Here, we present a protocol for niobium oxide films deposition by reactive sputtering with different oxygen flow rates for use as an electron transport layer in perovskite solar cells.
Reactive sputtering is a versatile technique used to form compact films with excellent homogeneity. In addition, it allows easy control over deposition parameters such as gas flow rate that results in changes on composition and thus in the film required properties. In this report, reactive sputtering is used to deposit niobium oxide films. A niobium target is used as metal source and different oxygen flow rates to deposit niobium oxide films. The oxygen flow rate was changed from 3 to 10 sccm. The films deposited under low oxygen flow rates show higher electrical conductivity and provide better perovskite solar cells when used as electron transport layer.
The sputtering technique is widely used to deposit high-quality films. Its main application is in the semiconductor industry, although it is also used in surface coating for improvement in mechanical properties, and reflective layers1. The main advantage of sputtering is the possibility to deposit different materials over different substrates; the good reproducibility and control over the deposition parameters. The sputtering technique allows deposition of homogeneous films, with good adhesion over large areas and at low-cost when compared with other deposition methods like chemical vapor deposition (CVD), molecular beam epitaxy (MBE) and atomic layer deposition (ALD)1,2. Commonly, semiconductor films deposited by sputtering are amorphous or polycrystalline, however, there are some reports on epitaxial growth by sputtering3,4. Nevertheless, the sputtering process is highly complex and the range of the parameter is wide5, so in order to achieve high-quality films, a good comprehension of the process and parameter optimization is necessary for each material.
There are several articles reporting on the deposition of niobium oxide films by sputtering, as well as niobium nitride6 and niobium carbide7. Among Nb-oxides, niobium pentoxide (Nb2O5) is a transparent, air-stable and water-insoluble material that exhibits extensive polymorphism. It is an n-type semiconductor with band gap values ranging from 3.1 to 5.3 eV, giving these oxides a wide range of applications8,9,10,11,12,13,14,15,16,17,18,19. Nb2O5 has attracted considerable attention as a promising material to be used in perovskite solar cells due to its comparable electron injection efficiency and better chemical stability compared to titanium dioxide (TiO2). In addition, the band gap of Nb2O5 could improve the open-circuit voltage (Voc) of the cells14.
In this work, Nb2O5 was deposited by reactive sputtering under different oxygen flow rates. At low oxygen flow rates, the conductivity of the films were increased without making use of doping, which introduces impurities on the system. These films were used as electron transport layer in perovskite solar cells improving the performance of these cells. It was found that decreasing the amount of oxygen induces the formation of oxygen vacancies, which increases the conductivity of the films leading to solar cells with better efficiency.
1. Etching and cleaning the substrate
2. Deposition of niobium oxide films
3. Constructing the solar cells
In the sputtering system, the deposition rate is strongly influenced by the oxygen flow rate. The deposition rate decreases when the oxygen flow is increased. Considering the present conditions of the target area used and plasma power, it is observed that from 3 to 4 sccm there is an expressive decrease on the deposition rate, however, when the oxygen is increased from 4 to 10 sccm it becomes less pronounced. In the regime of 3 sccm the deposition rate is 1.1 nm/s, decreasing abruptly to 0.1 nm/s for 10 sccm as seen in Figure 1.
The niobium oxide phase formed is dependent on the oxygen flow rate. For flows less than 3 sccm, niobium dioxide (NbO2) is the main phase formed. For flows higher than 3.5 sccm the oxygen amount is too high to originate NbO2, instead, Nb2O5 is observed as the main phase (Figure 2). Electron microscopy images (Figure 2) show the nanometric spherical particles of the films deposited at 3.5, 4 and 10 sccm. In contrast, the film deposited at 3 sccm shows sheets shape particles.
The films deposited by reactive sputtering in different oxygen flow rates show different electrical properties. The conductivity of the films increases when less oxygen is used, 3 sccm or less. Increasing the oxygen flow rate to 3.5, 4 and 10 sccm, a decrease in the conductivity is observed (Figure 3A,B). This represents a simple and easy way to increase the conductivity of oxide films by adjusting the flux of oxygen during film deposition.
The niobium oxide films deposited by sputtering were used as electron transport layer (ETL) in perovskite solar cells. For these solar cells, the film deposited at 3 sccm was not used because transparency is essential for ETLs. The performance of the solar cells also depends on the niobium oxide used (Figure 4). The cell made with the films deposited at 3.5 sccm has the best performance with the highest short-circuit current, a clear influence of the ETL film’s properties on the final performance of the cells.
Figure 1: Deposition rate as a function of the oxygen flow rate during niobium oxide films deposition.
Images of the oxide films surfaces are shown as insets. This figure has been modified from Fernandes et al.20. Please click here to view a larger version of this figure.
Figure 2: Electron microscopy images and X-ray difractograms of niobium oxide films deposited under different oxygens flow rates, 3 sccm (A), 3.5 sccm (B), 4 sccm (C) and 10 sccm (D).
The main NbO2 (JCPDS #82-1142) and Nb2O5 (JCPDS #28-317) peaks are indicated. The other peaks are referred to FTO. This figure has been modified from Fernandes et al.20. Please click here to view a larger version of this figure.
Figure 3: Current vs voltage of diferent niobium oxide films (A), and corresponding conductivity (B).
This figure has been modified from Fernandes et al.20. Please click here to view a larger version of this figure.
Figure 4: Schematic architecture of the solar cell devices (A), J-V curves of the perovskites solar cells using niobium oxide films deposited at different oxygen flow rate, 3.5 sccm (B), 4 sccm (C) and 10 sccm (D).
This figure has been modified from Fernandes et al.20. Please click here to view a larger version of this figure.
The niobium oxide films prepared in this work was used as electron transport layer in perovskite solar cells. The most important characteristic required for an electron transport layer is to prevent recombination, blocking holes and transferring efficiently electrons.
In this respect the use of reactive sputtering technique is advantageous since it produces dense and compact films. Also, as already mentioned, compared to sol-gel, anodization, hydrothermal, and chemical vapor deposition synthesis methods14,21,22, reactive sputtering is the most suitable to deposit large areas1,2,14. However, understanding the role of the deposition parameters on film properties is a challenge5,15,20, especially in the case of niobium-oxide that can form many different stable crystal structures.
Nb can be found in tree oxidation state as II, IV and V, which are predominant in NbO, NbO2, and Nb2O5 phase respectively14. Although niobium pentoxide (Nb2O5) is the most stable phase, controlling the amount of oxygen in the chamber during deposition can produce different phases. This is a critical and important step, requiring fine control. In our system oxygen flow of 3 sccm favors the formation of NbO2. The use of oxygen flow rate higher than 3 sccm leads to the formation of Nb2O5.
An excess of oxygen in the chamber leads to oxygen contamination of the target. This results in a decrease in oxide film deposition rate and the formation of different phases as explained in a previous publication20. On the contrary, oxygen deficiency in the chamber significantly decreases film transparency. Besides the formation of different phases, changing the oxygen flow results in films with different density of oxygen vacancies. This leads to significant changes in film properties, for example its conductivity. NbO2 shows high conductivity, while Nb2O5 is a more resistive phase. The lower electrical conductivity of Nb2O5 compared to NbO2 films is explained by the chemical nature of the bond, Nb has a charge state of 5+, with all of its 4d electrons bond to O 2p-orbital. For the Nb2O5 films (films deposited with 3.5, 4 and 10 sccm of oxygen flow), the highest conductivity is observed in films deposited with 3.5 sccm of oxygen flow, which is attributed to an increase in oxygen vacancies20.
The cell made with a film deposited with 3.5 sccm of oxygen flow has the best performance with the highest short circuit current. This high performance is due to the better conductivity of the correspondent niobium oxide film. As the resistivity of the niobium oxide film increases, the devices show less efficiency.
Clearly, sputtering is a powerful deposition technique that allows a finer control of deposition parameters compared to other chemical deposition techniques. The major limitation of sputtering is the use of ultra-high vacuum required to avoid contaminants, which implies relatively long waiting times for pumping. The waiting time can be partially avoided in system equipped with a pre-chamber, or a differential pumping system. Nevertheless, this requirement allows the production of films of high purity.
In conclusion, the use of sputtering allows the formation of dense and compact films with controlled stoichiometry. In our case, good conductivity was achieved by adjusting the oxygen content in the chamber. Sputtering is a promissing technique to deposit film in large areas to produce efficient solar cells.
The authors have nothing to disclose.
The work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Centro de Desenvolvimento de Materiais Cerâmicos (CDMF- FAPESP Nº 2013/07296-2, 2017/11072-3, 2013/09963-6 and 2017/18916-2). Special thanks to Professor Máximo Siu Li for PL measurements.
2-propanol | Merck | 67-63-0 | solvent with maximum of 0.005% H2O |
4-tert-butylpyridine | Sigma Aldrich | 3978-81-2 | chemical with 96% purity |
acetonitrile | Sigma Aldrich | 75-05-8 | anhydrous solvent , 99.8% purity |
bis(trifluoromethane)sulfonimide lithium salt | Sigma Aldrich | 90076-65-6 | chemical with ≥99.95% purity |
chlorobenzene | Sigma Aldrich | 108-90-7 | anhydrous solvent , 99.8% purity |
ethanol | Sigma Aldrich | 200-578-6 | solvent |
Fluorine doped tin oxide (SnO2:F) glass substrate | Solaronix | TCO22-7/LI | substrate to deposit films |
Kaptom tape | Usinainfo | 04227 | thermal tape used to cover the substrates |
Kurt J Lesker magnetron sputtering system | Kurt J Lesker | —— | Sputtering equipment used to deposit compact films |
Lead (II) iodide | Alfa Aesar | 10101-63-0 | PbI2 salt- 99.998% purity |
methylammonium iodide | Dyesol | 14965-49-2 | CH3NH3I salt |
N2,N2,N2′,N2′,N7,N7,N7′,N7′-octakis (4-methoxyphenyl)-9,9′-spirobi [9H-fluorene]-2,2′,7,7′-tetramine | Sigma Aldrich | 207739-72-8 | Spiro-OMeTAD salt, 99% purity |
Niobium target of 3” | CBMM- Brazilian Metallurgy and Mining Company | —— | niobium sputtering target used in the sputtering system |
N-N dimethylformamide | Merck | 68-12-2 | solvent with maximum of 0.003% H2O |
TiO2 paste | Dyesol | DSL 30NR-D | titanium dioxide paste |
tris(2-(1H-pyrazol-1-yl)-4-tert-butylpyridine)cobalt(III) tri[bis(trifluoromethane)sulfonimide] | Dyesol | 329768935 | FK 209 Co(III) TFSL salt |