorption spectrum of TiO$_{2}$/P3HT hybrid film, the absorption peak between 410 and 650 nm corresponds to the one of P3HT.

    The current density-voltage ($J$-$V)$ characteristics of the hybrid solar cell was measured under simulated AM 1.5 illumination and shown in Figure 3. Meanwhile, the parameters such as open-circuit potential ($V_{\rm OC})$, short-circuit current ($J_{\rm SC})$, fill factor (FF), and the photoelectron conversion efficiency ($\eta$) are summarized in Table 1. The $J_{\rm SC}$, $V_{\rm OC}$ and $\eta$ are 2.47 mA/cm$^{2}$, 0.51 V and 0.28 % when the TiO$_{2}$ nanorod arrays were modified with ACA; while the $J_{\rm SC}$, $V_{\rm OC}$ and $\eta$ for the TiO$_{2}$/P3HT solar cell without interface modification are 1.29 mA/cm$^{2}$, 0.26 V and 0.06 % only. The underlying mechanism for the improved photovoltaic performance by ACA modification was investigated through employing TRPL, OCVD and EIS.

    PL and TRPL were used to investigate the exciton dissociation process occurring at the TiO$_{2}$/P3HT interface. In Figure 4(a), the PL emission between 650 nm and 750 nm is due to the radiation recombination of P3HT. The PL intensity of TiO$_{2}$/ACA/P3HT is significantly reduced in comparison with the TiO$_{2}$/P3HT. The quenching of PL intensity indicates that part of the excitons is dissociated into free electrons and holes at the interface before the radiation recombination occurs. The enhanced exciton dissociation efficiency can be demonstrated further by the TRPL spectroscopy in Figure 4(b). As shown, the luminescence decays much more quickly after the interface modification of ACA. By exponential decay fitting, the average PL lifetimes of TiO$_{2}$/P3HT and TiO$_{2}$/ACA/P3HT hybrid films are induced to be 661 and 581 ps, respectively.

    Figure 5 displays the OCVD curves of the TiO$_{2}$/P3HT hybrid solar cell with and without interface modification. As shown, the open voltage decays promptly as the light illumination is off for the solar cell without interface modification; it becomes slower after the insertion of ACA. The decay of the open-circuit voltage was mainly caused by the recombination of free electrons in TiO$_{2}$ with the reductive species in P3HT. The decay-time constants ($\tau_{\rm n})$ were calculated to quantify the decay rate using the following equation and shown in the inset[5]:

    As seen in the inset, the decay-time constant is increased obviously when the interface modification by ACA is performed. So the slower the decay in TiO$_{2}$/ACA/P3HT means the slower the charge back recombination.

    The reduced electron back recombination is further demonstrated by EIS. Figure 6 shows the Nyquist plot of EIS results of TiO$_{2}$/P3HT and TiO$_{2}$/ACA/P3HT solar cells. Each plot is composed of one large semicircle. This type of impedance pattern belongs to the responses usually encountered in systems where the carrier transport is mainly determined by diffusion-recombination at the interface. The recombination resistance of EIS reflected the transfer and recombination process of electrons at the donor-acceptor interface[6]. By means of Z-view simulation, the recombination resistance of TiO$_{2}$/ACA/P3HT and TiO$_{2}$/P3HT solar cells is induced to be 90 k$\Omega $ and 28 k$\Omega $, respectively. It means that the electron recombination rate at the interface is decreased after the insertion of ACA. The decrease of electron back recombination is reasonable. The lowest unoccupied molecular orbital energy level is higher than the conduction band bottom of TiO$_{2}$[17]. When ACS is inserted between P3HT and TiO$_{2}$, a barrier is formed, which inhibits the back recombination of electrons.

    Through the above discussion, the role of the interfacial modifier ACA is summarized as: (i) The molecular structure of ACA consisted of anthracene and carboxylic acid group--the carboxylic group can bond to TiO$_{2}$ and the anthracene presents good compatibility with P3HT. In general, the ACA is like a linker and improves the compatibility between TiO$_{2}$ and P3HT. (ii) It enhanced the dissociation efficiency of excitons at the TiO$_{2}$/P3HT interface. (iii) It suppressed the electron back recombination.

4.   Conclusions
  • We have demonstrated the hybrid photovoltaic devices based on TiO$_{2}$/P3HT and TiO$_{2}$/ACA/P3HT. The performance of the corresponding solar cell can be significantly improved from 0.06 % to 0.28 % through interfacial modification. Various measurements have been employed to investigate the function of the ACA modifier. We find that the suppressing of electron back recombination is another important factor leading to the enhanced photovoltaic performance except for the increased exciton dissociation efficiency recognized by the previous report. Considering the unoptimized experimental process, the efficiency of the solar cell is lower even after interface modification. However, we report an effective method to improve the efficiency and revealed the underlying mechanism of why the efficiency is enhanced.

Figure (6)  Table (1) Reference (17) Relative (20)

Journal of Semiconductors © 2017 All Rights Reserved