Abstract
Nowadays, the copper nitride (Cu3N) is of great interest as a new solar absorber material, flexible and lightweight thin film solar cells. This material is a metastable semiconductor, non-toxic, composed of earth-abundant elements, and its band gap energy can be easily tunable in the range 1.4 to 1.8 eV. For this reason, it has been proposed for many applications in the solar energy conversion field. The main aim of this work is to evaluate the properties of the Cu3N thin films fabricated by reactive radio-frequency (RF) magnetron sputtering at different RF power values to determine its potential as light absorber. The Cu3N films were fabricated at room temperature (RT) from a Cu metallic target at the RF power ranged from 25 to 200 W on different substrates (silicon and glass). The pure nitrogen flux was set to 20 sccm, and the working pressures were set to 3.5 Pa and 5 Pa. The XRD results showed a transition from (100) to (111) preferred orientations when RF power increased; the AFM images revealed a granular morphology, while FTIR and Raman spectra exhibited the characteristics peaks related to Cu-N bonds, which became narrower when the RF power increased. Finally, to stablish the suitability of these films as absorber, the band gap energy was calculated from transmission spectra.
Keywords: Copper nitride, reactive magnetron sputtering, pure nitrogen atmosphere, absorber, photovoltaic devices.
Introduction
The Cu3N is of great interest as a new solar absorber material, flexible and lightweight thin film solar cells [1, 2]. This increasing interest as absorber for photovoltaic (PV) applications is due to the fact that is a metastable, non-toxic semiconductor, composed of earth-abundant elements, and its band gap energy can be easily tunable in the range 1.4 to 1.8 eV [3] . These values mean that the Cu3N film can be considered as a possible substitute for the conventional silicon. Furthermore, this material can show both n and p-type character, which can be obtained by doping it with different elements that placed in interstitial position, such as fluorine [3].
Cu3N is a metastable semiconductor with an anti-ReO3 cubic crystal structure[3]. The structure exhibits nitrogen (N) atoms at the corners of the cube and copper (Cu) atoms in the middle of each edge. The material is thermally and structurally stable at temperatures up to 200°C [4]. However, its thermal decomposition occurs in the temperature range of 100–470 °C [5]. In this sense, it has been reported that at high temperatures, visible structural changes can be appreciated and they may be related to the release of pure Cu [6].
The Cu-N family of bonding compounds can offer different optoelectronic properties, i.e. Cu4N that shows metallic behaviour, while Cu3N presents a semiconductor one, where there is a strong relationship between its chemical bonds and its electronic properties. These different behaviours can be achieved depending on the fabrication technique employed and modifying the deposition conditions. Such different behaviours are basically attributed to the vacant interstitial sites, because Cu atoms do not occupy the face-centered cubic (FCC) close packing sites[7]. Thanks to this, it is possible to tune the optoelectronic properties by incorporating different atoms in such vacant sites. In addition, the films can show different properties depending on the technique used. Among the main methods used for the Cu3N deposition, we can find molecular beam epitaxy (MBE) [8], pulsed magnetron sputtering (PMS) [9], pulsed laser deposition (RPLD) [10], and reactive direct current (DC) magnetron sputtering [11] or reactive radio-frequency (RF) magnetron sputtering [12-15]. In this last case, the film stoichiometry and the properties can be easily changed. This can be achieved by modifying the main deposition parameters: (i) the gas pressure due to the interaction between N and Cu that could modify the band gap [6], (ii) the substrate temperature and the power that can affect the structural, electrical and optical properties of the films [16], (iii) the type of the substrate [13], and (iv) the substrate to target distance [8]. Among these parameters, the plasma gas composition and the gas pressure control the N fraction in the gas mixture, which has a strong impact on the film structure, i.e. the lattice constant ranged from 0.3815 to 0.3885 nm, and consequently, on their optoelectronic properties [17].
In view of the above, in this work we evaluate the effect of the RF power on the structural and optoelectronic properties of Cu3N thin films sputtered at room temperature (RT) in a pure N atmosphere, without introducing argon, to obtain a suitable material for being used as light absorber. It is expected to determine the optimized RF values to achieve the Cu3N films with appropriated morphologies, structures and band gap energies suitable for the chosen application.
Materials and Methods
The deposition of the Cu3N thin films were carried out with a commercial MVSystem INC. monochamber sputtering system, where there is only one gun that can be vertically adjustable and is RF operated. The target used is a pure commercial Cu target (99.99%) with a 3-inch diameter, from Lesker Company. In this work, we used two types of substrates: 1737 Corning glass (Corning Inc., New York, USA) and <100> polished n-type floating zone crystalline silicon (c-Si) wafers. Prior to be loaded into the sputtering chamber, both substrates were prepared in a different way. The native silicon dioxide on the surface silicon wafer is removed with a 1% HF solution in deionized water and ethanol for 5 minutes. In the case of the glass, it was cleaned for 10 minutes in ultrasound, with ethanol and deionized water, and finally immersed in isopropyl alcohol. Subsequently, the substrates were dried by blowing nitrogen over them. After loading the substrates, the chamber was pumped to a pressure of 10-5 Pa. The working pressure was set to 3.5 and 5.0 Pa, adjusting it with a “butterfly” valve. The process gas was pure N2 (purity 99.999%) at a flow rate of 20 sccm, controlled by a mass flow controller (MFC). The distance between the target and the substrate was set to 10 cm. Cu3N thin film depositions were performed at different powers ranged from 25 to 200W, and the deposition time was modified from 420 to 1800 s, depending the RF power value, to counteract the effect of the power on deposition rate.
The structure of the thin film was determined by X-ray diffraction (XRD), using a commercial Panalytical powder diffractometer model X’Pert MPD/MRD, with a Cu anode and secondary monochromator. The radiation used was Cu-kα radiation (λ=0.15406nm), and the scanned 2θ range was 10°-60° at a step size of 0.01°. To determine the chemical composition and molecular structure, Raman and Fourier transform infrared spectroscopy (FTIR) were used. Raman measurements were carried out with the dispersive spectrometer Horiba Jobin-Yvon LabRam HR 800 coupled, to an optical microscope Olympus BXFM, using a solid-state laser as an excitation source emitting at 532 nm. The Raman spectra were obtained with a laser power at the sample of about 5 mW and using a 100× microscope objectives. FT-IR measurements were carried out with a Perkin Elmer Spectrum 100 FT-IR in range of 400-4000 cm-1. The morphology of the films was studied by atomic force microscopy (AFM) using an AFM model III A multimode nanoscope (Bruker). The roughness of the samples was determined from the root mean square (RMS) calculated with the commercial Gwyddion software. Finally, the optical properties were calculated from the transmittance spectra in the range 300-2000 nm, measured at RT and normal incidence with a Perkin Elmer Lambda 1050 spectrophotometer.
Results and discussion
Figure 1 shows the XRD spectra of the Cu3N films deposited on glass at different RF powers and at the different working pressures of 3.5 and 5.0 Pa. Regardless the pressure used, the XRD patterns show Cu3N crystallites with an anti-ReO3 structure [18]. The Cu3N films deposited at low RF power values (below 150 W) present as preferential orientation the (100) plane that corresponds to the N-rich planes. As the RF power increases above 100 W, the intensity of the main (100) diffraction peak begins to decrease, while the (111) diffraction peak, corresponding to a Cu-rich plane, starts to appear as preferred from the sample deposited at 150 W [5]. In addition, this sample also presents several diffraction peaks of weaker intensity such as the (110) and (210) planes, referenced from the XRD card number 00-047-1088 (The Joint Committee on Powder Diffraction Standards, 47–1088). This change observed in the phase transition orientation from the preferred (100) plane to the (111) one as the RF power increases can be attributed to the presence of a larger number of Cu atoms in the plasma. The high presence of Cu atoms could cause the formation of a Cu3N material saturated with Cu, which retains a cubic structure that begins to disappear and/or change at very high RF powers.
On the other hand, the main effect that the total pressure exerts on the films is the amorphous character, observed in the sample deposited at the highest pressure and the highest RF power of 200 W. Upon increasing the deposition pressure from 3.5 Pa to 5.0 Pa produces more N ions, are bombarding the Cu target, leading to a higher amount of Cu ions or atoms in the plasma, with the Cu2+ and Cu+ionization states.The sample deposited at 200 W and 5.0 Pa shows amorphous structure, in comparison with that deposited at 200 W and 3.5 Pa. This can be attributed to a high presence of Cu atoms, because of the high RF power, that would reach the substrate surface with less energy due to the increase of the collisions between the particles within the plasma at the increased pressure The saturation effect of the Cu in the film at higher pressures is more evident. This is the reason why the Cu-rich phase disappears. In addition, under these extreme conditions, there is an increase in ion bombardment, resulting in more defects produced in the sputtered films. This indicates that the crystallinity of the film is seriously deteriorated due to the excessive RF power and subsequently, excessive amount of Cu atoms.
Conversely, at low RF power values, more N+2 and N+ free radicals are generated due to the high proportion of low-energy electrons that are formed in the plasma[15]. Hence, the samples deposited at low RF powers show the preferential N-rich planes, and a better crystal quality.