Ayat Nasr El-Shazly‡
*a,
Ahmed Esmail Shalan‡*a,
Mohamed Mohamed Rashada,
Elsayed Ali Abdel-Aala,
Ibrahim Ahmed Ibrahima and
Mohamed F. El-Shahatb
aCentral Metallurgical Research and Development Institute, P.O. Box 87, Helwan, 11422, Cairo, Egypt. E-mail: a.shalan133@gmail.com; a.shalan@cmrdi.sci.eg; ayatelshazly@gmail.com; Fax: +202-25010639; Tel: +202-25010640-43
bChemistry Department, Faculty of Science, Ain Shams University, 11566, Cairo, Egypt
First published on 2nd July 2018
Solid-state dye-sensitized solar cells (ss-DSSCs) comprising Sn2+-substituted ZnO nanopowder were purposefully tailored via a co-precipitation method. The solar cells assembled in this work were sensitized with N719 ruthenium dye and insinuated with spiro-OMeTAD as a solid hole transport layer (HTL). Evidently, significant enhancement in cell efficiency was accomplished with Sn2+ ions-substituted ZnO photoelectrodes by maintaining the weight ratio of SnO at 5%. The overall power conversion efficiency was improved from 3.0% for the cell with pure ZnO to 4.3% for the cell with 5% SnO substitution. The improvement in the cell efficiency with Sn2+-substituted ZnO photoelectrodes is attributed to the considerably large surface area of the nanopowders for dye adsorption, efficient charge separation and the suppression of charge recombination provided by SnO. Furthermore, the energy distinction between the conduction band edges of SnO and ZnO implied a type II band alignment. Moreover, the durability as well as the stability of 15 assembled cells were studied to show the outstanding long-term stability of the devices made of Sn2+ ion substituted ZnO, and the PCE of each cell remained stable and ∼96% of its primary value was retained for up to 100 h. Subsequently, the efficacy was drastically reduced to ∼35% after 250 h of storage.
![]() | ||
Fig. 1 (a) Schematic diagram and (b) energy level diagram of the FTO/TiO2 BL/Zn1−xSnxO/N719 dye/spiro-OMeTAD/Al device. |
Fig. 2a shows the XRD patterns of the ZnO nanoparticles with 0, 5, 10, and 20% weight ratios of SnO, thermally treated at 500 °C for 1 h. The obtained diffraction patterns of the nanopowders suggest that ZnO is crystalline with a hexagonal wurtzite structure. All the XRD patterns showed that no excess impurity phases were present, evidencing that the as-prepared nanopowders were of high purity. The pure ZnO nanoparticles (black color) have a hexagonal structure, as indexed in JCPDS # 89-1397. The substitution of Zn2+ with different Sn2+ content did not change the hexagonal crystal structure of the wurtzite phase (Fig. 2a; red, purple and blue colors). The diffraction peaks of the 5 wt% SnO-containing nanopowders correspond to the wurtzite ZnO, and there was no SnO peak. With an increase in SnO content to 10 and 20%, peaks from both wurtzite ZnO and tetragonal SnO phases (JCPDS # 01-085-0423) were observed. The crystallinity of the ZnO samples was distorted as the SnO weight ratio was increased, due to degradation in grains as well as crystallite quality. This result can be ascribed to the decrease in the extended to localized state transitions resulting from the band tails formed owing to defects or disorders.36 The change in the degree of crystallinity of the composite framework was followed by a moderate change in the absorption edges, as will be discussed while describing the optical properties. Moreover, Fig. 2b clearly depicts the cross section-scanning electron microscopy analysis of the solar cells consisting of N719 Ru dye loaded on the Sn2+-substituted ZnO with a 5% weight ratio electrode and infiltrated with spiro-OMeTAD, which exhibited the maximum power conversion efficiency. This confirms the layer by layer formation of the cell. To determine the elemental composition of the as-prepared nanopowders, we performed (EDX) analysis, as illustrated in Fig. S1(a–c) in the ESI†. EDX analysis indicated that the different molar ratios of Zn to Sn were 1:
0.05, 1
:
0.1 and 1
:
0.2, which were close to the theoretical values. The atomic percentage of tin (Sn) ions in the framework of the formed Zn1−xSnxO nanocomposite was 5, 10 and 20%, as confirmed by the EDX spectrum and indicated in the inset tables of Fig. S1(a–c).†
The mean particle size of the ZnO–SnO nanocomposites was determined by high resolution transmission electron microscopy (HRTEM), as presented in Fig. 3a–d. Evidently, by comparing the four HRTEM images, we see that the particle size of Sn2+-substituted ZnO samples (Fig. 3b–d) are smaller than that of the pure ZnO sample (Fig. 3a). The smaller the particle size of the nanoparticles, the more porous is the generated film, resulting in higher specific surface area and greater amount of dye loading. Furthermore, the inset of Fig. 3a displays the crystallinity of the obtained powders via selected area electron diffraction (SAED), which was used to determine the structure of the samples.36 Moreover, the obtained results verify the good crystallinity of the pure ZnO sample compared to the Zn1−xSnxO samples. Remarkably, the diameter of the ZnO nanoparticles was ∼25 nm and that of Sn2+-substituted ZnO was ∼15 nm, which are in good agreement with the crystallite sizes obtained from the XRD profiles.
The BET surface area of the as-prepared nanopowders was substantial for the DSSCs due to its impact on the dye loading at the surface of the photoelectrode. A large number of small pores were recognized between the nanoparticles in all samples, suggesting the presence of a mesoporous structure. This phenomenon was assessed by N2 sorption analysis. The N2 adsorption–desorption isotherm of the pure ZnO and Sn2+-substituted ZnO with different Sn content exhibited a type IV isotherm with type H2 hysteresis loop (Fig. 4a and b), which is exemplary for mesoporous materials based on the IUPAC classification.39 The detailed physical characteristics of the pure ZnO and the Sn2+-substituted ZnO samples with different Sn content are given in Table 1. The specific surface area of Sn2+-substituted ZnO is relatively high in comparison with that of pure ZnO, which is attributed to the smaller particle sizes of Zn1−xSnxO and is fully consistent with the TEM data.40 In order to obtain more evidence for the presence of Sn ions in the formed nanocomposites, we conducted the XPS analysis to determine the chemical components and changes, as shown in Fig. S2(a–c), ESI.† The obtained results illustrate the spectra of the Zn1−xSnxO in the wide 8 range as shown in Fig. S2a† and assigned the existence of (Zn, O and Sn) element peaks. Besides, the narrow scan of ZnO and SnO shows the elemental composition for both of them (Fig. S2b and c†). The Zn 2p3/2 peak, (Fig. S2b†) at approximately 1021.6 eV, is assigned to the Zn–O bonds,25 which indicates the formation of ZnO composite comprising SnO nanoparticles via the solution processable wet chemical method. The XPS spectra of Sn 3d (Fig. S2c†) indicates the presence of spin orbit components, 3d3/2 and 3d5/2 at binding energies of 486.86 and 493.56 eV, respectively.25 These results confirm the existence of Sn2+. Notably, Sn ions can enhance the electron transfer between the dye molecules and the photoanodes due to its strong chemical activity through the formation of coordination bonds and oxygen vacancies to enhance the quantity of dye adsorption.25 Fig. 5a and b depict the absorption spectrum and band gap energy of pure ZnO and Sn2+-substituted ZnO with different Sn content. It is well known that ZnO is transparent and has an absorption band centered at around 3.15 eV.40 The Tauc model is the easiest tool to calculate the band gaps of Sn2+-substituted ZnO with different Sn concentrations.41 The optical band gap energy decreased slightly from 3.15 to 3.05 eV when the Sn content was 5%, i.e., the band gap red shifted, but remained constant with the increase in Sn percentage. The insignificant redshift (∼0.1 eV) of the band gap is ascribed to both the compositional change and the structural modification of the composite materials.36 Incorporating about 5–20% weight ratio of SnO, as an ionized donor, produces deep states in the band gap that lead to an increase in the scattering of photons by the crystal defects created by the substitution, thus reducing the band gap.42,43 More evidence for the formation of Zn1−xSnxO nanocomposites and the positive effect on the cells net efficiency can be obtained by studying dye loading through the absorbance of the dye on each photoanode used in this study, as shown in Fig. S3, ESI.† The results indicate that the amount of dye absorbed by the formed nanocomposite is enhanced compared to the amount that is absorbed by ZnO or SnO nanoparticles alone. This confirms that the charge separation are enhanced and the resistance as well as charge recombination decreases due to the limited penetration of the dye molecules caused by the blocking effect induced by the coated Zn1−xSnxO nanocomposites. The difference in the absorption results clearly indicate and confirm the formation of the nanocomposite.
![]() | ||
Fig. 4 (a) Nitrogen adsorption–desorption isotherms surface area and (b) pore size distribution of pure ZnO, Zn0.95Sn0.05O, Zn0.9Sn0.1O and Zn0.8Sn0.2O. |
Sample | SBET (m2 g−1) | Pore volume (cm3 g−1) | Pore size (nm) |
---|---|---|---|
Pure ZnO | 34.04 | 0.36 | 42.3 |
Zn0.95Sn0.05O | 49.12 | 0.35 | 29.3 |
Zn0.9Sn0.1O | 61.37 | 0.41 | 26.8 |
Zn0.8Sn0.2O | 65.16 | 0.39 | 24.1 |
Fig. 6a shows the characteristic J–V curves of the devices fabricated using pure ZnO and Sn2+-substituted ZnO with different SnO weight ratios. Furthermore, Table 2 summarizes the photovoltaic parameters for different cells: short circuit photocurrent (Jsc), open-circuit voltages (Voc), fill factor (FF) and solar conversion efficiency (η). It was found that the power conversion efficiency (η) enhanced with the addition of SnO, reaching a maximum of 4.3% when the weight ratio of SnO was minimum (5%), with a current density (JSC) of 12.45 mA cm−2 and an open circuit voltage (Voc) of 0.740 V, which is almost one and half times larger than for pure ZnO. According to the results obtained for the different substituted Zn1−xSnxO samples, there is no significant change in Voc upon substitution by SnO due to the fixed difference between the Fermi level of Zn1−xSnxO and the oxidation–reduction potential of HTL. Therefore, the reason for the different performances is mainly due to the different BET surface areas induced by Sn2+ substitution. These results can be explained by the electron transfer from the N719 dye into the conduction band (CB) of Zn1−xSnxO. This might be induced due to the coverage of ZnO particles over the SnO particles which favoring the charge injection from the low unoccupied molecular orbital (LUMO) of N719 to the CB of Zn1−xSnxO. Moreover, the photogenerated hole (h+) transfer takes place from the high occupied molecular orbital (HOMO) of the N719 dye to the VB of HTL.44,45 Increased photocurrent requires good electron transport, large surface area, better dye loading and excellent layer connection strength.46 By increasing the weight ratio of SnO to ZnO by 10 and 20%, the efficiency of the cell reduces to 3.9 and 3.4%, respectively. This occurs because the substitution of excess Sn in the ZnO lattice which leads to the formation of a separate SnO phase, resulting in a network of n–n heterojunctions in the ZnO system.47 As a result, the free carrier density is decreased, leading to reduce conductivity. Given the parameters of our assembled cells, we can conclude that the difference in fill factor in the tested cells may be related to the difference in shunt and series resistance of the cells. Every cell offers different resistance depending on its structure formation and causes a difference in FF. Moreover, the recombination current in the space charge region of the cell, which is confirmed by the obtained low efficiency, may be one of the reasons behind the difference in FF between the assembled cells. Subsequently, the cell efficiency and the conductivity can be controlled by changing the SnO ratio in the ZnO crystal lattice, which agrees with the results obtained from XRD analysis.42
![]() | ||
Fig. 6 (a) J–V curves, (b) IPCE spectra and (c) Nyquist diagrams of the impedance spectra of pure ZnO, Zn0.95Sn0.05O, Zn0.9Sn0.1O and Zn0.8Sn0.2O. |
Samples | JSC/mA cm−2 | VOC/V | FF | η/% |
---|---|---|---|---|
ZnO | 8.31 | 0.706 | 51.10 | 3.0 |
Zn0.95Sn0.05O | 12.45 | 0.740 | 46.70 | 4.3 |
Zn0.9Sn0.1O | 11.40 | 0.719 | 47.60 | 3.9 |
Zn0.8Sn0.2O | 10.76 | 0.719 | 43.90 | 3.4 |
Fig. 6b describes the current conversion efficiencies (IPCE) of ZnO and Sn2+-substituted ZnO. The ss-DSSC based on the Sn2+-substituted ZnO electrode exhibits higher IPCE value over a wide range (from 300 to 700 nm) than pure ZnO. It was reported that IPCE linearly increases with an increase in dye loading. The main functional parameter affecting IPCE is Jsc, which is well correlated with increased dye loading. In fact, typical increase in Jsc occurs in ss-DSSCs by increasing the optical density of the semiconductor electrode, which enhances the amount of photogenerated charge with a minor effect on Voc. It can be noticed that the ZnO sample with 5% SnO to ZnO weight ratio exhibits the most significant light harvesting achieved by the dye, compared to that achieved by Sn2+-substituted ZnO with 10 or 20% weight ratio and pure ZnO. Thus, we can conclude that the highest IPCE occurred because of the good assembly of dye molecules on the semiconductor electrode, while the small amount of light absorbed by the dye on the pure ZnO electrode is attributed to the deficiency in dye aggregation on the surface of the ZnO electrode.48 Fig. 6b also shows the integrating photo-current using IPCE results over the AM1.5 solar spectrum at irradiance 100 mW cm2. The differences between the current estimated from the IPCE integration and the J–V measurements are within the experimental uncertainties. We also note that we tested these cells in air without encapsulation: there is typically a slightly degraded performance for these cells following testing.17 To study and obtain better insight into the kinetics and dynamics of the interfacial charge transfer process within the assembled cells, electrochemical impedance spectroscopy (EIS) measurements were performed. Fig. 6c shows the Nyquist plots of the cells based on the different photoanodes used in this study. The results show semicircles for every photoanode that correspond to the electron recombination at the photoelectrode/dye/electrolyte interface. Also, they show that the recombination resistance decreases in the order of pure ZnO > Zn0.8Sn0.2O > Zn0.9Sn0.1O > Zn0.95Sn0.05O photoelectrode prepared using the facile wet chemical method. There is a good relationship between the charge transfer and the net efficiency of the assembled cells. The sample with the higher surface area and faster electron transfer rate had higher cell efficacy and vice versa.
The reproducibility of the devices was tested by fabricating 15 identical devices according to the same experimental procedure for each electrode. The corresponding photovoltaic parameters for the ZnO and Sn2+-substituted ZnO devices with ratios of 5, 10 and 20% are summarized in Tables S1–S4, in the ESI,† respectively. Fig. 7a shows histograms of the PCE distributions for the devices made with these electrodes for comparison; the corresponding histograms for the other photovoltaic parameters are shown in Fig. S4, ESI.† Our results show that the performances of these devices showed good reproducibility with an average PCE (%) = 2.98 ± 0.02, 4.22 ± 0.04, 3.85 ± 0.04, and 3.30 ± 0.04 for devices made of ZnO and Sn2+-substituted ZnO with ratios of 5, 10 and 20%, respectively. To test their long-term stability, we also checked the durability of the devices with different ratios under dark conditions; the corresponding stability results are described in Fig. 7b. Clearly, the cells display good stability over a time period of 250 h. The efficiency of each cell remained stable and 96% of its primary value was retained until 100 h. Subsequently, the efficiency decreased to 35% after 250 h of storage.
![]() | ||
Fig. 7 (a) The PCE distribution histogram and (b) the PCE evolution of encapsulated devices stored in ambient air for different time periods [pure ZnO, Zn0.95Sn0.05O, Zn0.9Sn0.1O and Zn0.8Sn0.2O]. |
Footnotes |
† Electronic supplementary information (ESI) available: EDX analysis of the produced nanocomposites; XPS analysis and optical absorbance of pure ZnO, pure SnO and the Zn1−xSnxO nanocomposite. Photovoltaic performance histograms of 15 ss-DSSCs devices manufactured using pure ZnO, Zn0.95Sn0.05O, Zn0.9Sn0.1O and Zn0.8Sn0.2O, and tables summarizing the corresponding photovoltaic parameters. See DOI: 10.1039/c8ra02852d |
‡ A. N. El-Shazly and A. E. Shalan contributed equally to this manuscript. |
This journal is © The Royal Society of Chemistry 2018 |