Figure 5.5 shows the IV curves of the QD-DSSC devices integrated with those Zn-CIS-coated ZnO nanowires, which are assembled into a layer-type film. The relevant solar-cell parameters for the three samples are given in Table 5.1, which include the current density at short circuit (JSC in mA cm−2), the voltage at open circuit (VOC in V), the fill factor (FF), and the efficiency of power conversion (η = JSC·VOC·FF/Pin with Pin = 100 mW cm−2).
A systematic trend with both VOC decreasing upon increasing the n value (i.e., decreasing Zn doping) was disclosed. This result is similar to the work of Braunger et al, [120] who reported that the incorporation of small amounts of Zn led to an increased VOC. However, the value of FF is lower than that received in TiO2 DSSC solar cell (0.6-0.7), which is attributed to recombination between photoexcited carriers in the ZnO nanowires and tri-iodide ions in the electrolyte.
n=1 n=2 n=3
400 500 600 700 800
n=1 n=2 n=3
Normalized Absorbance (a.u)
Wavelength/nm
(a) (b)
(c)
Figure 5.4 Photographs of different Cu/Zn ratio Zn-CIS quantum dots (a) in toluene, (b) anchored on ZnO NWs films. (c) Absorption spectra of different Cu/Zn ratios Zn-CIS quantum dots anchored on ZnO nanowires plate.
Because of the improvements in Jsc, and FF, the overall efficiency of conversion of photons to current exhibits a systematic increase from η = 0.23 % at n = 1 to η = 0.28 % at n = 2, but shows a decrease at n = 3 (η =0.24 %). As can be seen, two important factors restrict the efficiency in the case of n = 1. The first limitation is the low FF, which indicates that electron transport in the ZnO nanowires has been obstructed. This obstruction can be attributed to two explanations: one is the presence of Zn2+-ZCIS aggregates arisen from the unstable surface chemistry of ZnO nanowires electrode. During the assembly process, the acidic Zn-CIS QDs partially dissolve ZnO surface to form Zn2+-ZCIS aggregates which deposited along the nano-pore walls in electrode and thus interfered the Zn-CIS QDs in generation of the photoelectrons. [121]The other one is the higher driving force for electron
0.0 0.1 0.2 0.3 0.4 0.5
injection. The conduction band edge of Zn-CIS with n = 1 is higher than others due to more amount of Zn doping. Therefore, a higher injection resistance exists at the ZnO/Zn-CIS interface and the photocurrent decreases more rapidly with the opposite potential applied, leading to a lower FF. Although the former situation exists in all three cases (n = 1~3), the later one is expected to dominate and explain the low FF.
The second limitation to restrict the efficiency (n = 1) is the slightly lower Jsc, which is attributed to insufficient photocurrent probably due to the lack of full spectrum of light absorption. However, it shows a slight reduction of the efficiency when n = 3, which is due to the decreasing Zn doping restricted open-circuit voltage and resulted in a low efficiency.
Figure 5.5 Current-voltage characteristics of Zn-CIS QDs-DSSC devices with different n value under stimulated AM 1.5 solar illumination (100 mW/cm2) and active area 0.28 cm2.
Comparing with the efficiency reported by Leschkies et al,[109] who obtained 0.4% power-conversion efficiency and short-circuit current 2.1 mA/cm2 for a system of CdSe QDs with ZnO nanowires, the best power-conversion efficiency of current study has reached a lower value of 0.28%. However, while being normalized with those technical data in terms of the length of the ZnO nanowires, the average current
density is 0.342 mA/cm2 per μm length, which is higher than that reported for the CdSe QDs with 12μm ZnO nanowires system having a value of 0.175 mA/cm2 per μm length. Since it is reasonably to believe the population of the QD deposited on the nanowires will influence the resulting power-conversion efficiency, where the higher number of the QD on the nanowire surface, the higher the photocurrent can be produced while other variables are held constant, the average deposited number of the Zn-CIS QD on unit surface area of the ZnO nanowire is then estimated. In this case, the surface area of a 5 μm length single ZnO nanowire with 50 nm diameter is ca. 0.785 μm2, which can accommodate about 45700 Zn-CIS quantum dots with 3.5 nm diameter for a hexagonal closed packing configuration, converting to the average adsorption number per μm2 is about 58000. This is technically accessible since the microscopic observation readily showed a full coverage of the nanowire by the QDs.
For the case of Reference 7d, the calculated surface area and average absorption number is 4.7 μm2 and 68000 per μm2 (12 μm length and 125 nm diameter of ZnO nanowire, 3 nm diameter of CdSe), respectively, which is higher by 10 times in surface area, but only larger by about 10% in QD population compared to the nanowires employed in current study. However, in practice, the population of QDs deposited on 12-μm ZnO nanowire is sparingly distributed, rather than closely packed, which probably explained the power-conversion efficiency for the latter system is not as high as expected but only slightly higher, by about 7.5%, than current experimental outcome. Such a comparison strongly indicates that the overall power conversion efficiency of the QD-DSSCs should be largely limited by (1) the surface area of the nanowires available for QD deposition and (2) the deposited population of the QD. The length and the size of the nanowire are unlikely to be major dominating factors for controlling the efficiency. This result proved our hypothesis regarding the average current density for the QDs-ZnO nanowires solar
cell system.
Table 5.1 Photovoltaic performance of the Zn-CIS/ZnO DSSCs under AM1.5 illumination (Power 100 mW/cm2) and active area 0.28 cm2.
n=1 n=2 n=3
JSC /mA cm−2 1.48 1.71 1.54 VOC /V 0.44 0.34 0.32
FF 0.36 0.48 0.50
η(%) 0.23 0.28 0.24
5.4 Design of the ZnS shell as a dual-function layer
Based on the results reported above, there are two major competing pathways that reduced the performance of the Zn-CIS-coated ZnO nanowire device when the electrons transfer toward the ZnO nanowire electrodes. First recombination dynamic is attributed to the complex reaction between excited electrons in Zn-CIS excited state and the I3- in electrolyte. Second recombination dynamic results from the capture of these excited electrons by Zn-CIS cations, resulting in the QD regeneration reaction and no longer competing efficiently with this reaction. These faster recombination dynamics, however, result in an enhanced dark current for these devices and consequently limit open-circuit voltage and fill factor of the device, as evidenced by the current/voltage data. As was known, partial QD corrosion from electrolyte and charge recombination between ZnO and Zn-CIS might be the major origin for the reduced efficiency. The source of charge recombination might have been due to the band discontinuity between ZnO and Zn-CIS. Given these considerations, a key strategy for optimization of the performance of Zn-CIS-based device is the minimization of interfacial charge recombination losses. To address this issue, a novel core-shell configuration with ZnS as a shell on the Zn-CIS QD (as core) was designed and the shell is acting as
0.0 0.1 0.2 0.3 0.4 0.5
both a protecting layer and a band-bridge layer.
Figure 5.6 shows I-V characteristics of two QD-DSSCs recorded during illumination with 100 mW/cm2 simulated AM1.5 spectrum. One of these cells (hollow circle) was constructed using the core-shell QDs whilst the other without the ZnS shell (solid circle, sample with n=2) for comparison. The inset parameters are the Voc, Jsc, FF, and η of the ZnS-coated sample. It was found that both the power conversion efficiency (0.66%) and the short-circuit current (3.12 mA/cm2) of the QD-DSSC modified by ZnS coating exhibited much higher values than those obtained without the coating. Noticeably, the efficiency was promoted more than 2 times after ZnS coating, which indicated that ZnS coating effectively enhanced the performance of the QD-based solar cell. The value of Voc, Jsc and FF were also increased after the ZnS coating, which suggests that the ZnS coating is able to effectively eliminate the excited electrons recombination. The influence of ZnS layer deposition for this QD-based solar will be elucidated in more detail in a later section.
Figure 5.6 Current-voltage characteristics of Zn-CIS QDs-DSSC devices fabricated with (spot) ZnS coating and (square) without. Samples were measured under stimulated AM 1.5 solar illumination (100 mW/cm2) and active area 0.28 cm2.
Under open-circuit conditions, it is believed that electrons accumulate within the ZnS shell under visible irradiation, which shifts the apparent Fermi level to negative potentials. Once the illumination is removed, the accumulated electrons are slowly discharged because they are scavenged by the defects on the ZnO nanowire surface.
On this basis, the electrons injected from the excited Zn-CIS should be survived in a longer time length and hence facilitate electron transport without undergoing any possible loss at grain boundaries.
To prove this assumption, we further employed picosecond emission decay measurement to analyze the excited electron lifetime, in order to gain better understanding on the recombination dynamics of excitons.