ISSN 1063-7850, Technical Physics Letters, 2015, Vol. 41, No. 11, pp. 1094–1096. © Pleiades Publishing, Ltd., 2015. Original Russian Text © V.S. Levitskii, V.I. Shapovalov, A.E. Komlev, A.V. Zav’yalov, V.V. Vit’ko, A.A. Komlev, E.S. Shutova, 2015, published in Pis’ma v Zhurnal Tekhnicheskoi Fiziki, 2015, Vol. 41, No. 22, pp. 55–60.
Raman Spectroscopy of Copper Oxide Films Deposited by Reactive Magnetron Sputtering V. S. Levitskii, V. I. Shapovalov*, A. E. Komlev, A. V. Zav’yalov, V. V. Vit’ko, A. A. Komlev, and E. S. Shutova St. Petersburg State Electrotechnical University, St. Petersburg, 197376 Russia *e-mail:
[email protected] Received June 2, 2015
Abstract—Raman spectroscopy has been used to study the influence of partial oxygen pressure during deposition and isothermal treatment on the chemical composition of copper oxide films deposited by reactive dc magnetron sputtering of copper target in a reactive gaseous medium. Three series of films deposited at various partial oxygen pressures (from 0.06 to 0.16 mTorr) possessed different chemical compositions. The subsequent thermal treatment of all samples was performed for 30 min in air at a constant temperature in a 300‒500°C interval. An increase in the annealing temperature led to chemical changes in the films. After isothermal treatment at 450°C, the films in all series acquired stoichiometric CuO composition. DOI: 10.1134/S106378501511022X
Films of copper oxides possessing semiconductor properties provide a unique possibility of using them as active layers in solar cells, sensors, and other devices [1]. In the Cu–O chemical system, practical applications have been found for films with two compositions: Cu2O and CuO [2]. Oxide films have been obtained using a broad range of methods, which can be subdivided into chemical syntheses, thermal evaporation and deposition methods, and physical sputtering techniques [3]. Technologies based on the reactive magnetron sputtering are most widely employed [4–6]. Despite extensive investigations and vast literature devoted to the process of reactive magnetron sputtering, no attention has been paid to recently revealed [7] intermediate stationary regimes of copper target operation, in which the target surface is only partly covered with oxide. These regimes are not inherent in all metal targets and only take place for some metals (W, Cu, and others) in certain intervals of variation of the discharge current density J and partial oxygen pressure pO2 , which are characteristic of each particular technological system. Data on the influence of subsequent annealing in atmosphere on the composition of copper suboxide films are also not available. In the present work, we have used Raman spectroscopy to study the influence of the regime of copper target operation and the subsequent heat treatment in atmosphere on the chemical composition of films deposited on glass substrates by reactive dc magnetron sputtering in Ar + O2 gaseous medium. The films were deposited in a vacuum chamber with a volume of 7.8 × 10–2 m3 equipped with a planar
magnetron and target with a diameter of 115 mm. The residual gas pressure did not exceed 10–2 mTorr. The process was carried out at an argon pressure of 0.3 mTorr, current density J = 10 mA/cm2, and substrate temperature 240°C. Taking into account specific features of the reactive magnetron sputtering of copper targets [7], we have studied the influence of partial oxygen pressure by depositing films in various regimes of target operation. Three series of samples were prepared, the samples of which will be referred to below as 1, 2, and 3. The first two series were prepared in intermediate stationary regimes of copper target operation (at pO2 = 0.06 and 0.1 mTorr, respectively), which were revealed in preliminary experiments. The third series was prepared using the “oxide” regime of target operation ( pO2 = 0.16 mTorr), in which the entire target surface is covered with an oxide layer [8– 10]. In samples of series 1, 2, and 3, the film thicknesses were 0.4, 0.8, and 0.2 μm, respectively. The samples of all series were subjected to subsequent heat treatment for 30 min at temperatures 300, 350, and 400, 450, and 550°C (each sample was annealed only at one temperature). The chemical compositions of films were studied by the method of Raman scattering. The Raman spectra were measured at room temperature in the backscattering geometry on a LabRam HR800 spectrometer using second-harmonic radiation (wavelength, 532 nm) of a Nd:YAG laser. The laser beam was focused on the sample surface into a spot of ~1–2 μm diameter. The scattered light was collected and focused by a ×100 objective. In order to eliminate
1094
RAMAN SPECTROSCOPY OF COPPER OXIDE FILMS DEPOSITED
215
520 296 345
440
Raman spectra of crystalline and amorphous-crystalline Cu2O [11–14]. It was established that the Raman spectra of this oxide contained, in addition to the mode at 515 cm–1 (allowed by the selection rules), peaks of forbidden modes at 109, 154, and 635 cm–1, as well as overtones of the second (218 and 308 cm–1) and fourth (436 cm–1) orders [11, 12]. The appearance of forbidden modes such as those manifested by peaks at 109, 154, and 635 cm–1 results from violation of the rules of phonon selection with respect to the wave vector (caused by the frequency resonance with other excitons), which points to imperfection of the crystalline structure of Co2O and its possible amorphization [11, 13]. The obtained results allow the composition of films in samples of series 1 to be expressed by the formula Cu2Ox with x < 1.
630
Intensity, a. u.
93 110 148
1
2 3 200
400 600 Raman shift, cm−1
800
Fig. 1. Raman spectra of Cu–O films deposited at different partial oxygen pressures: (1) pO 2 = 0.06 mTorr, (2) pO 2 = 0.1 mTorr, and (3) pO 2 = 0.16 mTorr.
laser-radiation-induced modification of the samples studied, the laser radiation power density was controlled so as not to exceed 5 kW/cm2. Figure 1 shows the Raman spectra of Cu–O films as-deposited at different partial oxygen pressures. The spectrum of sample 1 (Fig. 1, curve 1) exhibits bands with maxima at 93, 110, 148, 215, 440, 520, and 630 cm–1. Analogous bands were observed in the
The spectrum of sample 2 (Fig. 1, curve 2) is close to that of sample 1, but the intensity of bands with maxima at 93 and 110 cm–1 is significantly increased. It can also be noted that the intensity of the band with a maximum at 440 cm–1 is significantly decreased. These changes indicate that the composition of films in samples of series 2 can be expressed by formula Cu2Ox with x ≈ 1. Significant changes are observed in the Raman spectrum of sample 3 (Fig. 1, curve 3), which displays bands with maxima at 296, 345, and 630 cm–1. The positions of band maxima in the spectrum of this sample are close to the corresponding values for CuO [15, 16]. Therefore, it can be ascertained that the composi-
PO2 = 0.1 mTorr
PO 2 = 0.06 mTorr
1095
PO2 = 0.16 mTorr
as dep
Intensity, a. u.
T = 300 °C T = 350 °C T = 400 °C T = 450 °C T = 550 °C 200 400 600 800 Raman shift, cm−1
200 400 600 800 Raman shift, cm−1
200 400 600 800 Raman shift, cm−1
Fig. 2. Raman spectra of Cu–O films as-deposited at different partial oxygen pressures and measured prior to annealing (asdeposited) or upon annealing in air for 30 min at various temperatures from 300 to 550°C. TECHNICAL PHYSICS LETTERS
Vol. 41
No. 11
2015
1096
LEVITSKII et al.
tion of films obtained at pO2 = 0.16 mTorr is close to CuO. However, this spectrum also contains an additional broad band within 400–600 cm–1, which may be related to the presence of an amorphous phase or a large number of structural defects in the crystalline lattice. Figure 2 compares the Raman spectra of as-deposited films to those of all samples upon isothermal annealing at various temperatures. As the annealing temperature increases, the spectra of all samples exhibit variations that ref lect changes in the chemical composition caused by the diffusion of oxygen. Changes in the composition of films deposited at pO2 = 0.06 mTorr can be expressed by the following sequence of conversions: Cu2Ox → Cu2Oδ → Cu2O → CuO1 – x → CuOδ, where x ≪ 1 and δ ≈ 1. During this, bands characteristic of CuO already appear in the spectra of samples annealed at 300°C. The spectra of films annealed in the interval of temperatures from 300 to 400°C are indicative of the coexistence of two phases: Cu2O and CuO. Upon the annealing at 400°C, the spectra show very weak bands of Cu2O, which is indicative if the termination of conversions. As the annealing temperature is increased further, only CuO is retained. In samples of series 2, the sequence of conversions is as follows: Cu2Oδ → Cu2O → CuO1 – x → CuOδ, where x ≪ 1 and δ ≈ 1. Already upon the annealing at 400°C, the spectra show evidence of the appearance of bands of CuO, while, at T = 450°C, the phase conversions are terminated by the formation of CuO. The spectra of samples representing series 3 contain the bands of CuO1 – x (x < 1) from the very beginning. As the annealing temperature is increased from 300 to 400°C, the intensity of maxima in the region of 400–600 cm–1 decreases, which indicates a decrease in the fraction of amorphous phase and reduction in the amount of structural defects. Annealing of these samples yields films with CuO composition. Thus, using the method of Raman spectroscopy, we have established that (i) The method of reactive dc magnetron sputtering in regimes corresponding to intermediate stationary states of the copper target allows suboxide films with
compositions Cu2O1 – x (x ≤ 1) to be obtained, while the oxide regime yields films with a composition close to stoichiometric CuO. (ii) Subsequent annealing in air initiates chemical transformations in as-deposited films, which, irrespective of the initial composition, are terminated at 450°C within 30 min with the formation of CuO films. Acknowledgments. This study was supported in part by the Russian Science Foundation, project no. 1519-00076. REFERENCES 1. Z. Ping, Z. Yurong, Y. Qingbo, et al., J. Semicond. 35, 103001 (2014). 2. A. S. Zoolfakar, R. A. Rani, A. J. Morfa, et al., J. Mater. Chem. C 2, 5247 (2014). 3. V. I. Shapovalov, Glass Phys. Chem. 36 (2), 121 (2010). 4. L. Wong, S. Chiam, J. Huang, et al., J. Appl. Phys. 108, 033702 (2010). 5. A. Ogwu and T. Darma, J. Appl. Phys. 113, 183522 (2013). 6. S. S. Noda, H. Shima, and H. Akinaga, AIP Conf. Proc. 1585, 9 (2014). 7. T. S. Shutova, A. E. Komlev, V. I. Shapovalov, et al., Proceedings of the 12th Int. Conf. “Films and Coatings2015” (St. Petersburg, Russia), pp. 157–159. 8. A. E. Komlev, V. I. Shapovalov, and N. S. Shutova, Tech. Phys. 57 (7), 1030 (2012). 9. A. A. Barybin and V. I. Shapovalov, J. Appl. Phys. 101, 054905 (2007). 10. A. A. Barybin, A. V. Zavyalov, and V. I. Shapovalov, Glass Phys. Chem. 38 (4), 396 (2012). 11. D. Powell, A. Compaan, and J. R. Macdonald, Phys. Rev. B 12, 20 (1975). 12. Y. P. Yu and Y. R. Shen, Phys. Rev. B 12, 1377 (1975). 13. J. C. W. Taylor and C. L. Weichman, Can. J. Phys. 49, 601 (1971). 14. P. F. Williams and S. P. S. Porto, Phys. Rev. B 8, 1782 (1973). 15. M. H. Chou, S. B. Liu, C. Y. Huang, et al., Appl. Surf. Sci. 254, 7539 (2008). 16. J. Chrzanowski and J. C. Irwin, Solid State Commun. 70, 11 (1989).
TECHNICAL PHYSICS LETTERS
Translated by P. Pozdeev
Vol. 41
No. 11
2015