Journal of the Korean Physical Society, Vol. 50, No. 4, April 2007, pp. 1099∼1103
Structural and Optical Properties of AgInSe2 Films Prepared on Indium Tin Oxide Substrates Jeoung Ju Lee,∗ Jong Duk Lee, Byeong Yeol Ahn, Hyeon Soo Kim and Kun Ho Kim Department of Physics and Research Institute of Natural Science, Gyeongsang National University, Jinju 660-701 (Received 5 January 2007) AgInSe2 films were prepared on the indium-tin-oxide (ITO)- coated glass substrates by using thermal evaporation. X-ray diffraction spectra showed that the AgInSe2 films were preferentially grown along the (112) orientation. The crystal structure was chalcopyrite with lattice constants a = 6.102 ˚ A and c = 11.69 ˚ A. X-ray photoelectron spectroscopy spectra showed that the asdeposited AgInSe2 film was Ag-rich while the AgInSe2 film annealed at 300 ◦ C was In-rich. From the atomic force microscope images, the root-mean-square roughness and the grain size decreased with increasing annealing temperature. The optical energy band gap, measured at room temperature, of the as-deposited AgInSe2 film was 1.76 eV, and it increased to about 1.82 eV upon annealing at 100 ∼ 300 ◦ C in vacuum. The dynamical behavior of the charged carriers in the AgInSe2 film was investigated by using the photoinduced discharge characteristics technique. PACS numbers: 68.55.-a, 78.20.-e Keywords: AgInSe2 , Annealing effect, Optical band gap, PIDC
I. INTRODUCTION The I-III-VI2 ternary compounds that crystallize in the chalcopyrite structure (space group I42d) form a large group of semiconducting materials with diverse physical and chemical properties [1–5]. The chalcopyrite structure can be deduced from that of the II-VI zincblende substrate by replacing the group II cations with those of groups I and III. Some of these compounds are of considerable interest for technological applications, such as optical detectors, light emitting diodes, solar cells, photovoltaic devices, and nonlinear optics. AgInSe2 belongs to this group of compounds and is a promising material in optical applications because it has a direct band gap, high optical absorption (∼10−5 cm−1 ), capability of pn-control, and relatively high mobility [1,2,6]. All of these characteristics are better than those of CuInSe2 [7, 8] for solar cell applications. However, there have been few reports concerning AgInSe2 . To synthesize a heterostructure device using AgInSe2 , one may employ CdS as a window material. Raviendra and Sharma  succeeded in forming n-CdS/p-AgInSe2 by electrodeposition and demonstrated solar cell performance with a 2 % conversion efficiency. Satyanarayana et al.  and Patel  reported variations in the electrical resistivity and the elemental composition of AgGaSe2 and AgInSe2 films grown at different substrate temperatures. El-Zahed et al.  reported the optical, electrical, and structural properties of AgSbSe2 films, and ∗ E-mail:
many experimental studies using Raman spectroscopy, X-ray absorption spectroscopy, luminescence, and reflectivity have been made on these compounds [7,13–15]. A variety of methods, e.g., spray pyrolysis , electrodeposition [9, 17], three- or four-source evaporation [18, 19], flash evaporation [20,21], RF magnetron sputtering , chalcogenization of metal layers , screen printing , and photo-chemical deposition , are used to form chalcopyrite films. However, it is very hard to grow stoichiometric single crystals, so studies of I-III-VI2 compound semiconductors are scarce. Although the first research on the AgInSe2 crystal was reported by Hahn et al.  in 1953, no significant improvement has been made since then, mainly due to the difficulties in growing single-crystal samples. Furthermore, as far as we know, studies on the optical properties of AgInSe2 single crystals have never been attempted. In this work, we report the dynamical behavior of carriers in AgInSe2 films studied by using the photo-induced discharge characteristics (PIDC) and the effect of thermal annealing on the structural and the optical properties of the films.
II. EXPERIMENTAL WORK The AgInSe2 films used in this study were prepared on cleaned indium-tin-oxide (ITO) glass by thermal vacuum evaporation using a AgInSe2 ingot as the source material. The elements for the ingot were stoichiometrically mixed and sealed in quartz tubes. The tubes were heated
Journal of the Korean Physical Society, Vol. 50, No. 4, April 2007
Fig. 1. X-ray diffraction patterns of AgInSe2 : (a) ingot material, films annealed at (b) 300 ◦ C, (c) 200 ◦ C, and (d) 100 ◦ C for 30 min in vacuum, and (e) as-deposited film.
up to 1,000 ◦ C at a rate of 50 ◦ C per hour, held at 1,000 ◦ C for 50 hours, and then cooled down to room temperature over 48 hours. The detailed experimental procedure for obtaining the AgInSe2 ingot is found elsewhere . AgInSe2 films were grown by thermal evaporation of the AgInSe2 ingot in a vacuum chamber. The distance between the substrate and the AgInSe2 ingot was about 15 cm. The over-all deposition time was about 30 min, and the substrate temperature was kept at about 200 ◦ C during the evaporation. After deposition, the samples were annealed at temperatures from 100 ◦ C to 300 ◦ C for 30 min in vacuum. The color of the samples gradually became dark silver with increasing annealing temperature. The crystal structure of the films was characterized by using X-ray diffraction (XRD) with Ni-filtered Cu Kα1 radiation. An ultraviolet-visible-near-infrared (UV-VisNIR) spectrophotometer was used to obtain the optical energy band gap at room temperature. X-ray photoelectron spectroscopy (XPS) was used to investigate the chemical information at the surface region of the AgInSe2 films. A VG Microtech ESCA2000 XPS spectrometer with a Mg-Kα source (FWHM: 0.826 eV) equipped with a CALM4 hemispherical electron energy analyzer (fixed pass energy of 23.4 eV) was used in the experiment. Atomic force microscopy (AFM) was used to investigate the surface morphology of the AgInSe2 films. We used the photoinduced discharge characteristics (PIDC) technique to extract the host carrier type, the transit time, the carrier concentration, and the mobility. The experimental configuration, which is commonly used in studying the transport of charge in a high-resistivity photoconductor, is found elsewhere [28–30]. Measurements
Fig. 2. XPS survey spectrum of (a) as-deposited and (b) 300 ◦ C annealed AgInSe2 films.
were carried out for as-deposited and for AgInSe2 films annealed at 100 ◦ C, 200 ◦ C and 300 ◦ C. III. RESULTS AND DISCUSSION The chemical compositions of the ingot and the films were confirmed by using the energy dispersive X-ray analysis (EDX). The compositions of the films and source materials showed no observable difference within the elemental detection limit of the EDX. The stoichiometric ratio of Ag : In : Se in the films was 25.2 : 26.7 : 48.1, which was almost the same as the stoichiometric composition (1 : 1 : 2) of the AgInSe2 films within the error limits of 2 %. The thickness of the AgInSe2 films grown on ITO glass was confirmed by using a crosssection scanning electron microscope, and the value was 1.5 µm. Figure 1 shows the XRD patterns for the ingot material and for the AgInSe2 thin films. Diffraction peaks for the (112), (220), (204), (312), and (116) planes of chalcopyrite AgInSe2 are observed in the XRD spectrum of ingot. No diffraction peaks were observed from the as-deposited film and from the film annealed at 100 ◦ C in vacuum. The diffraction peak for the (112) plane of chalcopyrite AgInSe2 appeared from the sample annealed at 200 ◦ C for 30 min in vacuum. The intensity of the (112) peak is enhanced, and another small peak, for the (204) plane, appeared upon further annealing at 300 ◦ C for 30 min. The lattice constants obtained from the XRD peak are a = 6.102 ˚ A and c = 11.69 ˚ A, which are comparable with those previously reported [22,23,31,32]. Figure 2 shows the XPS survey spectra of the (a) asdeposited and the (b) 300 ◦ C annealed AgInSe2 films.
Structural and Optical Properties of AgInSe2 Films Prepared· · · – Jeoung Ju Lee et al.
Fig. 4. AFM images of (a) as-deposited and (b) 300 ◦ C annealed AgInSe2 films. Fig. 3. Narrow scan XPS spectrum of the (a) Ag-3d, (b) In-3d, and (c) Se-3p core levels.
In the spectrum of the as-deposited film, In-3d, Ag-3d, Se-3d, and O-1s peaks are clearly observed. The survey spectrum of the 300 ◦ C annealed AgInSe2 film is similar to that of as-deposited film. The peak at 520 eV in both spectra is the O-1s peak, which probably originates from the absorbed gaseous oxygen molecules. The peak at a binding energy of about 380 eV is the Ag-3d peak, and the small peak at about 610 eV is the Ag-3p peak. The peaks at about 460 eV and 170 eV correspond to In-3d and Se-3p peaks, respectively. The small peak at 61.5 eV is the Se-3d peak. The intensity of the Ag-3d peak in the 300 ◦ C annealed AgInSe2 film is lower than that of the Ag-3d peak in as-deposited film. On the other hand, the In-3d peak in the 300 ◦ C annealed AgInSe2 film is higher than that of the as-deposited film. This implies that the as-deposited film is Ag-rich, but the 300 ◦ C annealed film is In-rich. Figure 3 shows the narrow range scan spectra of the (a) Ag-3d, (b) In-3d, and (c) Se-3p levels for the asdeposited (solid line) and the 300 ◦ C annealed (dotted line) AgInSe2 films. In all the spectra, the spin-orbit coupling effects are well developed. The total intensity of the Ag-3d peak of the as-deposited film is about 7 times higher than that of the 300 ◦ C annealed film. The
intensity of Se-3p peak of the as-deposited film is about 2.5 times higher than that of the 300 ◦ C annealed film. On the other hand, the total intensity of the In-3d peak of the as-deposited film is lower than that of the 300 ◦ C annealed film. The surface morphologies of the as-deposited and the annealed AgInSe2 films were characterized by using AFM. Figure 4 shows the surface morphologies of the (a) as-deposited and the (b) 300 ◦ C annealed AgInSe2 films. The area covered in all AFM pictures is 2 × 2 µm2 . The root mean square (RMS) roughness was measured by taking the average roughness of several scans measured under a length scale of 0.5 µm. As Figures 4 (a) and (b) show the RMS roughnesses of the as-deposited and the 300 ◦ C annealed AgInSe2 films were 2 nm and 0.6 nm, respectively. On the other hand, the mean grain size was 15.4 nm for the as-deposited AgInSe2 film and 2.3 nm for the 300 ◦ C annealed AgInSe2 film. From the AFM images, we can see that the RMS surface roughness and the grain size of the AgInSe2 films are decreased by thermal annealing in vacuum, which implies that nonuniform grains in the as-deposited films converted to uniform and dense AgInSe2 grains due to crystallization by thermal annealing at 300 ◦ C in vacuum. The optical absorption spectra of the as-deposited and the annealed AgInSe2 films were measured at room temperature. In the optical absorption spectra, an abrupt
Journal of the Korean Physical Society, Vol. 50, No. 4, April 2007
Fig. 5. hν vs. (αhν)2 plots for the as-deposited AgInSe2 film and for AgInSe2 films annealed at 100 ◦ C and 300 ◦ C for 30 min in vacuum.
increase is observed near 681 nm. This abrupt increase can be ascribed to optical absorption at the fundamental absorption edge of a semiconductor with a direct energy band structure [33, 34]. The optical absorption coefficient, α, is calculated from the optical absorption spectra near the fundamental absorption edge and is plotted in Figure 5 as a function of the incident photon energy, hν. As we can see in Figure 5, the relation between the optical absorption coefficient and the incident photon energy follows the formula for an allowed direct transition [33, 34], (αhν)2 ∼ (hν − Eg ),
where h is Planck’s constant and ν is the frequency of the incident photon. The optical energy gap (Eg ) can be obtained by extrapolating the straight-line portion of the plot to (αhν)2 = 0, as shown in Figure 5. A direct optical transition is confirmed by the straight line in this figure, and the optical band gaps of the as-deposited films and the annealed AgInSe2 film are about 1.76 eV and 1.82 eV, respectively. These values can be compared with those for other similar systems [22,30,31,35] but are almost the same as that of bulk AgGaSe2 . Figure 6 shows the oscilloscope trace for the PIDC of the (a) as-deposited and the (b) 300 ◦ C annealed (b) AgInSe2 films. Transit time characteristics are obtained from time-of-flight experiments. In such experiments, the sample, which is covered with two contacts, one of which is semi-transparent, is illuminated for a short time by an intense light pulse. This illumination causes a current of unipolar carriers to flow through the sample, across which a constant voltage is applied. The drifts of either electrons or holes could be observed by choosing
Fig. 6. Oscilloscope trace of the discharge of photoexcited carriers in (a) the as-deposited AgInSe2 film and (b) the 300 ◦ C annealed AgInSe2 film.
the polarity of the external power supply. In the ideal case, the injected carrier sheet reaches the back electrode after a definite transit time tT , from which a drift mobility can be derived. We estimated the transit time from the oscilloscope trace in Figure 6 by using the pulse technique method. The hole drift produced an upward (or positive) deflection, and the electron drift produced a downward (or negative) deflection on the oscilloscope screen, as shown in Figures 6 (a) and (b), respectively. The electron mobility and the carrier concentration were calculated by using the physical model and the theory of Batra et al . The estimated transit time was about 9 µs. The transit time (tT ) is expressed by tT = dD/µV,
where d is the sample thickness, D is the distance between the top and the bottom electrodes, µ is the drift mobility, and V is the applied voltage. The mobility of photo-induced carriers was evaluated by plotting the transit time obtained from the oscilloscope trace as a function of the reciprocal applied voltage and fitting it with a straight line. The drift electron mobility was calculated as µe ∼ 333 cm2 /Vs for the 300 ◦ C annealed sample. This value is very large compared with those for other compound semiconductors like Cd2 GeSe4 , and bulk AgInSe2 . The carrier concentration (ne ) and the transit time at the sample surface were ∼2.2 × 1016 /cm3 and 9 µs, respectively. This value agrees well with that reported by Ema and Harakawa . The drift
Structural and Optical Properties of AgInSe2 Films Prepared· · · – Jeoung Ju Lee et al.
hole mobility was calculated as µh ∼ 75 cm2 /Vs for the as-deposited film. The carrier concentration (nh ) and the transit time at the sample surface were ∼1.2 × 1017 /cm3 and 40 µs, respectively. In I-III-VI2 chalcopyrite semiconductors, the conduction type and the conductivity are greatly affected by the I/III atomic ratio or the metal/VI ratio: an I-rich sample shows p-type conductivity, and a III-rich sample, n-type, in general . By applying the same procedure as described above to all the samples, we found the as-deposited AgInSe2 film to be Ag-rich and the 300 ◦ C annealed AgInSe2 film to be In-rich. Thus, the PIDC result is consistent with the XPS result. IV. CONCLUSION AgInSe2 films were grown on ITO substrates by using conventional thermal vacuum evaporation. Post annealing at temperatures from 100 ◦ C to 300 ◦ C for 30 min in a vacuum enhanced the crystalline quality. Single-phase AgInSe2 with a (112) orientation was present in the 300 ◦ C annealed film. With XPS we found the as-deposited AgInSe2 film to be Ag-rich and the 300 ◦ C annealed AgInSe2 film to be In-rich. AFM images showed that the RMS roughness and the grain size decreased with increasing annealing temperature. The optical band gap energy was estimated to be 1.76 eV for the as-deposited film and 1.82 eV for the 300 ◦ C annealed film. The PIDC measurement showed that a hole drift was produced in the as-deposited AgInSe2 film and an electron drift was produced in the 300 ◦ C annealed AgInSe2 film. The electron mobility, the carrier concentration, and the transit time for the 300 ◦ C annealed AgInSe2 film were 333 cm2 /Vs, 2.2 × 1016 /cm3 , and 9 µs, respectively. The hole mobility, the hole concentration, and the transit time for the as-deposited film were 75 cm2 /Vs, 1.2 × 1017 /cm3 , and 45 µs, respectively. ACKNOWLEDGMENTS This work was supported in part by the Research Year Fund of Gyeongsang National University, 2004, and by the Korea Research Foundation (KRF) grant (KRF2003-005-C00014). REFERENCES  J. L. Shay and J. H. Wernick, Ternary Chalcopyrite Semiconductors: Growth, Electronic Properties and Applications, (Pergamon Press, Oxford, 1974).  J. E. Jaffe and A. Zunger, Phys. Rev. B 29, 1882 (1984).  S. M. Wasim, C. Rincon, G. Marin, P. Bocaranda, E. Hernandez, I. Bonalde and E. Medina, Phys. Rev. B 64, 195101 (2001).  D. Xue, K. Betzler and H. Hess, Phys. Rev. B 62, 13546 (2000).
 J. B. Caceres and C. Rincon, Phys. Stat. Sol. B 234, 541 (2002).  R. Noufi, R. Axton, C. Herrington and S. K. Deb, Appl. Phys. Lett. 7, 668 (1984).  B. Tell, J. L. Shay and H. M. Kasper, J. Appl. Phys. 43, 2469 (1972).  Y. Ema and N. Harakawa, Jpn. J. Appl. Phys. 34, 3260 (1995).  D. Raviendra and J. K. Sharma, Phys. Stat. Sol. (a) 88, 365 (1985).  Y. Satyanarayana, B. Srinivasula and P. Jayarama, Phys. Stat. Sol. (a) 115, K175 (1989).  S. M. Patel and A. D. Patel, Thin Solid Films 111, 53 (1984).  H. El-Zahed, Thin Solid Films 238, 104 (1994).  A. Jayaraman, V. Narayanamurti, H. M. Kasper, M. A. Chin and R. G. Maines, Phys. Rev. B 14, 3516 (1976).  A. K. Arora, T. Sakuntala and L. Artus, J. Phys. Chem. Solids 54, 381 (1993).  T. Tinoco, A. Polian, J. P. Itie, E. Moya and J. Bonzalez, J. Phys. Chem. Solids 56, 481 (1995).  T. Terasako, S. Shirakata and S. Isomura, Jpn. J. Appl. Phys. 38, 4656 (1999).  M. Boumerzoug and Le. H. Dao, J. Mater. Sci.: Material in Electronics 1, 123 (1990).  Y. L. Wu, H. Y. Lin, C. Y. Sun, M. H. Yang and H. L. Hwang, Thin Solid Films 168, 113 (1989).  W. E. Devaney, W. S. Chen, J. M. Stewart and R. A. Mickelsen, IEEE Trans. Electron Dev. 37, 428 (1990).  R. D. L. Kristensen, S. N. Sahu and D. Haneman, Appl. Surf. Sci. 33-34, 1285 (1988).  P. Paul Ramesh, O. Md. Hussain, S. Uthanna, B. Srinivasulu Naidu and P. Jayarama Reddy, Mater. Lett. 34, 217 (1998).  R. D. Weir, P. E. Jessop and B. K. Garside, Can. J. Phys. 65, 1033 (1987).  J. Szot and U. Prinz, J. Appl. Phys. 66, 6077 (1989).  S. A. AL Kuhaimi and S. Bahammam, Jpn. J. Appl. Phys. 29, 1499 (1990).  K. Moriya, K. Tanaka and H. Uchiki, Jpn. J. Appl. Phys. 44, 715 (2005).  H. Hahn, G. Frank, W. Klingler, A. D. Meyer and G. Strorger, Z. Anorg. Chem. 271, 153 (1953).  J. J. Lee, C. S. Yang, Y. S. Park, K. H. Kim, M. S. Jin, H. L. Park and W. T. Kim, J. Appl. Phys. 89, 3270 (2001).  J. L. Hartke, Phys. Rev. 125, 1177 (1962).  H. Scher and E. W. Montroll, Phys. Rev. B. 12, 2455 (1975).  Y. S. Park, K. H. Kim, J. J. Lee and T. W. Kang, J. Korean Phys. Soc. 44, 875 (2004).  P. Paul Ramesh, S. Uthanna, B. Srinivasulu Naidu and P. Jayarama Reddy, Vacuum 47, 211 (1996).  R. Poerschke, Data in Science and Technology (SpringerVerlag, Berlin, 1992).  J. I. Pankove, Optical Processes in Semiconductors (Dover, New York, 1971), Chaps. 1-3.  D. T. Kim, M. S. Jin and W. T. Kim, J. Korean Phys. Soc. 47, 331 (2005).  C. D. Kim, G. C. Park, M. S. Jin and D. T. Kim, J. Korean Phys. Soc. 48, 951 (2006).  I. P. Batra, K. Keiji Kanazawa and H. Seki, J. Appl. Phys. 41, 3416 (1970).