Effect of Mn Doping on Properties of CdS Quantum Dot-Sensitized Solar Cells

Quantum dot-sensitized solar cells (QDSSCs) have received extensive attention in recent years due to their higher theoretical conversion efficiency and lower production costs. However, the photoelectric conversion efficiency of QDSSCs is still lower than the DSSCs because of the severe recombination of electrons of quantum dots conduction band. In order to improve the photoelectric conversion efficiency of QDSSCs, impurity element Mn is doped into the precursor solution of cadmium sulfide (CdS). By optimizing the experimental parameters, the photoelectric conversion efficiency of QDSSCs can be greatly improved. For the deposition of a fixed number of six times, the photoelectric conversion efficiency shows the maximum value (1.51%) at the doped ratio of 1 : 10.


Introduction
Quantum dot-sensitized solar cells (QDSSCs) have been recently drawing great attention due to the characteristics of quantum dots (QDs), such as high absorption coefficient, tunable band gap, and multiple exciton generation (MEG) effect [1][2][3][4]. However, its photoelectric conversion efficiency is still low compared to the theoretical value. The main factors limiting the efficiency of QDSSCs include the carriers recombination with redox couple on the semiconductor interface, a slower rate of hole transport, and the properties of the electrode. Recently, many efforts on the improvement of quantum dots (QDs), electrolyte, and electrode have made quite a lot of progress [5][6][7]. Among those researches doping optically active transition metal ions, such as Mn, Mg, Co, Cu, and In, has achieved obvious improvement in the photoelectric conversion efficiency of QDSSCs [8][9][10][11][12]. The doped system modifies the electronic and photophysical properties of QDs [13]. In addition, it is also possible to tune the optical and electronic properties of semiconductor nanocrystals by controlling the type of dopant [14]. Very recently, Santra and Kamat fabricated Mn-doped-CdS/CdSe quantum dot solar cells and yielded the efficiency (5.42%) for the QDSSCs [8]. And, the next year, Lee et al. reported on a PbS:Hg QD-sensitized solar cells with an extremely high photocurrent density (38 mA/cm 2 ) [15].
CdS, as an important II-VI semiconductor compound, with the band gap of 2.42 eV at room temperature [16], has good optical absorption properties in the visible region. Simultaneously, it has a large extinction coefficient and photochemical stability. Moreover, CdS quantum dots of different sizes can be synthesized to achieve broad spectral absorption through the relationship between the particle size and energy level of QDs. Therefore, the CdS quantum dots became a favorite QDs material for many researchers [17].
By the way of doping Mn atoms into CdS quantum dots, we can further regulate the properties of quantum dots for expanding the range of visible light absorption and improve the stability of cells [18,19].
In this paper, we introduced impurity element Mn 2+ into the precursor solution of cadmium sulfide (CdS). And Mn-doped-CdS QDs have been in situ deposited onto TiO 2 mesoporous substrates by the successive ionic layer adsorption and reaction (SILAR). Then, we assembled Mn-doped-CdS quantum dot-sensitized solar cells with the structure of photoanode\polysulfide electrolyte\Pt counter and incidentally discussed the impact of doping. We emphatically analyzed the effect of doped ratio and SILAR cycles on the properties of QDSSCs. At last, the power conversion efficiency of Mn-doped-CdS QDSSCs showed 1.51% under air mass (AM) 1.5 condition (100 mW/cm 2 ) with the optimized parameters (the doped ratio of Mn : CdS fixed on 1 : 10 and the six cycles of SILAR) obtained by experiment. Although the conversion efficiency is still low, it is much higher than the undoped CdS QDSSCs. And, with increased doping ratio, the efficiency had no further improvement but beginning to reduce.

Experimental
The TiO 2 nanoparticles are uniformly coated on the surface of the FTO (fluorine-doped tin oxide) by screen printing method and then annealed at 450 ∘ C for 30 min. And the counter electrode is prepared with Pt which is obtained from the thermal decomposition of chloroplatinic acid. Like the preparation of TiO 2 photoanode, the counter electrode has also been annealed at 450 ∘ C for 30 min. A mixed methanol Here, in this work, we used C 4 H 6 MnO 4 which contains impurity elements incorporated into the precursor solution by doping Mn in situ for the fabrication of Mn-doped-CdS QDs on the TiO 2 photoanode. In brief, specific concentration of the C 4 H 6 MnO 4 was mixed with Cd(NO 3 ) 2 (0.1 M) in the ethanol solution as cation source. Na 2 S (0.1 M) in methanol was used as anion source. The prepared TiO 2 photoanode was immersed in the mixed ethanol solution of C 4 H 6 MnO 4 and Cd(NO 3 ) 2 for 5 min, followed by rinsing with ethanol and drying with nitrogen. Subsequently, the TiO 2 photoanode was dipped into Na 2 S (0.1 M) methanol solution for 5 min at 30 ∘ C. And, then, one cycle of SILAR has been completed. During the experiment, to meet the research needs, we changed the ratio of molar concentration of C 4 H 6 MnO 4 in the cation source solution (i.e., doping ratio of Mn-doped-CdS).  The incident photon-to-electron conversion efficiency (IPCE) spectra of the doped and undoped QDSSCs at different incident light wavelengths were shown in Figure 2(b). The maximum IPCE value of two samples is about 67%. However, the spectral response range of Mn-doped-CdS QDSSCs is wider than the undoped one. This is due to the fact the introduction of impurities of Mn enhances the spectral response of the CdS QDSSCs, so that more electrons transferred to the external circuit and the generation of dark  current is also suppressed, which is consistent with the UVvisible absorption spectra of samples. The -characteristics of those QDSSCs samples based on diverse doped ratio are presented in Figure 3, with 4 cycles of SILAR. Simultaneously, the corresponding performance parameters of those samples are presented in Table 1.

Result and Discussion
As can be seen from the chart, Mn-doped-CdS QDSSCs exhibited considerable increment close to 58% (from 0.71% to 1.24%) in the power conversion efficiency, compared to the corresponding undoped sample. The highest overall power conversion efficiency (1.24%) was achieved with the most optimized Mn doped amount. And the actual Mn concentration, applied on an inductively coupled plasma optical emission spectroscopy (ICP-OES), shown in Table 2, was found to be 1.054% in the CdS film. Higher Mn concentration might damage the performance. Therefore, as the proportion of Mn doping gradually reduced, the sc increased from 4.13 mA/cm 2 (1 : 1) to 5.88 mA/cm 2 (1 : 10). But when the concentration continues to decrease, the performance of cells is declined. Meanwhile, the fill factor (FF) remained stable. The higher sc of the Mn-doped-CdS films is mainly attributed to the midgap states created by Mn doping, which not only enhances the capture for electrons and slows the speed of electrons transport but also screens partly the charge recombination with the hole and polysulfide electrolyte after oxidation. Indeed, the improved photovoltage in the QDSSCs with doped semiconductor films indicates that Mn-doping assists in electron accumulation within the QDs layers, thus shifting the Fermi level to more negative potentials and increasing the conduction band of CdS. As shown in Figure 4, the Fermi energy is adjusted to an even lower electric potential, due to the promotion of electron collection by doping.
In addition, the sc of the CdS QDSSCs will be improved by the Mn doping with the relatively high doped ratio. However, a large number of recombination centers, which have been introduced along with the Mn ions doped on the CdS QDs, are an unfavorable factor for the performance of solar cells. The enhancement for the sc by Mn doping increased firstly when the doping ratio is decreased. In contrast, as the doped ratio decreased, the enhancement is weakened. And the crystal defects caused by Mn doping still exist, so that the short-circuit current density and photoelectric conversion efficiency began to decline. In this experiment, we just changed the ratio of molar concentration of C 4 H 6 MnO 4 and Cd(NO 3 ) 2 . It has little effect on the fill factor, so there is no significant change about it. Based on the above discussion, the doped ratio is fixed on 1 : 10 as the most optimized Mn-doped concentration. Figure 5(a) gives the UV-Vis absorption spectra of four samples based on different SILAR cycles. It shows that the exciton absorption peak of those samples moved in a direction toward the long wavelength. And there is a sample exciton absorption peak at about 680 nm, which means manganese sulfur compounds may be formed during the deposition.
The IPCE recorded at different SILAR cycles for QDSC that employ three different photoactive semiconductor electrodes is shown in Figure 5(b). The overall photocurrent response matching the absorption features also increases along with the repetition of SILAR cycles.
The -characteristics and the corresponding performance parameters of those Mn-doped-CdS QDSSCs are presented in Figure 6 and Table 3. Obviously, as the times of SILAR cycles increased, there was a clearly different performance from the sc , FF, and oc . The best performance is obtained with the sample of 6 cycles of SILAR. When we International Journal of Photoenergy repeat the process of deposition by Mn-doped-CdS QDs, more QDs will be generated on the mesoscopic TiO 2 film, leading to the accumulation of photo-generated electron. Therefore, the quasi-Fermi level of Mn-doped-CdS QDs is shifted to more negative potentials like the effect of Mn doping. But excessive deposition of QDs caused stoppage of interspace of the mesoscopic TiO 2 film and severe recombination of electrons. So the sc is enhanced at first but got down finally. In contrast, because the conduction band of QDs increased with the repetition of SILAR cycles, the oc keep increasing.

Conclusions
In summary, the doping of CdS films with Mn has enabled us to achieve a significant improvement as compared to undoped films. After the optimization of experimental parameters, the photoelectric conversion efficiency of Mndoped-CdS QDSSCs shows the maximum value (1.51%). Although the performance is still lower than the expected value, it gives us another optional approach for making QDSSCs competitive.