Journal of ELECTRONIC MATERIALS, Vol. 43, No. 11, 2014
DOI: 10.1007/s11664-014-3325-9 Ó 2014 The Minerals, Metals & Materials Society
Photoassisted Chemically Deposited Tin Sulfide Thin Films Based on Two Different Chemical Formulations T.L. REMADEVI,1,2 A.C. DHANYA,1,3 and K.DEEPA1,2 1.—School of Pure and Applied Physics, Swami Ananda Theertha Campus, Kannur University, Edatt, Kerala, India. 2.—Department of Physics, Pazhassi Raja N.S.S. College, Mattannur, Kerala, India. 3.—e-mail:
[email protected]
Photoassisted chemical deposition is a customized form of chemical bath deposition where the reaction is carried out in the presence of ultraviolet light. Deposition of tin sulfide films was carried out by this method using two different chemical baths. The as-prepared samples from the acetone bath were crystalline, exhibiting the orthorhombic structure of the Sn2S3 phase, but those from the glacial acetic acid bath were amorphous. The crystallinity of the films was improved on annealing. The deposition rate was found to depend on the pH of the bath and the chemical formulation. Distinct morphology was obtained for as-grown films. The films from the acetone bath were compact with uniform morphology of needle-shaped grains having equal diameters and lengths. The films from the glacial acetic acid bath were similar, with smaller needles. The high absorption coefficients of as-grown and annealed films show their potential application as absorber layers in photovoltaic devices. The refractive index was estimated from the reflectance of the films. The estimated activation energies of the as-prepared films from the acetone and glacial acetic acid baths were 0.4 eV and 0.46 eV, whereas those of the annealed samples were 0.2 eV and 0.44 eV, respectively. The activation energy was found to decrease for annealed films due to a decrease in trap sites. Key words: UV source, SnS, various solvents, activation energy
INTRODUCTION IV–VI binary semiconductors have established a prominent place in the field of photovoltaic devices1 working in the visible region due to their nearoptimum bandgap. They also have an important position in optoelectronic devices2 and infrared (IR) detectors working in the IR region due to their high transmittance. The relative abundance of the nontoxic components Sn and S of tin sulfide in the Earth’s crust makes this material technologically significant in the photovoltaic industry for reducing the production cost of thin-film solar cells. The theoretical photovoltaic conversion efficiency of tinsulfide-based solar cells may reach up to 25% with optimum bandgap, making them competitive with (Received October 16, 2013; accepted July 9, 2014; published online August 14, 2014)
3984
currently available CdTe and CuInS2 absorber materials.3 Tin sulfide is a layered semiconductor with orthorhombic structure having high absorption coefficient and both direct and indirect bandgaps in the range of 1.2 eV to 1.5 eV and 1.0 eV to 1.2 eV, respectively.4,5 Akkari et al. obtained an indirect bandgap of 1.76 eV by the chemical bath deposition method6 and concluded that this higher value may be due to the presence of orthorhombic phases of SnS. Though various physical and chemical techniques have been applied by many researchers, development of films with required properties suitable for specific applications in a cost-effective manner remains a challenge. The optical and transport properties of as-prepared tin sulfide can be engineered by simple chemical methods even without the help of doping. The absorption coefficient and electrical conductivity of as-prepared and
Photoassisted Chemically Deposited Tin Sulfide Thin Films Based on Two Different Chemical Formulations
annealed films are found to depend on the host solutions which acted as either solvent or complexing agent, or both. Various combinations of precursors in different host solutions have been reported previously.1,3,6 Orthorhombic tin sulfide films were deposited chemically by Hankare et al.7 by dissolving tin chloride in tartaric acid solution. Tin sulfide films deposited using acetone and glacial acetic acid as solvents by the chemical bath deposition method have also been reported in the literature.8,9 Films deposited via ammonium citrate solution by the chemical bath deposition method exhibited better electrical properties.10,11 Various phases of nanocrystalline tin sulfide with different morphologies have been synthesized by choosing different alkaline solutions as the reaction medium, due to the difference in their chelating ability. Stoichiometric orthorhombic SnS phase has been achieved starting from SnCl2 dissolved in a mixture of isopropyl alcohol and deionized water with thiourea as the sulfur source at Sn/S ratio of 1.06 by chemical spray pyrolysis (CSP).12 Using the same precursors in an acidic bath, Sn2S3 has also been deposited by CSP.13 The growth conditions were optimized to obtain tin monosulfide with orthorhombic structure by CBD.14 Use of an acidic cationic precursor and alkaline anionic precursor resulted in SnS films of nanosized spherical grains when using the successive ionic layer adsorption and reaction (SILAR) method.15 In this paper, we report studies on the physical, optical, and transport properties of as-grown and annealed tin sulfide thin films prepared using two different host solutions (acetone and glacial acetic acid) in alkaline medium keeping the precursors invariant, by the photoassisted chemical deposition technique.16 EXPERIMENTAL PROCEDURES All chemicals used for preparing the baths were of analytical grade. Soda lime glass substrates were cleaned using chromic acid and distilled water prior to deposition. The chemical composition of the first bath included 1 M SnCl2 dissolved in 5 mL acetone, 3 mL triethanolamine (TEA), and 10 mL NH3 with pH of 9.22. Finally, 1 M thioacetamide was added to this and stirred well. The glass substrates were introduced into the reaction bath, and deposition was carried out for 5 h under UV illumination and then 14 h at room temperature. The other bath contained the same components with the difference that the same cationic precursor was dissolved in 10 mL glacial acetic acid followed by addition of 1 mL HCl, 15 mL TEA, and 20 mL ammonia solution to obtain a pH of 8.07. After thorough stirring of this solution with 1 M thioacetamide, deposition was carried out for 3 h under UV illumination. All samples were removed after deposition, washed in distilled water, and dried in air using a hot air blower. Some of the samples were annealed in air to
3985
100°C for 15 min. The as-prepared and annealed samples from the two separate baths are named AR, A100, GR, and G100, respectively. Photoassisted Chemical Deposition Mechanism A 125-W UV lamp with wavelength of 355 nm was used to irradiate the reaction bath. The experimental setup is shown in Fig. 1. The UV light energy was absorbed by the bath and utilized to increase the kinetic energy and hence the interdiffusion of adsorbing particles. This photoexcitation helps to dissociate thioacetamide, whereas the complexing agents help the slow release of metallic ions. The ion-by-ion mechanism in the reaction bath leads to the formation of tin sulfide samples. Absorption of UV rays by the glass substrate also supports the adherence of ions onto it. The amount of UV energy absorbed by the reaction vessel is small, as it is made of quartz. Photoassisted chemical deposition is less intense and slower than thermally activated processes, hence the deposition rate will also be lower. It is seen that, after completion of the reactions in CBD, the vessel containing the solution exhibits some precipitate of the components at the bottom. Chemical bath deposition is inefficient to convert the preparatory materials effectively into thin-film form due to either the homogeneous reaction or deposition on the reaction vessel. However, in photoassisted chemical deposition (PCD), heterogeneous reaction predominates and the deposition on the vessel is very thin. In the case of PCD, for deposition of most sulfides, even after reaction for 5 h to 6 h, the reaction solution can be used again for further deposition. Thus, cent percent utilization of the metal ions can be ensured to form thin films when using optimized parameters.17 The solution can be continuously stirred using a magnetic stirrer to avoid sedimentation on the substrate. Photoassisted chemical deposition is widely accepted since it is cost effective and suitable for
Fig. 1. Experimental setup of the photoassisted chemical deposition technique.
3986
Remadevi, Dhanya, and Deepa
large-area deposition, hence being highly useful in solar cell technology. The kinetics of the formation of tin sulfide thin films from tin metal ions and thioacetamide has been discussed elsewhere.16 It is to be noted that the surface tension values of the solvents (25 mN/m) are lower than that of water (72 mN/m), hence the attractive force between ions in the solution and the surface of the substrate will be more effective. It is observed that SnCl2 dissolved more easily in acetone than in glacial acetic acid. However, the dissociation of metal ions was faster in the latter. First, it appeared milky white in color, then turned reddish black and then to complete black after just 15 min, indicating the start of deposition. However, the milky-white appearance of the acetone bath turned complete black only after 1 h to 2 h, then deposition started. Hence, the deposition rate of the samples was higher in the second bath, indicating its dependence on the pH of the bath and the chemical formulation. In the present case, cent percent utilization of the reaction bath was completed within 1 h to 2 h. The as-prepared films were observed to be smooth, pin hole/crack free, and highly adherent to the substrate. Characterization Techniques The thickness of the samples was determined by the gravimetric method. The as-prepared samples were characterized with respect to their structural, morphological, optical, and electrical behavior. Structural analysis was carried out using an x-ray diffractometer (Bruker AXS D8) with Cu Ka radia˚ as the source. Mortion at wavelength of 1.5405 A phological studies were carried out using a JEOL model JSM6490. Optical studies were carried out using a Hitachi U-3410 UV–Vis–NIR spectrophotometer in the range from 200 nm to 1000 nm. The dependence of the conductivity of the samples on temperature was measured using a four-probe setup. RESULTS AND DISCUSSION Structural Characterization X-ray diffraction (XRD) studies enable identification of crystal structure and phase, and estimation of grain size. The XRD profiles of as-grown and annealed samples from the two different baths are shown in Fig. 2. The films from the first bath showed a well-defined peak at 32.398° corresponding to the (2 4 0) plane of Sn2S3 with orthorhombic structure [Joint Committee on Powder Diffraction Standards (JCPDS) card no. 72-0031]. Due to the difference in the chemical reactions between the two solvents, the manner of deposition is entirely different. The reaction speed with acetone is comparatively slower than with glacial acetic acid. As a result, the average deposition rate with acetone is slower than with glacial acetic acid. Increasing the
deposition rate with the glacial acetic acid bath will definitely result in an improvement in thickness. The crystallites rearranged on annealing, and the number of crystallites along the preferential direction reduced. Though the peak intensity reduced, there was no change in the predominant orientation. The annealed samples showed a slight redshift in peak position, but the compressive strain was not appreciable. The as-grown samples from the glacial acetic acid bath were amorphous or formed of nanocrystallites. However, annealing of the samples resulted in polycrystalline behavior, which may be due to a decrease of internal strain. The preferential orientation along (0 0 4) at 30.36° and other peaks along (1 0 0) at 28.18° and (1 0 3) at 37.13° match the SnS2 phase of tin sulfide (JCPDS card no. 89-3198). The presence of elemental sulfur (JCPDS card no. 77-0228) is also observed in the annealed film, as shown in the figure. Formation of herzenbergite SnS and mixed phase was reported previously in tin sulfide thin films chemically deposited from acetone and glacial acetic acid baths, respectively.8 However, in this work, the occurrence of more specific crystal structures in all samples may be the consequence of the UV assistance. Literature also reports that lower deposition temperature favors formation of SnS2 and Sn2S3 phases over SnS.4 A histogram showing the thickness of the films is depicted in Fig. 3. It was found that the maximum thickness was observed for the glacial acetic acid deposited films. Due to the difference in the chemical reactions between the two solvents, the manner of deposition is entirely different. The reaction speed with acetone is comparatively slower than with glacial acetic acid. The average deposition rate with acetone is slower than with glacial acetic acid. Increasing the deposition rate with the glacial acetic acid bath will definitely result in an improvement in thickness. All the annealed films exhibited lower thickness than the as-prepared films. This might be due to escape of hydroxide content from the films on annealing. The average crystallite sizes of the films calculated using the Debye–Scherrer formula18 were in the range from 23 nm to 46 nm. Improvement in thickness from 22 nm to 24 nm was also reported earlier with increase in deposition time.19,20 Shaaban et al. in their study on ellipsometric investigations of the optical constants of material formed by the thermal evaporation technique, reported that crystallite size increases with film thickness.21 However, in our study, there was no appreciable change in grain size on annealing films from the first bath, but it increased on annealing for samples from the second bath. Morphological Characterization The optical and electrical properties of the samples mainly depend on the surface structure. Scanning
Photoassisted Chemically Deposited Tin Sulfide Thin Films Based on Two Different Chemical Formulations
3987
Fig. 2. XRD profiles of samples from the glacial acetic acid bath: (a) as prepared and (b) annealed, and the acetone bath: (c) as prepared and (d) annealed.
places, there are agglomerations or coalescence where the grain size is larger. Various morphologies have been reported when using different reaction baths and deposition techniques. Coalescence of spherical grains was observed when using SnCl2 in tartaric acid and in ethylene glycol by CBD and microwave-assisted CBD, respectively.7,21 However, in the present case, a distinct morphology was achieved by utilizing UV assistance for these chemical combinations. Optical Characterization
Fig. 3. Histogram of film thickness.
electron microscopy (SEM) images of as-grown samples are presented in Fig. 4, showing a compact and uniformly covered morphology of the samples from the acetone bath on the surface of the substrate. The grains are needle shaped, having equal diameters and lengths. The length is observed to be greater than the diameter. However, they are randomly directed. The pin-hole-free structure can enhance the conductivity of these samples, as demanded for photovoltaic devices. The surface structure of the samples from the glacial acetic acid bath also shows a similar texture, but the needles are smaller in size, emerging as nanoworms. At
The absorption spectra of the samples from the acetone and glacial acetic acid baths are depicted in Figs. 5 and 6. The absorbance of all samples was found to be high over the entire visible region. The samples from the acetone bath showed a decrease in absorbance on annealing, whereas those from the glacial acetic acid bath showed an increase. This may be due to their distinct surface morphology. The absorption edge of the annealed samples was red-shifted. The surface of the films plays an important role in determining their application in various fields. As the grains grew bigger, there was an enhancement in the absorbance of the films. Hence, the annealed films from the glacial acetic acid bath exhibited higher absorption than all other films. This might be due to the reduction in the number of grain boundaries in the film. The optical absorption coefficient determined from the Beer–Lambert law22 showed a high value
3988
Remadevi, Dhanya, and Deepa
Fig. 4. SEM images of as-prepared samples from the (a) acetone and (b) glacial acetic acid baths.
Fig. 5. Absorption spectra of samples prepared from the acetone bath: (a) as prepared, and (b) annealed.
(105/cm), confirming the potential application of these samples as absorber material.23 The reflection spectra of both as-prepared and annealed samples from the acetone and glacial acetic acid baths are presented in Figs. 7 and 8. In the case of samples AR and A100, the surface is smoother than for samples GR and G100, and less internal scattering leads to less reflection. The higher reflection for samples GR and G100 than AR and A100 may also be due to the dependence on the host solution. The decreased reflection and absorbance of the annealed samples from the acetone bath show a corresponding increase in transmission. The annealed samples from both baths showed low reflection, since the increase in grain size reduces scattering at boundaries. As the defect levels are removed, the absorption edge becomes red-shifted. The decrease in transmittance of the samples from the glacial acetic acid bath may be due to the elemental sulfur present. The transmission is sensitive to the distribution of grains and their size.
Fig. 6. Absorption spectra of samples prepared from the glacial acetic acid bath:(a) as prepared, and (b) annealed.
For parabolic band structure, the absorption coefficient (a) is related to the bandgap of the material by the Tauc relation24
Photoassisted Chemically Deposited Tin Sulfide Thin Films Based on Two Different Chemical Formulations
3989
ahm ¼ Aðhm Eg Þn ; (1)
Fig. 7. Reflection spectra of samples prepared from the acetone bath: (a) as prepared and (b) annealed.
Fig. 9. Direct bandgap of samples prepared from the glacial acetic acid bath: (a) as prepared and (b) annealed, and the acetone bath: (c) as prepared and (d) annealed.
Fig. 8. Reflection spectra of samples prepared from the glacial acetic acid bath: (a) as prepared, and (b) annealed.
defect density near the band edge. The observed bandgap of the tin sulfide thin films lies in the range of 1.59 eV to 2.06 eV. Solid-State Properties
where n = 1/2 for allowed direct transition and n = 2 for indirect transition. A is a parameter which depends on the transition probability. The absorption coefficient can be deduced from the absorption spectrum using the relation25 a ¼ 2:303A=t;
(2)
where t is the thickness of the as-deposited SnS thin film. In the case of direct transition in the fundamental absorption, (aht)2 shows a linear dependence on the photon energy ht. Extrapolating the linear portion of the graph to the ht axis as shown in Fig. 9, we obtained the direct bandgap energy. It can be seen that the films have a steep optical absorption feature, indicating good homogeneity in the shape and size of grains as well as low
The refractive index of a material influences its optical properties. The refractive index can be estimated from the reflectance. Figures 10 and 11 illustrate the variation of the refractive index of the samples from the acetone and glacial acetic acid baths. Mariappan et al.26 reported that the refractive index of electrodeposited SnS films decreases from 2.2 to 1.6 as the wavelength increases. This decrease in refractive index may be due to the decrease in film density. The refractive index varies from 1.1 to 1.2 linearly in the visible region for AR and A100 and has a curly nature. In the IR region, the refractive index is found to be constant. For the samples from the glacial acetic acid bath, the refractive index variation lies in the range of 1.1 to 1.2 for GR and G100. The behavior remains the
3990
same in the IR region. Consistent with the lower reflection of samples AR and A100, they also possess low refractive index values. The observed range of refractive index is in good agreement with reported values.10 Furthermore, the refractive index is a measure of film density, which could provide information about the voids present in the films. Figures 12 and 13 show the variation of the extinction coefficient for the samples from the acetone and glacial acetic acid baths. The evaluated extinction coefficient shows a maximum value near the band edge and becomes steady in the visible and IR region. High values of extinction coefficient indicate a high degree of surface nonuniformity, whereas low values indicate a higher degree of surface uniformity or surface smoothing. The values of the extinction coefficient were high for AR and
Remadevi, Dhanya, and Deepa
A100, whereas they were comparatively lower for GR and G100. This result is in accordance with the surface morphology. The extinction coefficient increases with wavelength for electrodeposited thin films. This means that more scattering occurs in these films.26 The numerical values of the structural, optoelectronic, and solid-state properties are presented in Table I. It can be concluded that the nature of the solvent plays an important role in determining the various properties of the thin films. Electrical Characterization The activation energy is one of the most important electrical parameters of thin films. Its measurement provides a measure of the trapping levels.
Fig. 10. Refractive index of samples prepared from the acetone bath: (a) as prepared and (b) annealed.
Fig. 13. Variation of extinction coefficient with wavelength for samples from the glacial acetic acid bath: (a) as prepared, and (b) annealed.
Fig. 11. Refractive index of samples prepared from the glacial acetic acid bath: (a) as prepared, and (b) annealed.
Fig. 12. Variation of extinction coefficient with wavelength for samples from the acetone bath: (a) as prepared, and (b) annealed.
Photoassisted Chemically Deposited Tin Sulfide Thin Films Based on Two Different Chemical Formulations
3991
Table I. Numerical values of structural, optoelectronic, and solid-state properties Sample Code
2h (°)
˚) d (A
hkl
Grain Size, D (nm)
n
k
Eg (eV)
AR A100 GR G100
32.398 32.401
2.739 2.748
(2 4 0) (2 4 0)
28.18 30.36 37.13
3.163 2.941 2.422
(1 0 0) (0 0 4) (1 0 3)
23 24 31 46
1.19 1.13 1.15 1.14
0.53 0.59 0.04 0.05
1.62 1.59 2.06 1.86
semiconducting behavior of the films. The increase in conductivity with temperature also confirms the absence of any metallic layer. Increase in the grain size leads to a decrease in the activation energy in both cases. This result was reported earlier in Seto’s model. The estimated activation energy for AR and A100 is 0.4 eV and 0.2 eV and for GR and G100 is 0.46 eV and 0.44 eV, respectively, in agreement with literature reports.7,15 CONCLUSIONS
Fig. 14. Arrhenius plots for samples prepared from the glacial acetic acid bath: (a) as prepared and (b) annealed, and the acetone bath: (c) as prepared and (d) annealed.
It also indicates the sensitivity of the reaction rate to temperature. It is found to depend on the thickness of the film, the chemical formulation, and the method of deposition.27 The temperature dependence of the electrical resistivity over the temperature range of 50°C to 180°C was determined using a four-probe setup for the samples from the two different chemical formulations. Figure 14 displays the corresponding Arrhenius plots for the as-prepared and annealed samples from the acetone and glacial acetic acid baths. In the present case, both annealed films showed higher conductivity and lower activation energy compared with the as-deposited films, which may be due to the crystalline nature of the annealed films, as reported earlier.28 The linear plot indicates
Tin sulfide thin films were fabricated by the photoassisted chemical deposition method using two different chemical formulations. The XRD profiles revealed that the as-prepared films from the acetone bath were polycrystalline, whereas those from the glacial acetic acid bath were amorphous; on annealing, they became polycrystalline, despite the presence of traces of elemental sulfur. The deposition rate was higher for the samples from the glacial acetic acid bath. Both samples showed high absorption over the entire visible region. The refractive index of the samples from the acetone bath was lower due to decreased internal scattering. The variation of the conductivity with temperature showed that, on annealing, the activation energy decreases due to a reduction in trap sites, being highly suitable for photovoltaic applications. ACKNOWLEDGEMENTS The authors express sincere gratitude to SAIF STIC, CUSAT and SAIF STIC, IIT Madras for offering technical support, Mr. Jithesh K, M.G. College Iritty, and Mr. Praveen Kumar K, M.G. University Kottayam. T.L.R. acknowledges KSCSTE for financial support under Project 001/SRSPS/ 2008/CSTE. REFERENCES 1. L. Amalraj, C. Sanjeeviraja, et al., J. Cryst. Growth 234, 683 (2002). 2. B. Ghosh, M. Das, et al., Appl. Surf. Sci. 254, 6436 (2008). 3. J.P. Singh and R.K. Bedi, Thin Solid Films 199, 9 (1991). 4. K.T. Ramakrishna Reddy and P. Purandhara Reddy, Mater. Lett. 56, 108 (2002). 5. T.H. Sajeesh, A.R. Warrier, et al., Thin Solid Films 518, 4370 (2010). 6. A. Akkari, C. Guasch, and N. Kamoun-Turki, J. Alloys Compd. 4901, 80 (2010).
3992 7. P.P. Hankare, A.V. Jadhav, et al., J. Alloys Compd. 463, 581 (2008). 8. M.T.S. Nair and P.K. Nair, Semicond. Sci. Technol. 6, 132 (1991). 9. D. Avellanede, M.T.S. Nair, and P.K. Nair, Thin Solid Films 517, 2500 (2009). 10. D. Avellanede, G. Deldado, et al., Thin Solid Films 515, 5771 (2007). 11. C. Gao and H. Shan, Thin Solid Films 520, 3523 (2012). 12. N. Koteeswara Reddy and K.T. Ramakrishna Reddy, Mater. Chem. Phys. 102, 13 (2007). 13. H. Benhaj Salah, H. Bouzouita, and B. Rezig, Thin Solid Films 480, 439 (2005). 14. Y. Jayasree, U. Chalapathi, et al., Appl. Surf. Sci. 258, 2732 (2012). 15. B. Ghosh, M. Das, et al., Appl. Surf. Sci. 254, 6436 (2008). 16. T.L. Remadevi and A.C. Dhanya, Arch. Phys. Res. 2, 128 (2011). 17. P.K. Nair and M.T.S. Nair, Sol. Energy Mater. Sol. Cells 52, 313 (1998). 18. E. Guneri, C. Ulutaset, et al., Appl. Surf. Sci. 257, 1189 (2010).
Remadevi, Dhanya, and Deepa 19. B.D. Cullity, Elements of X-ray Diffraction (Reading, MA: Addison-Wesley, 1978). 20. B.G. Jayaprakash, A. Amalarani, et al., Chalcogenide Lett. 6, 455 (2009). 21. E.R. Shaaban, M.S. Abd, E.L. Sadek, M. El-Hagary, and I.S. Yahia, Phys. Scr. 86, 015702 (2012). 22. B. Lambert and I.N. Levine, Physical Chemistry, 4th ed. (New York: McGraw-Hill, 1995). 23. R. Bayon, R. Musembi, et al., Sol. Energy Mater. Sol. Cells 89, 13 (2005). 24. J. Tauc, Amorphous and Liquid Semiconductors (New York: Plenum, 1974). 25. F.N. Dultsev, L.L. Vasilieva, et al., Thin Solid Films 510, 255 (2006). 26. R. Mariappan, T. Mahalingam, and V. Ponnuswamy, Optik 122, 2216–2219 (2011). 27. M. Devika, N. Reddy, et al., Semicond. Sci. Technol. 211, 125 (2006). 28. S. Varghese and M. Iype, Orient. J. Chem. 27, 265 (2011).