Issue
Eur. Phys. J. Appl. Phys.
Volume 89, Number 2, February 2020
Disordered Semiconductors: Physics and Applications
Article Number 20301
Number of page(s) 5
Section Thin Films
DOI https://doi.org/10.1051/epjap/2020190240
Published online 09 April 2020

© EDP Sciences, 2020

1 Introduction

Indium compounds are used for various electronic applications such as indium tin oxide as transparent electrodes and indium gallium zinc oxide for thin-film transistors of flat panel displays [1,2]. The role of indium is considered to be extremely important for excellent electronic conduction properties. Understanding the electronic structure of indium oxides helps improving the performance of the devices. As indium oxides, indium sulfides (In-S) are important for optoelectronics and photovoltaic applications because of their semiconducting properties. They have the appropriate band gap and n-type conductivity to be used as buffer layers for solar cell devices [35]. In-S can be one of the materials constituting the next generation of cost-effective and environment-friendly solar cell. The properties of In-S nanowires and columnar structures are also of interest [6]. Several synthesis methods such as chemical synthesis, ionic layer adsorption and reaction method, and co-evaporation method have been studied to obtain high-quality In-S [612]. Despite the increasing importance of materials, the relation between defect absorption and carrier generation in In-S remains unclear. Therefore, understanding the defect absorption and carrier generation in In-S is valuable. In this study, the optical and electrical properties of In-S films with different heat treatments were investigated. As a starting material, In-S powder was prepared by the sulfur gas heat treatment of indium. The structure of the In-S powder was evaluated using X-ray diffraction (XRD). The In-S films were deposited by vacuum evaporation of the In-S powder. In-S films were heat treated in Ar gas in the temperature range of 100–400 °C. In addition, some annealed samples were heat treated at 300 °C with sulfur powder. The band gap and subgap absorption of the In-S films were investigated by photothermal deflection spectroscopy and optical transmittance. The carrier properties were evaluated by electrical conductivity and thermopower measurements.

2 Experimental details

An indium plate (purity 99.9%) and sulfur powder (purity 99%) were used as the sources of In and S, respectively, and placed in a test tube with Ar gas. The indium plate and sulfur powder were heated at 400 °C for 10 min. The product was in the form of powder and pieces and the color was red. Small pieces of the product were crushed before analysis. To determine the crystal phase of the In-S powder, XRD (RINT TTR-II, Rigaku) measurement was carried out. The In-S films were deposited on a quartz and Si substrate by vacuum evaporation of the In-S powder. The thickness of the thin film evaluated by an atomic force microscope (SPA-300, Seiko Instruments) was 140 nm. The films were annealed at different temperatures of 100–400 °C for 10 min in Ar gas. Heat treatment in Ar gas is referred to as Ar heat treatment. Some of the annealed samples at 300 °C were annealed again at 300 °C in Ar gas with 5 mg of sulfur powder. The heat treatment with sulfur powder is referred to as S heat treatment. The optical absorption spectra of the In-S films were evaluated by optical transmittance and photothermal deflection spectroscopy (PDS). PDS is suitable for the evaluation of optical absorption in submicron-thick films [13,14]. The PDS system was composed of an excitation light source and optical detection. Monochromatic light was obtained by passing light from a 50 W tungsten-halogen lamp through a grating monochromator (CM110, Spectral Products). The photothermal signal was measured using an optical system consisting of He-Ne laser light with a wavelength of 632.8 nm, a lock-in amplifier (SR830, Stanford Research), an optical chopper, and a position sensor (S3979, Hamamatsu Photonics). To detect a sufficiently strong photothermal signal, the sample was placed in a liquid as a deflection medium. Fluorine-based liquid FC-43 was used as the deflection medium. This liquid has the advantages, of being transparent in a wide wavelength range, stable, and difficult to react with substances. The transmittance spectra were measured using a spectrometer (UV-1800, Shimadzu). Structural and spectroscopic measurements were performed at room temperature.

3 Results and discussion

XRD was used to clarify the red powder obtained by thermal sulfidation. The XRD patterns of the red powder and representative In-S are shown in Figure 1. Although the signal intensity of the diffraction was small, several diffraction peaks were confirmed. The diffraction peaks related to cubic In2S3 (JCPDS No. 25-0456), tetragonal In2S3 (JCPDS No. 32-0390) [8] and In6S7 (JCPDS No. 19-0587) [15] were observed. The XRD result showed that the main component of the red powder was a mixture of In2S3 and In6S7.

Raman spectroscopy provides important information about substances. Figure 2 shows the Raman spectra of the In-S films. Raman peaks were observed at 240, 307, and 520 cm−1. According to [16], the Raman peaks at 240 and 307 cm−1 were attributed to the vibration of InS6 and InS4, respectively. The peak at 520 cm−1 was derived from the Si of the substrate. No significant change in the Raman spectra due to heat treatment was observed. From the results of the Raman spectroscopic analysis, thin-film samples were identified as an In-S compound.

The optical properties of the In-S films are important in understanding the electronic structure. Information on the localized states helps to understand the carrier generation and transport phenomena. By contrast, the weak absorption spectra of submicron-thick film samples using optical transmittance are difficult to measure. To overcome this difficulty, PDS was used in this study. The concept of PDS [13,14] is the sensitive detection of photothermal signals generated by the optical absorption of samples. The deflection of the probe laser through a transparent liquid is the signature of the photothermal signal.

Figure 3 shows the optical absorption spectra of the In-S films with different heat treatments. The absorption coefficient below 2 × 104 cm−1 was determined by PDS, and that above 2 × 104 cm−1 was determined by optical transmittance. The absorption due to band-to-band transitions was observed in the region of energies larger than 2 eV, and the weak absorption related to the localized states was observed in the region of energies smaller than 2 eV. The optical band gap Eg was estimated by using Tauc's relation, (αhν) n  = B( − Eg ). Here, n is the integer that depends on optical transition, and it has the value of n = 2 and 1/2 for direct and indirect transition, respectively. The direct and indirect bandgaps of the as-deposited In-S films were estimated to be 2.7 and 1.9 eV, respectively. The energy of the indirect gap almost agrees with the band gap Eg  = 2.01 eV of In2S3 found in a previous work [11]. The energy of the direct gap also agrees with the band gap Eg  = 2.81 eV [9] and 2.84 [12]. In-S thin films prepared by vacuum deposition have optical properties similar to those of In2S3 thin films. Several changes by heat treatment were confirmed in the optical absorption spectra. The thermal history of the sample was classified into an as-deposited state, a heat treatment state of 100–400 °C (Ar heat treatment), and a heat treatment state of 300 °C with sulfur (S heat treatment). The spectra around the band edge region were almost the same except in the S heat treatment sample. The absorption of the S heat treatment sample in the energy region above 1.8 eV was shifted to a higher energy than that of as-deposited sample by 0.3 eV. The blue shift of the band gap due to heat treatment has been reported in In2S3, and the localized states at the band edge are involved [12].

Optical absorption in the energy region lower than 1.8 eV showed a large change by heat treatment. In the region of energies lower than the band gap, tail absorption related to the localized states was observed in many semiconductors, but no reports of subgap absorption in the In-S films could be found. Therefore, the absorption coefficient at 1 eV in the In-S films was examined. The absorption coefficient of the as-deposited sample was the smallest, and the absorption coefficient increased with the heat treatment temperature. The value after heat treatment at 300 °C was 20 times the initial value. After S heat treatment, absorption decreased significantly, indicating the possibility of recovering to the initial state.

Changes in the subgap absorption were also discussed. The origin of the subgap absorption was unclear, but it was inferred from previous results. In2S3 has been reported to exhibit n-type conductivity, as sulfur vacancies act as donors and produce mobile electrons [17]. Sulfur vacancies are assumed to exist in In-S [5]. The optical absorption in the low energy region below 1.8 eV may be linked to the absorption of localized levels caused by sulfur vacancies.

Comparing the changes in properties due to heat treatment is important. Figure 4 shows the heat treatment temperature dependences of the (a) indirect band gap Eg , (b) electrical resistivity ρ, (c) Seebeck coefficient S, and (d) absorption coefficient α at 1 eV. All data were measured at room temperature. The samples were in the different heat treatment states: as-deposited, Ar heat treatment, and S heat treatment. Changes in the properties mainly appeared above 200 °C. As shown in Figure 4a, the band gap hardly changed in heat treatment up to 300 °C, but increased to 2.0 eV in heat treatment at 400 °C. The band gap increased to 2.3 eV by S heat treatment. The amount of change in the band gap after S heat treatment corresponded to a blue shift of 0.3 eV. The cause of the increase in Eg due to S heat treatment is unknown.

Figure 4b shows the electrical resistivity change after the heat treatment. The resistivity of as-deposited In-S film was 1.2 × 103 Ωm. This value is almost the same as the resistivity of In2S3 thin films [10]. Resistivity was hardly changed by heat treatment up to 200 °C, but it was greatly decreased by heat treatment at 300 °C. By contrast, resistivity was almost recovered to the initial value by S heat treatment.

Figure 4c and d shows the changes in the Seebeck coefficient and the absorption coefficient at 1 eV after heat treatment. As the sign of the Seebeck coefficient was negative, majority of the carriers of In-S were identified as electrons. The decrease in the absolute value of the Seebeck coefficient reflects the increase in carrier density [18]. If the change in carrier mobility is ignored, the significant decrease in resistivity in Figure 4b corresponds to an increase in carrier density. Both the Seebeck coefficient and absorption coefficient at 1 eV showed a tendency to increase up to 400 °C by heat treatment. By contrast, the recovery of the Seebeck coefficient and absorption coefficient at 1 eV resulted from the S heat treatment. These results suggest that the carrier densities were increased by Ar heat treatment and lowered by S heat treatment. This is an important experimental result showing the possibility of In-S carrier control by heat treatment.

Figure 5 shows the temperature dependence of electrical resistivity for as-deposited In-S films. The arrows indicate heating and cooling. After reaching the maximum temperature, the heater was turned off and cooled. In this experiment, the sample was placed in vacuum. The thermal activation energies were estimated to be 0.8 eV for heating and 0.5 eV for cooling. As the InS films exhibit n-type conductivity, it suggests that a donor level exits below the conduction band fixing the Fermi level position. An Ar heat treatment results in an increase of these defects as revealed by the increase of α around 1 eV. To maintain electrical neutrality the Fermi level moves from 0.8 eV below the conduction band up to 0.5 eV. As a consequence the carrier density increases as shown by the decrease of the resistivity.

Figure 6 shows the schematic diagrams of the density of states in the (a) as-deposited and (b) annealed In-S films based on the experimental results. The optical band gap is 1.9 eV, and the Fermi level Ef estimated from the activation energy is located at 0.5–0.8 eV below the bottom of the conduction band. The subgap absorption observed in the PDS corresponds to the optical transition from the valence band to the localized level. The difference in localized density reflects the intensity of the subgap absorption at 1 eV. The density of states in Figure 6 can well explain the changes in various properties due to heat treatment.

What happens to In-S during heat treatment? The answer to this question depends on the behavior of sulfur. A known example of desorption of sulfur occurring by heat treatment, there is tin sulfide. In the case of tin sulfide, resistivity decreases with heat treatment [19]. Similarly, a small amount of sulfur is desorbed from In-S during Ar heat treatment. According to the experimental results in the In-S films, sulfur desorption seems to be promoted by heat treatment at 300–400 °C. The effect of S heat treatment is slightly complicated. As the sulfurization reaction occurs, sulfur vacancies decrease. At the same time, the process of forming metal sulfide molecules and detaching them from the film occurs. The behavior of sulfur affects the properties of In-S films. The sulfurization reaction results in the widening of the band gap. This idea is in good agreement with the experimental result in Figure 4a. The change in the carrier concentration is described based on desorption and incorporation of sulfur. Desorption of S occurs in Ar heat treatment, increasing the concentration of vacancies. Therefore, to maintain electrical neutrality, the Fermi level moves up toward the conduction band, finally, increasing the electron concentration. In S heat treatment, the concentration of vacancies decreases by incorporating S, and the Fermi level moves back toward the mid-gap, finally decreasing the electron concentration. Evidently, it was found that the reversible characteristic changes shown in Figure 4b–d can be explained by sulfur desorption and incorporation in the In-S films. This is an important finding that enables the control of the carrier characteristics of the In-S films.

thumbnail Fig. 1

XRD patterns of sulfide powder and representative In-S. The corresponding peaks of c-In2S3, t-In2S3, and In6S7 are marked by closed squares, open circles, and closed circles, respectively.

thumbnail Fig. 2

Raman spectra of the as-deposited and thermally annealed In-S films.

thumbnail Fig. 3

Optical absorption spectra of the as-deposited and thermally annealed In-S films. The arrows indicate the evolution of the defect absorption with Ar heat treatment (red arrow) and S heat treatment (blue arrow).

thumbnail Fig. 4

Heat treatment temperature dependencies of the properties of In-S films. (a) indirect band gap Eg , (b) electrical resistivity ρ, (c) Seebeck coefficient S, and (d) subgap absorption coefficient α at 1 eV. Closed squares show the data of the sample after S heat treatment (blue arrow).

thumbnail Fig. 5

Temperature dependence of electrical resistivity in In-S films. The arrows indicate heating and cooling.

thumbnail Fig. 6

Schematic diagram of the density of states in the (a) as-deposited and (b) annealed In-S films. Ef is the Fermi energy and EA is the energy of the subgap absorption.

4 Conclusions

The optical and electrical properties of In-S films were investigated. In-S films were heat treated in Ar gas in the temperature range of 100–400 °C. Some of the annealed samples were heat treated at 300 °C with sulfur powder. The band gap of the as-deposited In-S film was 1.9 eV, which changed to 2.0 eV by Ar heat treatment at 300 °C and to 2.3 eV by S heat treatment. The electrical resistivity, Seebeck coefficient, and absorption coefficient at 1 eV showed reversible changes between Ar heat treatment and S heat treatment. Desorption or incorporation of sulfur into the In-S film by Ar or S heat treatment caused the carrier concentration to increase or decrease, respectively. These results show the feasibility of carrier control in In-S films by heat treatments.

Acknowledgments

This work was supported by Gunma University, Japan. The author would like to thank Mr. K. Imai (Gunma Univ.) for experimental assistance.

References

  1. T. Kamiya, H. Hosono, NPG Asia Mater. 2 , 15 (2010) [Google Scholar]
  2. J.S. Park, W.J. Maeng, H.S. Kim, J.S. Park, Thin Solid Films 520 , 1679 (2012) [Google Scholar]
  3. D. Braunger, D. Hariskos, T. Walter, H.W. Schock, Sol. Energy Mater. Sol. Cells 40 , 97 (1996) [Google Scholar]
  4. M.A. Mughal, R. Engelken, R. Sharma, Solar Energy 120 , 131 (2015) [CrossRef] [Google Scholar]
  5. U.G. Gutiérrez, M.L.A. Aguilera, C.H. Vasquez, J.M.F. Márquez, M.A.G. Trujillo, D.J. Olarte, J.R.A. Hernández, A.R. Millán, Phys. Status Solidi A 215 , 1700428 (2018) [CrossRef] [Google Scholar]
  6. J. Bartolom, D. Maestre, A. Cremades, Phys. Status Solidi A 215 , 1800261 (2018) [CrossRef] [Google Scholar]
  7. C.D. Lokhande, A. Ennaoui, P.S. Patil, M. Giersig, K. Diesner, M. Muller, H. Tributsch, Thin Solid Films 340 , 18 (1999) [Google Scholar]
  8. S. Gorai, P. Guha, D. Ganguli, S. Chaudhuri, Mater. Chem. Phys. 82 , 974 (2003) [Google Scholar]
  9. T.T. John, S. Bini, Y. Kashiwaba, T. Abe, Y. Yasuhiro, C.S. Kartha, K.P. Vijayakumar, Semicond. Sci. Technol. 18 , 491 (2003) [Google Scholar]
  10. R.S. Mane, C.D. Lokhande, Mater. Chem. Phys. 78 , 15 (2003) [Google Scholar]
  11. N. Barreau, A. Mokrani, F. Couzinié-Devy, J. Kessler, Thin Solid Films 517 , 2316 (2009) [Google Scholar]
  12. S.P. Nehra, S. Chander, A. Sharma, M.S. Dhaka, Mater. Sci. Semicond. Process. 40 , 26 (2015) [Google Scholar]
  13. A.C. Boccara, D. Fournier, J. Badoz, Appl. Phys. Lett. 36 , 130 (1980) [Google Scholar]
  14. T. Gotoh, J. Non-Cryst. Solids 358 , 2366 (2012) [CrossRef] [Google Scholar]
  15. Z. Wang, S. Wang, J. Liu, W. Jiang, Y. Zhou, C. An, J. Zhang, Materials 9 , 329 (2016) [CrossRef] [Google Scholar]
  16. E. Kärber, K. Otto, A. Katerski, A. Mere, M. Krunks, Mater. Sci. Semicond. Process. 25 , 137 (2014) [Google Scholar]
  17. W. Rehwald, G. Harbeke, J. Phys. Chem. Solids 26 , 1309 (1965) [Google Scholar]
  18. D. Emin, in Thermoelectrics Handbook: Macro to Nano , edited by D.M. Rowe (CRC Press, Boca Raton, FL, 2006), Chap. 5, pp. 1–7 [Google Scholar]
  19. T. Gotoh, K. Yazawa, K. Imai, Key Eng. Mater. 596 , 21 (2014) [Google Scholar]

Cite this article as: Tamihiro Gotoh, Effect of heat treatments on the electronic properties of indium sulfide films, Eur. Phys. J. Appl. Phys. 89, 20301 (2020)

All Figures

thumbnail Fig. 1

XRD patterns of sulfide powder and representative In-S. The corresponding peaks of c-In2S3, t-In2S3, and In6S7 are marked by closed squares, open circles, and closed circles, respectively.

In the text
thumbnail Fig. 2

Raman spectra of the as-deposited and thermally annealed In-S films.

In the text
thumbnail Fig. 3

Optical absorption spectra of the as-deposited and thermally annealed In-S films. The arrows indicate the evolution of the defect absorption with Ar heat treatment (red arrow) and S heat treatment (blue arrow).

In the text
thumbnail Fig. 4

Heat treatment temperature dependencies of the properties of In-S films. (a) indirect band gap Eg , (b) electrical resistivity ρ, (c) Seebeck coefficient S, and (d) subgap absorption coefficient α at 1 eV. Closed squares show the data of the sample after S heat treatment (blue arrow).

In the text
thumbnail Fig. 5

Temperature dependence of electrical resistivity in In-S films. The arrows indicate heating and cooling.

In the text
thumbnail Fig. 6

Schematic diagram of the density of states in the (a) as-deposited and (b) annealed In-S films. Ef is the Fermi energy and EA is the energy of the subgap absorption.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.