Abstract

Quantum dots are drawing great attention as a material for the next-generation solar cells because of the high absorption coefficient, tunable band gap, and multiple exciton generation effect. In search of the viable way to enhance the power conversion efficiency of quantum dot-sensitized solar cells, we have succeeded in preparing the quantum dot solar cells with high efficiency based on CdSe:X (Mn2+ or Cu2+) nanocrystal by successive ionic layer absorption and reaction. The morphological observation and crystalline structure of photoanode were characterized by field-emission scanning electron microscopy, X-ray diffraction, and the EDX spectra. In addition, the electrochemical performance of photoelectrode was studied by the electrochemical impedance spectra. As a result, we have succeeded in designing QDSSCs with a high efficiency of 4.3%. Moreover, the optical properties, the direct optical energy gap, and both the conduction band and the valence band levels of the compositional CdSe:X were estimated by the theory of Tauc and discussed details. This theory is useful for us to understand the alignment energy structure of the compositions in electrodes, in particular, the conduction band and valence band levels of CdSe:X nanoparticles.

1. Introduction

In recent years, inorganic semiconductor materials, which are also called Quantum Dots (QDs), have emerged as a powerful light-absorbing material that generates electrons for the quantum dot-sensitized solar cell (QDSSC) application. Various QDs were applied in the QDSSCs such as CdS, CdSe [1], PbS, PbSe [2, 3], and InP [1]. These had a number of advantages over dye molecules because the band gap could be controlled through particle size, modification [4], the high optical absorption coefficient [5], and generating several e–h+ pairs as absorbing photons [6]. So far, QDSSCs have still low optical conversion efficiency compared to dye-sensitized solar cells [711].

There are now a number of ways to improve performance of QDSSCs such as QDSSCs based on the CdS/CdSe combination synthesized by chemical bath deposition methods and successive ionic layer absorption and reaction (SILAR) adsorbed directly onto TiO2 nanoparticles [12, 13]. The improved performance was also achieved on core-shell QDs. Yu et al. reported the fabrication results on the CdS/CdSe core-shell system, resulting in higher current density, voltage, and performance than single QDs [13]. Yu said that the improved performance was done because of the reduced recombination processes at the surface trapping states of QDs and the increased electron mobility to TiO2 nanoparticles. In addition, the combination of electrons and holes in the contact surfaces and in semiconductor oxides such as TiO2 and ZnO causes the reduction in performance of QDSSCs. To reduce the recombination processes, Jung and colleagues coated CdS QDs with a ZnS layer to protect QDs from the electrolyte [1416]. This result was also confirmed in the QDSSCs based on PbS QDs with a CdS protection layer [17]. There were many methods that can reduce recombination in devices, but the most common method of surface treatment was used. Surface treatment means protection of QDs, which helped stabilize QDs in an electrolyte solution [18, 19].

Today, CdSe QDs are widely used by scientists in QDSSCs because of their easy manufacture, low cost, and high stability. However, their resistance is high and their CB energy is lower than that of TiO2 in the bulk material. These results do not facilitate the shift of electrons from CdSe QDs to the TiO2 film. By doping metal ions into QDs such as Cu [6], In [20], Co [21], Mn [22, 23], Hg [24], Eu [25, 26], La [27], and Mg [28], the electrical and optical properties of photoanode can be improved by the boosting absorption of photons of QDSSCs.

As far as reasons are concerned, QDSSCs based on X ions doped on CdSe nanoparticles with the different compositions of the Mn2+ ( between 0% and 40%) and the Cu2+ ( from 0% to 0.5%) by SILAR were illustrated. Moreover, the significant effects of X dopant on optical, physical, chemical, and photovoltaic properties of QDSSCs can be studied by the UV-Vis spectra and Tauc equation, which can determine the , CB, and VB positions of X ions doped on pure CdSe nanoparticles. This theory is useful for us to understand the alignment energy structure of the compositions in electrodes, in particular, CB and VB levels of CdS, CdSe:X nanoparticles. As a result, there is a rise or a drop of the CB and the VB levels when X content was changed. Correspondingly, the electrochemical impedance spectra were carried out to determine dynamic resistances in QDSSCs.

2. Experiment

In this experiment, TiO2/CdS and TiO2/CdS/CdSe:Cu2+ were synthesized similar to Ref [29]. In Ref [29], the thickness of TiO2/CdS/CdSe:Cu2+ was investigated to get the optimization. However, in this paper, we investigated the effect of Cu2+ molar concentrations as doping it into CdSe nanoparticle. Besides, we also prepared QDSSCs based on Mn2+ ion-doped CdSe nanoparticle and compared with that of QDSSCs based on TiO2/CdS/CdSe:Cu2+ photoanode.

2.1. TiO2 and TiO2/CdS Films Were Prepared As Follows Ref [29]

Briefly, a fluorine-doped tin oxide (FTO) glass substrate with sheet resistance 7 Ωsq−2 was used for the photoanode and the counter electrode. First, the FTO substrate was cleaned with ethanol for 30 min by ultrasonic, followed by deionized (DI) water for 15 min. The nanoporous TiO2 film was made on the well-cleaned substrate by the print method followed by sintering at 500°C for 30 min. TiO2/CdS film: TiO2 films were dipped into a Cd(CH3COO)2·2H2O ethanol solution for 5 min, rinsed with ethanol, and dried, then successively dipped into a Na2S·9H2O methanol solution for another 5 min, rinsed with methanol, and dried. The two-step dipping procedure was termed as one SILAR cycle. The process was repeated three times, and the obtained TiO2 film decorated with CdS QDs was named as the TiO2/CdS film [30].

2.2. A TiO2/CdS/CdSe:Mn2+ Film

Similarly, the Mn doped on CdSe QDs was attached to the TiO2/CdS film using SILAR. 0.740304 g Cd(NO3)2·2H2O and 0.47054 g Mn(CH3COO)2·2H2O were dissolved in 30 ml ethanol and DI water with ratio 1 : 1 to obtain 0.1 M of Cd2+ and Mn2+ solution. The process involved subsequent immersions of the TiO2/CdS film in a solution of 0.1 M of Cd2+ and Mn2+ solution for 5 min and then rinsed with ethanol. It then was immersed in 0.3 M of Se2- solution for 5 min at 50°C and rinsed with DI water. To accommodate the doping of Mn2+ ion, relevant molar concentrations of 0.1 mM, 0.2 mM, 0.3 mM, 0.4 mM, and 0.5 mM of Mn(CH3COO)2·2H2O were mixed with Cd(CH3COO)2·2H2O anion source. The process was repeated 3 times and the obtained TiO2/CdS film decorated with CdS/Mn-CdSe multilayers.

2.3. A TiO2/CdS/CdSe:Cu2+ Film

Preparation of TiO2/CdS/CdSe:Cu2+ photoanodes: briefly, the TiO2/CdS layers were immersed into Cd2+ and Cu2+ ionized solution (molar concentrations of 0.1 mM, 0.2 mM, 0.3 mM, 0.4 mM, and 0.5 mM of Cu(NO3)2·3H2O were mixed with Cd(CH3COO)2·2H2O anion source) for 5 min at room temperature, then rinsed with ethanol to remove excess precursors and dried. The films were next dipped into Se2- solution for 5 min at room temperature followed by rinsing with methanol and drying. This cycle was repeated 3 times [30]. The polysulfide electrolyte and Cu2S counter electrode were followed Ref [30].

2.4. Characterization

Field-effect scanning electron microscope (FESEM, 7401F) of Ho Chi Minh City Institute of Physics was used to investigate the surface morphology and composition of photoanode using the voltage of 10 kV. The UV-Vis spectra were characterized by the JASCO V-670 device of Applied Physical Chemistry Lab of University of Science, Vietnam National University-Ho Chi Minh City. The crystal structure was analyzed using an X-ray diffractometer (Philips, PANalytical X’Pert, CuK_radiation) and photocurrent-voltage measurements were performed on a Keithley 2400 source meter using simulated AM 1.5 sunlight with an output power of produced by a solar simulator (Solarena, Sweden). The electrochemical impedance spectroscopy (EIS) was carried out with the use of an impedance analyzer (ZAHNER CIMPS).

3. Results and Discussion

Figures 1(a) and 1(d) are FESEM images of TiO2/CdS/CdSe:Mn2+ and TiO2/CdS/CdSe:Cu2+ photoanodes with 20% and 0.3% doping concentration, respectively. The highly porous nearly spherical-shaped TiO2 nanoparticles can be observed very clearly from Figures 1(a) and 1(d) with an average size of approximately 70 nm. We can clearly see each layer of film, thickness of the FTO layer about 0.563 nm. The thickness of TiO2/CdS/CdSe:Mn2+ and TiO2/CdS/CdSe:Cu2+ photoanodes is approximately 12.056 μm and 12.675 μm, respectively. However, we did not observe both CdSe:Mn2+ and CdSe:Cu2+nanoparticles in the film because of their very small size. They can be filled in the space between the TiO2 nanoparticles and absorbed onto the TiO2 surface.

The compositional EDX analysis of the CdSe:X is shown in Figures 1(c) and 1(f), which confirm the presence of Mn2+ and Cu2+ dopant in photoanodes. As can be seen from Figures 1(g) and 1(h), they are the same structures of TiO2, CdS, and CdSe:Mn2+ (Cu2+). It is immediately obvious that the Raman spectroscopy of TiO2/CdS/CdSe:Mn2+ electrode shows the modes at 144 cm-1, 395 cm-1, 515 cm-1, and 636 cm-1positions corresponding to the TiO2 anatase [31]. Similarly, the above modes of TiO2 anatase in the Raman spectroscopy of TiO2/CdS/CdSe:Cu2+ electrode appear at 206 cm-1, 266 cm-1, 414 cm-1 and 622 cm-1positions. It is immediately obvious that there was a shift of the modes towards the high frequency due to the increase of CdSe:Cu2+-TiO2 associate compared to CdSe:Mn2+-TiO2 associate. Besides, there are one Longitudinal-Optical (1LO) and 2LO modes of CdSe:Mn2+ Zinc Blende at 201 cm-1 and 395 cm-1positions. However, the modes are shifted towards high frequency for Cu2+ ion-doped CdSe QDs. In addition, the 2LO mode of CdSe Cubic is shown in both Figures 1(g) and 1(h), but there was no 1LO mode. This implies that CdS, CdSe:Cu2+(Mn2+) nanoparticles have absorbed on TiO2 films.

To determine the of the TiO2/CdS/CdSe:Mn2+ and TiO2/CdS/CdSe:Cu2+ electrodes, UV-Vis absorption measurements are carried out. By analysis of the absorption coefficients for the electrodes, optical energy gap can be estimated using Tauc plot of versus (hv) and extrapolating of the linear portions of the curves to the energy axis according to [32]

is the absorption coefficient, which is determined by about 9.104 cm-1 [33], where is the thickness of the film and is the transmittance; is the reflectance. hv is the photon energy, is the direct band gap energy, and is a constant (about 2.059 10-2 cm-1) [33]. The Tauc plots are shown in Figure 2.

Moreover, the CB and VB of CdS, CdSe nanoparticles can also be determined by Tauc equation. The CB and VB of materials can be calculated if the electron affinity energy and the ionization energy are known. Similarly, the CB and VB of CdS and CdSe can be determined by Tauc equation [34]. The calculated parameters are listed in Tables 1 and 2.

Optical properties of undoped and doped photoanodes, which were prepared by SILAR method and calcined in a vacuum environment, were investigated by UV-Vis absorption spectra. Figures 2(a) and 2(b) show the UV-Vis absorption spectra of the photoelectrode with undoped (, ) and doped concentration of Mn2+ ions from 5% to 40% and Cu2+ ions from 0.1% to 0.5%. In general, there are two major differences between UV-Vis absorption spectra of undoped films and doped films such as, firstly, there is a very sharp increase in absorption spectra intensity when we doped Mn2+ ions from 5% to 40% and Cu2+ ions from 0.1% to 0.5% on CdSe QDs. This implied that the presence of metal ion energy levels in the band gap of CdSe QDs causes an increase in the absorption of incoming photons [35]. This result will be explained in more detail when we determine the CB energy and the VB of both CdSe:Mn2+ and CdSe:Cu2+ QDs. Secondly, there is a shift of the absorption peak in the red region compared to the undoped photoelectrodes (from 570 nm to 615 nm) [3638]. This is also understandable because of the change in band gap of CdSe QDs after doping; the CB position of CdSe QDs may be raised and its band gap may be narrowed. Figures 2(c) and 2(d) show the Tauc plot of Mn2+ ions from 5% to 40% and Cu2+ ions from 0.1% to 0.5% on CdSe QDs as shown in Tables 1 and 2. The band gap values of both CdSe:Mn2+ and CdSe:Cu2+ QDs reduced compared to the band gap of pure CdSe QDs.

Figure 3 shows photocurrent density-voltage curves of QDSSCs based on the metal ion-doped photoanodes measured under AM 1.5, 100 mW/cm2, and the photovoltaic parameters are listed in Table 3. The QDSSCs on undoped TiO2/CdSe films were measured with open circuit (), FF (0.38), short-current density (), and efficiency (). In general, the QDSSCs based on CdSe:Mn2+ and CdSe:Cu2+ QDs have better performance compared to the QDSSCs based on undoped films. When doping, open-circuit parameters, short-circuit currents, and efficiency increase due to the conduction band position of CdSe:Mn2+ and CdSe:Cu2+ QDs, which is raised higher than that of TiO2 semiconductor and listed in Tables 1 and 2. Similar to the UV-Vis absorption spectra, there was a strong increase in efficiency of the QDSSCs from 1.95% (with pure CdSe QDs) to 4.3% (with CdSe:Cu2+ QDs) and 3.8% (CdSe:Mn2+ QDs). The result of an increase performance is due to the effect of metal concentration impurities in CdSe QDs. This result is also explained by the calculation of CB energy levels, the VB of TiO2 semiconductor, CdSe:Mn2+ and CdSe:Cu2+ QDs [29, 39]. As we know, in the bulk material, the CB position of CdSe QDs (-4.5 eV) is lower than the CB of TiO2 (-4.3 eV), which is difficult for excited electrons to move from CdSe QDs to TiO2 nanoparticles [27]. However, after doping, the CB energy of CdSe:Mn2+ and CdSe:Cu2+ QDs is significantly increased from -4.5 eV to -4.05 eV (for CdSe:Cu2+ QDs) and -3.74 eV (for CdSe:Mn2+ QDs), which is higher than the CB of TiO2 nanoparticles. Therefore, like a waterfall, electrons can easily move from CdSe:X QDs to TiO2 film. This result also causes an increase in current density from 12.5 mA to 20 mA (CdSe:Cu2+ QDs) and 19 mA (CdSe:Mn2+ QDs).

Figures 4(a) and 4(b) are the EIS curves of photoanodes with the doped concentration of Mn2+ ions from 5% to 40% and Cu2+ ions from 0.1% to 0.5% under one-sun illumination and open circuit of the device; the parameter values shown are in Tables 3 and 4. Basically, the EIS curves have two semicircles corresponding to the resistance at the surface of the polyelectrolyte/counter electrode (denoted as ) and the diffuse resistance in the TiO2 film and TiO2/CdSe:X surface (denoted as ) [40]. After measuring the EIS curve, we used Nova software to fit and obtain the circuit corresponding to each EIS curve. From the circuit obtained, we determined and resistances, respectively. The lifetime of the excited electron () is determined by the formula , and the capacitance of QDSSCs () is determined by . The calculated values are presented in Tables 3 and 4. In general, the QDSSCs based on TiO2/CdSe:Mn2+ with the doped concentration from 5% to 40% and TiO2/CdSe:Cu2+ with the doped concentration from 0.1% to 0.5% have sharply changed the photovoltaic, which causes the change of and resistance values, the capacitance, and the excited electron lifetime of CdSe:X QDs. In particular, the and resistances correspond to pure CdSe QDs recorded approximately 630.5 Ω and 194.8 Ω; they are much higher than both the (254.4 Ω) and (8.68 Ω) resistances with the Cu2+-doped concentration from 0.1% to 0.5% and the (204.5 Ω) and (24.65 Ω) resistances with the Mn2+-doped concentration from 10% to 40%, while the excited electron lifetime and capacitances of QDSSCs are much lower [29]. This result implies that Cu2+ and Mn2+ doping into CdSe QDs can cause a decrease in diffusion resistance in TiO2 films and surfaces. Furthermore, the cause of the increase in performance may be due to the increase in the CB position of the CdSe:X QDs, which is higher than that of the TiO2 conduction band. In addition, the energy level of metal ion appears in the band gap of CdSe QDs, which reduces the band gap energy of CdSe:X QDs. This result is evidenced by the shift and an increase in the intensity of the UV-Vis absorption spectra (Figure 2). From the above reasons, we can show an increase in the current density and decrease the resistance of the device.

4. Conclusions

The QDSSCs based on TiO2/CdS/CdSe(3)Cu2+() with from 0% to 0.5% and TiO2/CdS/CdSe(3)Mn2+() with from 0% to 40% were successfully fabricated by the SILAR method. The TiO2/CdS/CdSe(3)Cu2+() cosensitized solar cells demonstrated better performance (4.3%) than the TiO2/CdS/CdSe(3)Mn2+() (3.8%) and TiO2/CdS(3)/CdSe(3) (1.96%). This result is a cause of the light harvesting for producing excitons, facility fast electron transfer at TiO2/CdS/CdSe:Mn2+(Cu2+) interfaces, and the reduced recombination in the QDSSCs as the concentration of Cu2+ (Mn2+) doped on the CdSe nanoparticles. In addition, the optical properties, the direct optical energy gap, and both the CB and VB levels of the compositional CdS, CdSe:Cu2+ and CdSe:Mn2+ were estimated by the theory of Tauc and discussed details. This theory is useful for us to understand the alignment energy structure of the compositions in electrodes, in particular, the CB and VB levels of CdS, CdSe:Cu2+, and CdSe:Mn2+ nanoparticles. As a result, there is a rise or drop of the CB and VB levels when Cu2+ (Mn2+) content was changed.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare no competing interests.

Acknowledgments

The authors would like to thank University of Science, Viet Nam National University Ho Chi Minh City, Vietnam.

Supplementary Materials

Supplemental Figure 1: FESEM image at 100 nm (a, d), cross section at 1 μm (b, e), and EDX of CdSe:Mn2+ (20%) and CdSe:Cu2+ (0.3%) (c, f). (g, h) Raman spectroscopy of CdSe:X-QDSSCs. The morphology of preparing as-photoelectrodes with 20% of Mn2+ and 0.3% of Cu2+ could be obtained from FESEM. In addition, the compositions and structure of films could be investigated from EDX and Raman spectra. In general, these are the main data sets, which provided all information about the morphology, sizes, and thickness of a sandwich layer and structure of preparing as-photoelectrodes. Supplemental Figure 2: UV-Vis spectra of (a) CdSe(3)Mn2+ and (b) CdSe(3)Cu2+. ()2 vs. () curves of (c) CdSe(3)Mn2+ and (d) CdSe(3)Cu2+ QDs. This is the supplementary material which supports for us to determine the band gap of sample as getting the peak of UV-Vis spectra, in particular, calculate the conduction band and valence band of material. Supplemental Figure 3: photocurrent density-voltage (J-V) curves of (a) CdSe:Cu2+-QDSSCs and (b) CdSe:Mn2+-QDSSCs. The performance of the quantum dot-sensitized solar cells could be obtained from the photocurrent density-voltage curves. These results support the discussion from the electrochemical impedance spectra. Supplemental Figure 4: Nyquist curves of QDSSCs. (a) Mn2+ concentration changing from 0% to 40% and (b) Cu2+ concentration changing from 0% to 0.5%. The data set obtained from the electrochemical impedance spectra such as Nyquist curves, Bode phase. They are used to determine resistance dynamic, lifetime, and capacitances of the photoelectrodes to correlate with the shift of excited electrons from quantum dots to TiO2 nanoparticles, the diffusion of electrons. (Supplementary Materials)