Electrical, optical and structural characteristics of ... - Semantic Scholar
Thin Solid Films 308–309 (1997) 1–7
Electrical, optical and structural characteristics of indium-tin-oxide thin films deposited on glass and polymer substrates A.K. Kulkarni a ,*, K.H. Schulz b, T.-S. Lim a, M. Khan b a
Department of Electrical Engineering, Michigan Technological University, 1400 Townsend Drive, Houghton, MI 49931, USA Department of Chemical Engineering, Michigan Technological University, 1400 Townsend Drive, Houghton, MI 49931, USA
b
Abstract The sheet resistance, optical transmittance and microstructure of tin-doped indium oxide (ITO) thin films (50–100-nm thick) rf sputter deposited on polymer substrates are investigated using a four-point probe, spectrophotometer, X-ray diffractometer and a transmission electron microscope (TEM). Sheet resistances vary from 250 Q/sq. to 170 kQ/sq. Sheet resistances for the ITO films on polycarbonate substrates are at least an order of magnitude higher than those ITO films deposited on glass substrates at the same time. Annealing ITO films on polycarbonate substrates at 100°C in air for 1 h decreased the sheet resistances significantly (almost by 50%). The X-ray diffraction data indicate polycrystalline films with grain orientations predominantly along (222) and (400) directions. TEM photographs show two distinct regions of growth: a dense growth close to the substrate and a sparse growth away from the substrate. The vertical growth is columnar and rod shaped. Changes in the ITO film sheet resistance either due to the types of substrate used or due to annealing can be correlated to the grain size and grain orientation. 1997 Elsevier Science S.A. Keywords: ITO; Sheet resistance; Transmittance; Grain orientation
1. Introduction Indium tin oxide, commonly referred to as ITO, is a degenerate n-type semiconducting material that has wide applications in optics and optoelectronics. These applications include flat panel display devices [1,2], heat reflecting mirrors [3], and heterojunction solar cells [4]. High electrical conductivity ( ≈ 2 × 104 Q−1 cm−1), and high transparency ( ≈ 90% in the visible spectrum range) of this material have been the focus of research throughout the world. Most of the research on ITO thin films is concentrated on the simultaneous improvement of the conductivity and the transparency of the ITO thin films deposited on glass substrates by a variety of techniques such as rf sputtering [4,5], electron beam deposition [6], chemical vapor deposition [7], and spray pyrolysis [8]. Attempts are also made to determine the fundamental properties of ITO based on the energy band diagram so that optimum theoretical values of conduc-
* Corresponding author. Tel.: +1 906 4872773; fax: +1 906 4872949; e-mail: [email protected]
0040-6090/97/$17.00 1997 Elsevier Science S.A. All rights reserved PII S0040-6090 (97 )0 0526-9
tivity and transparency can be obtained [9,10]. However, very little work is reported on the ITO thin films deposited on polymer substrates since these substrates have low thermal stability [11,12]. Modern applications employing ITO films require light polymer substrates such as plastics for use in liquid crystal display devices. In this paper, we report on the sheet resistance, optical transmittance and microstructure of the ITO thin films rf sputter deposited on polymer substrates.
2. Fabrication and processing Details of the rf sputtering technique and system, and sample preparation are given elsewhere [13]; only highlights of the fabrication and processing methods are given here. A water cooled pressed In2O3 target with 10 wt.% SnO2 is used inside a Perkin-Elmer model 2400 rf sputtering system. Three types of substrates were used: (i) glass, (ii) PET (poly ethylene terephthalate), and (iii) polycarbonate. These substrates were pre-cleaned with 2-propanol in an ultrasonic cleaner for approximately 6 min. During deposition, the oxygen partial pressure was varied from 0% to 20%
2
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Table 1 Fabrication, processing and sheet resistances of ITO thin films Sample
1C 2C 2D 3C 3D 4C 4D 5B 5C 5D a
Substrate
Polycarbonate Polycarbonate Glass Polycarbonate Glass Polycarbonate Glass PET Polycarbonate Glass
Deposition time (min) 30 30 30 30 30 20 20 30 30 30
RF power (W) 50 50 50 50 50 100 100 100 100 100
Oxygen partial pressure (%) 0 15 15 21 21 5 5 ? ? ?
Sheet resistancea (Q/sq.) Before annealing
After annealing
6.15 × 10 170 × 103 129 × 103 131 × 103 18.8 × 103 23.5 × 103 998 250 587 293
5.26 78.3 – 75.6 31.1 – 2.10 – – –
3
× 103 × 103 × 103 × 103 × 103
The error on the sheet resistance is about ±3%.
of the total pressure ( ≈ 2.67 Pa or 20 mTorr) to achieve the highest conductivity and transparency. Deposition parameters such as target bias and deposition time were held
constant during deposition. A few of the samples were annealed at 100°C in air for 1 h. Table 1 shows the details of sample fabrication and annealing.
Fig. 1. TEM pictures of sample 2C. (a) Electron diffraction patterns. (b) Before annealing (R S ≈ 170 kQ/sq.). (c) After annealing (R S ≈ 78.3 kQ/sq.).
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Fig. 2. TEM picture of annealed sample 4C. Inset shows columnar morphology.
3. Experimental results
3
with intensities similar to those expected from the JCPDS (Joint Committee on Powder Diffraction Standards) card on In2O3. No appreciable changes are observed in line intensities upon tilting the sample by 10°, indicating the absence of extreme preferred orientations. Fig. 1b shows the plan view of sample 2C (before annealing) showing clearly the columnar vertical growth with voids in between columns. The column length is approximately 60 nm and the columns appear to coalesce at the bottom where the film joins the substrate. The sheet resistance of this sample is 170 kQ/sq. After annealing in air at 100°C for 1 h, the sheet resistance drops to 78.3 kQ/sq.; the TEM picture of the annealed sample is shown in Fig. 1c. Here it is possible to observe parallel (222) and (112) fringes crossing multiple columns, suggesting grain growth beyond the vertical columns resulting in a multiple columnar growth. Vertical columnar growth is very apparent in the TEM picture of sample 4C shown in Fig. 2. As shown in the figure, the individual columns are single crystals with a fringe separation of 0.29 nm corresponding to the (222) planes. Fig. 3 shows the cross-sectional view of the ITO/ polymer interface. The ITO film is separated from the polymer during specimen preparation, but several polymer ligaments remain attached to the ITO film base (indicated by arrows) which would imply good adhesion. The ITO film appears to be quite dense at the base.
3.1. Sheet resistance measurements The sheet resistances of as-deposited and annealed samples were measured by a standard four-point probe technique and are listed in Table 1. Several interesting observations are made based on the oxygen partial pressure, the type of substrate and post deposition annealing. Sheet resistances increase with increasing oxygen partial pressure on both glass and polymer substrates. The increase in sheet resistances is significant in the ITO films deposited on polycarbonate substrates with 0–15% of oxygen partial pressure. The sheet resistances of the ITO films deposited on glass substrates are lower than those deposited on polycarbonate substrates. After annealing in air at 100°C for 1 h, the sheet resistances of the ITO films on glass substrates increased in both samples 3D and 4D whereas the sheet resistances of the ITO films deposited on polymer substrates decreased in all three samples 1C, 2C and 3C. This decrease is substantial (54% and 42%) for samples 2C and 3C, respectively, which are fabricated under high oxygen partial pressures (15% and 21%, respectively). 3.2. Transmission electron microscope (TEM) measurements The TEM used to characterize the structure of ITO thin films is a JEOL 4000 TEM operated at 200-kV accelerating voltage. Fig. 1a shows the electron diffraction ring patterns of sample 2C before annealing. As seen in Fig. 1a, the electron diffraction ring patterns give four strong lines
Fig. 3. TEM cross-sectional view of the ITO film/polymer interface. The length of the line at the top is 10 nm.
4
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Fig. 4. X-ray diffraction data on (a) sample 4C (before annealing, R S ≈ 23.5 kQ/sq.), and (b) sample 4D (before annealing, R S ≈ 1 kQ/sq.).
3.3. X-ray diffraction measurements The X-ray diffraction (XRD) patterns were obtained using an XDS 2000 Diffractometer. The X-ray source used is Cu Ka radiation (l = 0.154 nm) with a graphite monochromator (2a = 26.6°). Fig. 4a shows XRD data on sample 4C (as-deposited ITO on polymer) which has a sheet resistance of 23.5 kQ/sq. As seen in this figure, the peaks are small and broad indicating growth in (222), (400), (440) and (622) orientations. The normalized peak intensities for these peaks are 100, 100, 28 and 14, respectively and FWHM (full width at half maximum) values are 0.703°, 0.600°, 1.076° and 1.633°, respectively. In Fig. 4b, the X-ray data on sample 4D (as-deposited ITO on glass substrate) shows a sharp peak for (400) orientation and small broad peaks for (222), (440) and (622) orientations. The normalized intensities for (222), (400), (440) and (622) peaks are 30, 100, 15 and 15 and the FWHMs for the corresponding peaks are 0.749°, 0.592°, 1.093° and 1.593°, respectively. The sheet resistance of this sample is approximately 1 kQ/sq. In Fig. 5a,b,c, we compare the X-ray diffraction data and sheet resistances of three different ITO samples deposited on PET (5B), polycarbonate (5C) and glass (5D) substrates in the same run under exactly the same deposition conditions. The sheet resistances of samples 5B, 5C and 5D are 250 Q/sq., 587 Q/sq. and 293 Q/sq., respectively. Fig. 5a
shows a very sharp peak at (400) orientation and almost negligible peaks for other orientations. The peaks centered at 2v values equal to 47° and 54° are due to the PET substrate. Similarly, Fig. 5b shows a sharp peak only for (400) orientation and very small and broad peaks for (222), (440) and (622) orientations. On the other hand, the XRD data of an ITO sample deposited on glass substrate shown in Fig. 5c shows not only a sharp peak for (400) orientation but also for (320) orientation corresponding to the JCPDS orientations of ITO rather than In2O3. Comparing the X-ray diffraction plots of Fig. 4a and 5a (samples 4C and 5C), it is obvious that the decrease in sheet resistance from 23.5 kQ/sq. to 587 Q/sq. is quite significant and is due to a highly oriented film in sample 5C versus a poorly oriented film in sample 4C. The FWHM values of the (400) peaks in samples 4C and 5C are 0.600° and 0.145°, respectively. Similarly, the XRD plots of samples 4D and 5D show substantial differences with a corresponding decrease in the sheet resistance from 1 kQ/sq. to 293 Q/sq. The FWHM values of the (400) peaks in samples 4D and 5D are 0.592° and 0.114°, respectively. These results strongly suggest the importance of oriented growth (particularly along (400) direction) and large grain sizes (low FWHMs) in improving the conductivity of ITO thin films. Similar results are also observed by other investigators [14].
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
3.4. Optical transmission measurements The optical measurements of the ITO samples were carried out using a HP8451A Diode Array Spectrophotometer. Fig. 6 shows the transmission spectra of a bare polycarbonate substrate and samples 2C, 3C and 5C which are ITO films deposited on polycarbonate substrates. As seen in the figure, the transmittance is roughly 60%, 69% and 64% (with respect to the bare polycarbonate substrate) for sam-
5
ples 2C, 3C and 5C, respectively, in the wavelength range from 0.39 mm to 0.77 mm (visible spectrum).
4. Discussion 4.1. Sheet resistance The lowest sheet resistivities of the ITO thin films depos-
Fig. 5. X-ray diffraction data on (a) sample 5B (before annealing, R S ≈ 250 Q/sq.), (b) sample 5C (before annealing, R S ≈ 587 Q/sq.), and (c) sample 5D (before annealing, R S ≈ 293 Q/sq.).
6
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
sheet resistances of the ITO films deposited on glass substrates (samples 3D and 4D) increased significantly after annealing in air at 100°C for 1 h. This is an expected result since the annealing of these ITO films decreased the number of vacancies resulting in more resistive films. 4.2. Microstructure
Fig. 6. Optical transmission spectra on a bare polycarbonate substrate and samples 2C, 3C and 5C.
ited on heated glass substrates reported in the literature are of the order 2 × 10−4 Q cm. Previous researchers working on substrates other than glass such as mylar, polyester, teflon and other polymer substrates have reported resistivity values in the range of 1 × 10−3 Q cm, which is an order of magnitude higher than the resistivity of the ITO films deposited on glass [11,12]. The resistivity values reported here on unheated polymer substrates are comparable to the values reported by other researchers [11,12]. The dependence of the resistivity on the oxygen partial pressure is a well known experimental result and is explained on the basis of oxygen deficiency in the film (each oxygen vacancy gives rise to two conduction electrons) [7]. Increasing oxygen content of the films either by increasing the partial pressure of oxygen during the growth or annealing the samples in air or oxygen should decrease the oxygen vacancies leading to less conductive films. However, a minimum in the resistivity of the ITO thin films deposited on glass substrates is reported between 15 and 20% of oxygen partial pressure by a few investigators indicating an improvement in the crystallinity of the films (mobility of the carriers is dependent on crystallinity) [15]. Our experimental results on ITO deposited on polycarbonate substrates did not show a maximum in the conductivity as a function of oxygen partial pressure. An interesting observation can be made on the effect of annealing the ITO films deposited on glass and polymer substrates. As shown in Table 1, the sheet resistance of these ITO films decreased by 50% in samples 2C and 3C after annealing in air at 100°C for 1 h. This improvement in conductivity is attributed to an increase in the grain size of the film as observed in these TEM pictures (Fig. 1b, c) of sample 2C. As seen in these figures, the voids are significantly less in the annealed sample. The X-ray diffraction data (not shown) on annealed sample 3C showed a peak at 2v ≈ 30.58° indicating (222) preferred orientation for this film. On the other hand, the
The grain orientations, grain size and grain boundaries determine the quality of the thin films suggesting a dependence of the electrical and optical properties on the growth parameters [16]. The TEM results shown in Figs. 1,2 and 3 clearly show the columnar growth of the ITO films on polycarbonate substrates. However, the vertical columns are separated by voids at certain places and it appears that the voids may be responsible for the high sheet resistances of unannealed samples. The decrease in voids as well as increase in grain size from a few nanometers to a few tens of nanometers after annealing in air at 100°C for 1 h is speculated to be the cause for the decrease in the sheet resistance. The two distinct regions of growth observed in the TEM pictures (dense growth close to the substrate and sparse growth away from the substrate) have also been reported [17]. The multiple orientations observed in TEM pictures are substantiated by XRD results. The X-ray diffraction results shown in Figs. 4 and 5 indicate mainly four different orientations (222), (400), (440) and (622) which are normally observed for ITO films [18]. Our XRD results strongly suggest that highly oriented films have lowest sheet resistances. The lower FWHM values are a good indication of the crystallinity of the films resulting in lower sheet resistances.
5. Conclusions The ITO thin films are deposited on glass, PET and polycarbonate substrates by rf sputtering technique. The sheet resistances vary from 250 Q/sq. to 170 kQ/sq. depending upon oxygen partial pressure during the growth and the types of substrate used. Annealing in air at 100°C for 1 h resulted in either an increase or decrease in sheet resistance values depending on the types of substrate. Optical transmittance is 60–70% in the visible region for the samples studied here. The TEM results indicate vertical columnar growth with multiple orientations. The grain sizes range from a few nanometers to a few tens of nanometers and control the sheet resistances of the films. A significant decrease in the sheet resistances of the ITO films on polycarbonate substrates after annealing is attributed to a corresponding change in the grain size and the absence of voids. Our results suggest that the low resistive ITO thin films are obtained in the films that are highly oriented in the (400) direction.
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Acknowledgements The authors gratefully acknowledge the support of the 3M corporation, GE plastics, Steve Hackney for assistance with TEM work and Edward Laitila for assistance with XRD work.
References [1] J.E. Costellamo, Handbook of Display Technology, Academic Press, New York, 1992. [2] S. Ishibashi, Y. Higuchi, Y. Ota and K. Nakamura, J. Vac. Sci. Technol., A8 (1990) 1399. [3] K.L. Chopra and S.R. Das, Thin Film Solar Cells, Plenum Press, New York, 1983, p. 321. [4] C.V.R. Vasant Kumar and A. Mansingh, J. Appl. Phys., 65 (1989) 1270. [5] S.A. Knickerbocker and A.K. Kulkarni, J. Vac. Sci. Technol., A13 (3) (1995) 1048. [6] S.A. Agnihotry, K.K. Sari, T.K. Saxena, K.C. Nagpal and S. Chandra, J. Phys. D: Appl. Phys., 18 (1985) 2087.
7
[7] K.L. Chopra, S. Major and D.K. Pandya, Thin Solid Films, 102 (1983) 1. [8] J.C. Manifacier, L. Szepessy, J.F. Bresse, M. Perotin and R. Stuck, Mater. Res. Bull., 14 (1979) 163. [9] A.K. Kulkarni and S.A. Knickerbocker, J. Vac. Sci. Technol., A14 (1996) 1709. [10] S.A. Knickerbocker and A.K. Kulkarni, J. Vac. Sci. Technol., A14 (1996) 757. [11] A. Mansingh and C.V.R. Vasant Kumar, Thin Solid Films, 167 (1988) L11. [12] B. Chiou, S. Hsieh and W. Wu, J. Am. Ceram. Soc., 77 (1994) 1740. [13] S.A. Knickerbocker, Ph.D. dissertation, Michigan Technological University, 1995. [14] Y. Shigesato and D.C. Paine, Thin Solid Films, 238 (1994) 44. [15] K. Sreenivas, T. Sudersena Rao and A. Mansingh, J. Appl. Phys., 57 (2) (1985) 384. [16] M. Ohring, The Materials Science of Thin Films, Academic Press, New York, 1992, p. 451. [17] M. Kamei, Y. Shigesato and S. Takaki, Thin Solid Films, 259 (1995) 38. [18] T.J. Vink, W. Walrave, J.L.C. Daams, P.C. Baarslag and J.E.A.M. Meerakker, Thin Solid Films, 266 (1995) 145.
Electrical, optical and structural characteristics of indium-tin-oxide thin films deposited on glass and polymer substrates A.K. Kulkarni a ,*, K.H. Schulz b, T.-S. Lim a, M. Khan b a
Department of Electrical Engineering, Michigan Technological University, 1400 Townsend Drive, Houghton, MI 49931, USA Department of Chemical Engineering, Michigan Technological University, 1400 Townsend Drive, Houghton, MI 49931, USA
b
Abstract The sheet resistance, optical transmittance and microstructure of tin-doped indium oxide (ITO) thin films (50–100-nm thick) rf sputter deposited on polymer substrates are investigated using a four-point probe, spectrophotometer, X-ray diffractometer and a transmission electron microscope (TEM). Sheet resistances vary from 250 Q/sq. to 170 kQ/sq. Sheet resistances for the ITO films on polycarbonate substrates are at least an order of magnitude higher than those ITO films deposited on glass substrates at the same time. Annealing ITO films on polycarbonate substrates at 100°C in air for 1 h decreased the sheet resistances significantly (almost by 50%). The X-ray diffraction data indicate polycrystalline films with grain orientations predominantly along (222) and (400) directions. TEM photographs show two distinct regions of growth: a dense growth close to the substrate and a sparse growth away from the substrate. The vertical growth is columnar and rod shaped. Changes in the ITO film sheet resistance either due to the types of substrate used or due to annealing can be correlated to the grain size and grain orientation. 1997 Elsevier Science S.A. Keywords: ITO; Sheet resistance; Transmittance; Grain orientation
1. Introduction Indium tin oxide, commonly referred to as ITO, is a degenerate n-type semiconducting material that has wide applications in optics and optoelectronics. These applications include flat panel display devices [1,2], heat reflecting mirrors [3], and heterojunction solar cells [4]. High electrical conductivity ( ≈ 2 × 104 Q−1 cm−1), and high transparency ( ≈ 90% in the visible spectrum range) of this material have been the focus of research throughout the world. Most of the research on ITO thin films is concentrated on the simultaneous improvement of the conductivity and the transparency of the ITO thin films deposited on glass substrates by a variety of techniques such as rf sputtering [4,5], electron beam deposition [6], chemical vapor deposition [7], and spray pyrolysis [8]. Attempts are also made to determine the fundamental properties of ITO based on the energy band diagram so that optimum theoretical values of conduc-
* Corresponding author. Tel.: +1 906 4872773; fax: +1 906 4872949; e-mail: [email protected]
0040-6090/97/$17.00 1997 Elsevier Science S.A. All rights reserved PII S0040-6090 (97 )0 0526-9
tivity and transparency can be obtained [9,10]. However, very little work is reported on the ITO thin films deposited on polymer substrates since these substrates have low thermal stability [11,12]. Modern applications employing ITO films require light polymer substrates such as plastics for use in liquid crystal display devices. In this paper, we report on the sheet resistance, optical transmittance and microstructure of the ITO thin films rf sputter deposited on polymer substrates.
2. Fabrication and processing Details of the rf sputtering technique and system, and sample preparation are given elsewhere [13]; only highlights of the fabrication and processing methods are given here. A water cooled pressed In2O3 target with 10 wt.% SnO2 is used inside a Perkin-Elmer model 2400 rf sputtering system. Three types of substrates were used: (i) glass, (ii) PET (poly ethylene terephthalate), and (iii) polycarbonate. These substrates were pre-cleaned with 2-propanol in an ultrasonic cleaner for approximately 6 min. During deposition, the oxygen partial pressure was varied from 0% to 20%
2
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Table 1 Fabrication, processing and sheet resistances of ITO thin films Sample
1C 2C 2D 3C 3D 4C 4D 5B 5C 5D a
Substrate
Polycarbonate Polycarbonate Glass Polycarbonate Glass Polycarbonate Glass PET Polycarbonate Glass
Deposition time (min) 30 30 30 30 30 20 20 30 30 30
RF power (W) 50 50 50 50 50 100 100 100 100 100
Oxygen partial pressure (%) 0 15 15 21 21 5 5 ? ? ?
Sheet resistancea (Q/sq.) Before annealing
After annealing
6.15 × 10 170 × 103 129 × 103 131 × 103 18.8 × 103 23.5 × 103 998 250 587 293
5.26 78.3 – 75.6 31.1 – 2.10 – – –
3
× 103 × 103 × 103 × 103 × 103
The error on the sheet resistance is about ±3%.
of the total pressure ( ≈ 2.67 Pa or 20 mTorr) to achieve the highest conductivity and transparency. Deposition parameters such as target bias and deposition time were held
constant during deposition. A few of the samples were annealed at 100°C in air for 1 h. Table 1 shows the details of sample fabrication and annealing.
Fig. 1. TEM pictures of sample 2C. (a) Electron diffraction patterns. (b) Before annealing (R S ≈ 170 kQ/sq.). (c) After annealing (R S ≈ 78.3 kQ/sq.).
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Fig. 2. TEM picture of annealed sample 4C. Inset shows columnar morphology.
3. Experimental results
3
with intensities similar to those expected from the JCPDS (Joint Committee on Powder Diffraction Standards) card on In2O3. No appreciable changes are observed in line intensities upon tilting the sample by 10°, indicating the absence of extreme preferred orientations. Fig. 1b shows the plan view of sample 2C (before annealing) showing clearly the columnar vertical growth with voids in between columns. The column length is approximately 60 nm and the columns appear to coalesce at the bottom where the film joins the substrate. The sheet resistance of this sample is 170 kQ/sq. After annealing in air at 100°C for 1 h, the sheet resistance drops to 78.3 kQ/sq.; the TEM picture of the annealed sample is shown in Fig. 1c. Here it is possible to observe parallel (222) and (112) fringes crossing multiple columns, suggesting grain growth beyond the vertical columns resulting in a multiple columnar growth. Vertical columnar growth is very apparent in the TEM picture of sample 4C shown in Fig. 2. As shown in the figure, the individual columns are single crystals with a fringe separation of 0.29 nm corresponding to the (222) planes. Fig. 3 shows the cross-sectional view of the ITO/ polymer interface. The ITO film is separated from the polymer during specimen preparation, but several polymer ligaments remain attached to the ITO film base (indicated by arrows) which would imply good adhesion. The ITO film appears to be quite dense at the base.
3.1. Sheet resistance measurements The sheet resistances of as-deposited and annealed samples were measured by a standard four-point probe technique and are listed in Table 1. Several interesting observations are made based on the oxygen partial pressure, the type of substrate and post deposition annealing. Sheet resistances increase with increasing oxygen partial pressure on both glass and polymer substrates. The increase in sheet resistances is significant in the ITO films deposited on polycarbonate substrates with 0–15% of oxygen partial pressure. The sheet resistances of the ITO films deposited on glass substrates are lower than those deposited on polycarbonate substrates. After annealing in air at 100°C for 1 h, the sheet resistances of the ITO films on glass substrates increased in both samples 3D and 4D whereas the sheet resistances of the ITO films deposited on polymer substrates decreased in all three samples 1C, 2C and 3C. This decrease is substantial (54% and 42%) for samples 2C and 3C, respectively, which are fabricated under high oxygen partial pressures (15% and 21%, respectively). 3.2. Transmission electron microscope (TEM) measurements The TEM used to characterize the structure of ITO thin films is a JEOL 4000 TEM operated at 200-kV accelerating voltage. Fig. 1a shows the electron diffraction ring patterns of sample 2C before annealing. As seen in Fig. 1a, the electron diffraction ring patterns give four strong lines
Fig. 3. TEM cross-sectional view of the ITO film/polymer interface. The length of the line at the top is 10 nm.
4
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Fig. 4. X-ray diffraction data on (a) sample 4C (before annealing, R S ≈ 23.5 kQ/sq.), and (b) sample 4D (before annealing, R S ≈ 1 kQ/sq.).
3.3. X-ray diffraction measurements The X-ray diffraction (XRD) patterns were obtained using an XDS 2000 Diffractometer. The X-ray source used is Cu Ka radiation (l = 0.154 nm) with a graphite monochromator (2a = 26.6°). Fig. 4a shows XRD data on sample 4C (as-deposited ITO on polymer) which has a sheet resistance of 23.5 kQ/sq. As seen in this figure, the peaks are small and broad indicating growth in (222), (400), (440) and (622) orientations. The normalized peak intensities for these peaks are 100, 100, 28 and 14, respectively and FWHM (full width at half maximum) values are 0.703°, 0.600°, 1.076° and 1.633°, respectively. In Fig. 4b, the X-ray data on sample 4D (as-deposited ITO on glass substrate) shows a sharp peak for (400) orientation and small broad peaks for (222), (440) and (622) orientations. The normalized intensities for (222), (400), (440) and (622) peaks are 30, 100, 15 and 15 and the FWHMs for the corresponding peaks are 0.749°, 0.592°, 1.093° and 1.593°, respectively. The sheet resistance of this sample is approximately 1 kQ/sq. In Fig. 5a,b,c, we compare the X-ray diffraction data and sheet resistances of three different ITO samples deposited on PET (5B), polycarbonate (5C) and glass (5D) substrates in the same run under exactly the same deposition conditions. The sheet resistances of samples 5B, 5C and 5D are 250 Q/sq., 587 Q/sq. and 293 Q/sq., respectively. Fig. 5a
shows a very sharp peak at (400) orientation and almost negligible peaks for other orientations. The peaks centered at 2v values equal to 47° and 54° are due to the PET substrate. Similarly, Fig. 5b shows a sharp peak only for (400) orientation and very small and broad peaks for (222), (440) and (622) orientations. On the other hand, the XRD data of an ITO sample deposited on glass substrate shown in Fig. 5c shows not only a sharp peak for (400) orientation but also for (320) orientation corresponding to the JCPDS orientations of ITO rather than In2O3. Comparing the X-ray diffraction plots of Fig. 4a and 5a (samples 4C and 5C), it is obvious that the decrease in sheet resistance from 23.5 kQ/sq. to 587 Q/sq. is quite significant and is due to a highly oriented film in sample 5C versus a poorly oriented film in sample 4C. The FWHM values of the (400) peaks in samples 4C and 5C are 0.600° and 0.145°, respectively. Similarly, the XRD plots of samples 4D and 5D show substantial differences with a corresponding decrease in the sheet resistance from 1 kQ/sq. to 293 Q/sq. The FWHM values of the (400) peaks in samples 4D and 5D are 0.592° and 0.114°, respectively. These results strongly suggest the importance of oriented growth (particularly along (400) direction) and large grain sizes (low FWHMs) in improving the conductivity of ITO thin films. Similar results are also observed by other investigators [14].
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
3.4. Optical transmission measurements The optical measurements of the ITO samples were carried out using a HP8451A Diode Array Spectrophotometer. Fig. 6 shows the transmission spectra of a bare polycarbonate substrate and samples 2C, 3C and 5C which are ITO films deposited on polycarbonate substrates. As seen in the figure, the transmittance is roughly 60%, 69% and 64% (with respect to the bare polycarbonate substrate) for sam-
5
ples 2C, 3C and 5C, respectively, in the wavelength range from 0.39 mm to 0.77 mm (visible spectrum).
4. Discussion 4.1. Sheet resistance The lowest sheet resistivities of the ITO thin films depos-
Fig. 5. X-ray diffraction data on (a) sample 5B (before annealing, R S ≈ 250 Q/sq.), (b) sample 5C (before annealing, R S ≈ 587 Q/sq.), and (c) sample 5D (before annealing, R S ≈ 293 Q/sq.).
6
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
sheet resistances of the ITO films deposited on glass substrates (samples 3D and 4D) increased significantly after annealing in air at 100°C for 1 h. This is an expected result since the annealing of these ITO films decreased the number of vacancies resulting in more resistive films. 4.2. Microstructure
Fig. 6. Optical transmission spectra on a bare polycarbonate substrate and samples 2C, 3C and 5C.
ited on heated glass substrates reported in the literature are of the order 2 × 10−4 Q cm. Previous researchers working on substrates other than glass such as mylar, polyester, teflon and other polymer substrates have reported resistivity values in the range of 1 × 10−3 Q cm, which is an order of magnitude higher than the resistivity of the ITO films deposited on glass [11,12]. The resistivity values reported here on unheated polymer substrates are comparable to the values reported by other researchers [11,12]. The dependence of the resistivity on the oxygen partial pressure is a well known experimental result and is explained on the basis of oxygen deficiency in the film (each oxygen vacancy gives rise to two conduction electrons) [7]. Increasing oxygen content of the films either by increasing the partial pressure of oxygen during the growth or annealing the samples in air or oxygen should decrease the oxygen vacancies leading to less conductive films. However, a minimum in the resistivity of the ITO thin films deposited on glass substrates is reported between 15 and 20% of oxygen partial pressure by a few investigators indicating an improvement in the crystallinity of the films (mobility of the carriers is dependent on crystallinity) [15]. Our experimental results on ITO deposited on polycarbonate substrates did not show a maximum in the conductivity as a function of oxygen partial pressure. An interesting observation can be made on the effect of annealing the ITO films deposited on glass and polymer substrates. As shown in Table 1, the sheet resistance of these ITO films decreased by 50% in samples 2C and 3C after annealing in air at 100°C for 1 h. This improvement in conductivity is attributed to an increase in the grain size of the film as observed in these TEM pictures (Fig. 1b, c) of sample 2C. As seen in these figures, the voids are significantly less in the annealed sample. The X-ray diffraction data (not shown) on annealed sample 3C showed a peak at 2v ≈ 30.58° indicating (222) preferred orientation for this film. On the other hand, the
The grain orientations, grain size and grain boundaries determine the quality of the thin films suggesting a dependence of the electrical and optical properties on the growth parameters [16]. The TEM results shown in Figs. 1,2 and 3 clearly show the columnar growth of the ITO films on polycarbonate substrates. However, the vertical columns are separated by voids at certain places and it appears that the voids may be responsible for the high sheet resistances of unannealed samples. The decrease in voids as well as increase in grain size from a few nanometers to a few tens of nanometers after annealing in air at 100°C for 1 h is speculated to be the cause for the decrease in the sheet resistance. The two distinct regions of growth observed in the TEM pictures (dense growth close to the substrate and sparse growth away from the substrate) have also been reported [17]. The multiple orientations observed in TEM pictures are substantiated by XRD results. The X-ray diffraction results shown in Figs. 4 and 5 indicate mainly four different orientations (222), (400), (440) and (622) which are normally observed for ITO films [18]. Our XRD results strongly suggest that highly oriented films have lowest sheet resistances. The lower FWHM values are a good indication of the crystallinity of the films resulting in lower sheet resistances.
5. Conclusions The ITO thin films are deposited on glass, PET and polycarbonate substrates by rf sputtering technique. The sheet resistances vary from 250 Q/sq. to 170 kQ/sq. depending upon oxygen partial pressure during the growth and the types of substrate used. Annealing in air at 100°C for 1 h resulted in either an increase or decrease in sheet resistance values depending on the types of substrate. Optical transmittance is 60–70% in the visible region for the samples studied here. The TEM results indicate vertical columnar growth with multiple orientations. The grain sizes range from a few nanometers to a few tens of nanometers and control the sheet resistances of the films. A significant decrease in the sheet resistances of the ITO films on polycarbonate substrates after annealing is attributed to a corresponding change in the grain size and the absence of voids. Our results suggest that the low resistive ITO thin films are obtained in the films that are highly oriented in the (400) direction.
A.K. Kulkarni et al. / Thin Solid Films 308–309 (1997) 1–7
Acknowledgements The authors gratefully acknowledge the support of the 3M corporation, GE plastics, Steve Hackney for assistance with TEM work and Edward Laitila for assistance with XRD work.
References [1] J.E. Costellamo, Handbook of Display Technology, Academic Press, New York, 1992. [2] S. Ishibashi, Y. Higuchi, Y. Ota and K. Nakamura, J. Vac. Sci. Technol., A8 (1990) 1399. [3] K.L. Chopra and S.R. Das, Thin Film Solar Cells, Plenum Press, New York, 1983, p. 321. [4] C.V.R. Vasant Kumar and A. Mansingh, J. Appl. Phys., 65 (1989) 1270. [5] S.A. Knickerbocker and A.K. Kulkarni, J. Vac. Sci. Technol., A13 (3) (1995) 1048. [6] S.A. Agnihotry, K.K. Sari, T.K. Saxena, K.C. Nagpal and S. Chandra, J. Phys. D: Appl. Phys., 18 (1985) 2087.
7
[7] K.L. Chopra, S. Major and D.K. Pandya, Thin Solid Films, 102 (1983) 1. [8] J.C. Manifacier, L. Szepessy, J.F. Bresse, M. Perotin and R. Stuck, Mater. Res. Bull., 14 (1979) 163. [9] A.K. Kulkarni and S.A. Knickerbocker, J. Vac. Sci. Technol., A14 (1996) 1709. [10] S.A. Knickerbocker and A.K. Kulkarni, J. Vac. Sci. Technol., A14 (1996) 757. [11] A. Mansingh and C.V.R. Vasant Kumar, Thin Solid Films, 167 (1988) L11. [12] B. Chiou, S. Hsieh and W. Wu, J. Am. Ceram. Soc., 77 (1994) 1740. [13] S.A. Knickerbocker, Ph.D. dissertation, Michigan Technological University, 1995. [14] Y. Shigesato and D.C. Paine, Thin Solid Films, 238 (1994) 44. [15] K. Sreenivas, T. Sudersena Rao and A. Mansingh, J. Appl. Phys., 57 (2) (1985) 384. [16] M. Ohring, The Materials Science of Thin Films, Academic Press, New York, 1992, p. 451. [17] M. Kamei, Y. Shigesato and S. Takaki, Thin Solid Films, 259 (1995) 38. [18] T.J. Vink, W. Walrave, J.L.C. Daams, P.C. Baarslag and J.E.A.M. Meerakker, Thin Solid Films, 266 (1995) 145.
