J. Mater. Sci. Technol., 2010, 26(1), 93-96. Properties of Gallium Phosphide Thick Films Prepared on Zinc Sulfide Substrates by Radio-Frequency Magnetron Sputtering Yangping Li† and Zhengtang Liu School of Materials Science and Engineering, Northwestern Polytechnical University, Xi0 an 710072, China [Manuscript received November 3, 2008, in revised form March 23, 2009] Radio-frequency (RF) magnetron sputtering was employed to prepare gallium phosphide (GaP) thick films on zinc sulfide (ZnS) substrates by sputtering a single crystalline GaP target in an Ar atmosphere. The infrared (IR) transmission properties, structure, morphology, composition and hardness of the film were studied. Results show that both amorphous and zinc-blende crystalline phases existed in the GaP film in almost stoichiometric amounts. The GaP film exhibited good IR transmission properties, though the relatively rough surface and loose microstructure caused a small loss of IR transmission due to scattering. The GaP film also showed a much higher hardness than the ZnS substrate, thereby providing good protection to ZnS. KEY WORDS: Radio-frequency magnetron sputtering; Gallium phosphide; Thick film; Infrared transmission 1. Introduction Zinc sulfide (ZnS) is one of the most promising materials used for long-wave infrared (LWIR, 8–12 µm) windows and domes owing to its wide transparent waveband, relatively low cost and availability in large sizes and complex shapes[1–3] . However, ZnS has high reflection, low hardness and poor environmental durability in harsh conditions. Thus, efforts have been made to prepare IR antireflective and protective films for ZnS[4–6] . To improve their protective performance, the films were usually prepared as thick films[7] . Among these films, gallium phosphide (GaP) is a preferred film due to its excellent mechanical, erosion protective and IR transmission properties. Moreover, GaP has frequently been incorporated into antireflective and protective multilayers[8] . So far, the preparation and properties of GaP thick films have been extensively studied. Klocek et al.[9] and Wilson et al.[10] reported the growth of polycrystalline and † Corresponding author. Ph.D.; Tel.: +86 29 88492178; E-mail address: liyp@nwpu.edu.cn (Y.P. Li). epitaxial GaP thick films by metal-organic chemical vapor deposition (MOCVD). Gibson et al.[11] prepared thick GaP films by reactive sputtering (RS) and plasma-assisted chemical vapor deposition (PACVD). Dong et al.[12] employed radio-frequency plasmaenhanced chemical vapor deposition (RF-PECVD) to deposit thick GaP films. In these preparation methods, PH3 and Ga(CH3 )3 , which are hazardous and poisonous, were usually utilized as the precursor material. Therefore, extra measures should be taken to meet stringent health and safety requirements. In this work, GaP thick films were deposited on ZnS substrates by RF magnetron sputtering starting from a single crystalline GaP target in an Ar atmosphere without the use of toxic reactive gas. The properties of IR transmission, structure, morphology, composition and hardness of the deposited GaP thick film were reported. 2. Experimental Thermo-pressed ZnS discs (φ20 mm×5 mm) and 94 Y.P. Li et al.: J. Mater. Sci. Technol., 2010, 26(1), 93–96. a single crystalline GaP disc (φ50 mm×5 mm) were used for the substrates and target, respectively. After being cleaned, the ZnS substrate was mounted on the earthed anode and heated at a fixed temperature throughout the deposition process. The target was mounted on the water-cooled cathode, into which a 13.56 MHz RF power was capacitively coupled. As the working gas, high purity Ar (99.99%) was introduced into the vacuum chamber through a gas flow controller after the base vacuum had been pumped to <3.0×10−5 Pa. In order to efficiently prepare thick GaP films with low IR absorption, the deposition parameters were optimized to deposit stoichiometric GaP films at a relatively high deposition rate[13] . The optimized parameters were: RF power 80W, Ar gas flow ratio 6.0 sccm (standard-state cubic centimeter per minute), working gas pressure 0.07 Pa, substrate temperature 390◦ C and target-substrate distance 60 mm. The deposition rate was approximately 14 nm/min. The IR transmittance spectra of the specimens were measured with a Nicolet 60 SXR Fourier transform infrared (FTIR) spectrometer. X-ray diffraction (XRD) was conducted for structure analysis on a PANalytical X0 Pert HighScore diffractometer (CuKα, λ=0.154 nm) at a grazing incidence angle of 0.5◦ . The morphologies of the GaP films were investigated by using a JSM5800 scanning electron microscope (SEM), and the elemental content in the films was studied by using an energy dispersive X-ray spectrometer (EDS) fitted to the SEM . The hardness of the GaP film was tested with an HX-1000 Vickers hardness tester. 3. Results and Discussion 3.1 IR transmission performance of the GaP film For IR transmission performance evaluation, the IR transmittance spectra of GaP/ZnS (i.e., GaP film on ZnS substrate) and the uncoated ZnS were analyzed and the results are shown in Fig. 1. The spectra were analyzed between 400 and 5000 cm−1 (2– 25 µm waveband) and the insets were between 833 and 1250 cm−1 (8–12 µm). The spectrum of GaP/ZnS was below that of ZnS because the refractive index of GaP (typically 2.9 for bulk GaP) is larger than that of ZnS (typically 2.2 for bulk ZnS). Due to interference between the light reflected from the film surface and the GaP/ZnS interface, the spectrum for GaP/ZnS was a smooth wavy curve having many maxima and minima. The maxima did not contact the spectrum curve for uncoated ZnS, indicating an IR transmission loss caused by absorption or scattering in the GaP film. The IR transmission loss can be estimated by the average transmission loss, ∆T , over the 8–12 µm waveband. ∆T is expressed as ∆T =∆T 0 − ∆T 00 , where ∆T 0 is the average transmittance of uncoated ZnS and ∆T 00 is the average value of the envelope of Fig. 1 Infrared transmittance spectra of uncoated ZnS and GaP/ZnS samples Fig. 2 XRD pattern of the GaP thick film the maxima over the 8–12 µm waveband. For the GaP film, the value ∆T is equal to 1.42%. Through fitting the transmittance spectrum of GaP/ZnS[14] , the film thickness and refractive index at a wavelength of 10 µm were found to be 14.83 µm and 3.02, respectively. Since the IR transmission loss caused by absorption or scattering was small, the absorption and scattering of the film were neglected for simplification of the fitting. Thus, the drop in the transmittance of GaP/ZnS caused by absorption or scattering is ascribed to the rise in refractive index of the film. As a result, the obtained refractive index was a little larger than that of the bulk GaP. 3.2 Structure, morphology and composition of the GaP film The optical performance of films is determined by their properties of structure, morphology and composition. Therefore, these properties of the GaP film were investigated. Figure 2 shows the XRD pattern of the GaP film. The X-ray diffractogram was characterized by the superposition of broad humps and peaks. The humps indicated the existence of an amorphous phase in the GaP film. The peaks are attributed to the (111), (220) and (311) diffraction planes of the zinc-blende crystalline GaP and these peaks are in good agreement with the standard JCPDS data file No. 32-0397. The average size of the crystalline grain Y.P. Li et al.: J. Mater. Sci. Technol., 2010, 26(1), 93–96. 95 Fig. 3 Surface (a) and cross-section (b) SEM images, EDS spectrum (c) and cross- section line-scan analysis (d) of the GaP thick film calculated by the (111) peak using Scherrer0 s formula was approximately 10 nm. Figure 3(a) and (b) show the surface and crosssection morphologies of the GaP film, respectively. Figure 3(a) reveals a relatively rough but uniform film surface. Micro holes and grains were found to exist in the GaP film. From Fig. 3(b), it can be seen that the part of the GaP film adjacent to the substrate was featureless and compact, while with increasing thickness, the GaP film tended to have a columnar growth and became a little loose. Micro-holes were also found in the cross-section micrograph. Columnar growth of a film always leads to a rough surface and the presence of micro-holes between the columns in the film. Moreover, when the deposition rate was high, the atoms deposited onto the film-growth surface would easily meet other atoms to form stable clusters prior to full diffusion. Hence, on one hand, the micro-holes that had formed on the film-growth surface would be covered and embedded within the film; while on the other hand, grains would be formed in the film. Therefore, the rough surface and micro-holes were attributed to the high deposition rate to some extent. Figure 3(c) shows the EDS spectrum of the GaP film, and the inset shows the atomic content of the film. The elements found were mainly Ga and P, plus a slight amount of O, which was incorporated into the GaP film due to the leakage of air into the vacuum chamber during deposition. The content of P and Ga in the film was 50.26% and 46.76% (in at. pct), respectively, showing that the GaP film was almost stoichiometric. Figure 3 (d) shows the cross-section line-scan analysis of the GaP film. The Lscanning path was shown in Fig. 3(b), beginning at . It can be seen in Fig. 3(d) that the composition of the GaP film was homogeneous in the thickness direction. At the GaP/ZnS interface, all the curves for S, Zn, P and Ga changed in a sloping manner, indicating the diffusion of P and Ga into the ZnS substrate. The diffusion can help to improve the adhesion of the film to the substrate. IR transmission loss may be caused by scattering and absorption. However, it is hard to distinguish the contributions of these two factors to the transmission loss. The content of Ga and P in the GaP film is almost stoichiometric, but the film has a relatively loose structure and rough surface. Hence, the major source of its IR transmission loss was due to the scattering occurring both at the grains and holes in the bulk and at the surface of the film, rather than caused by the absorption. 3.3 Hardness of the GaP film The hardness of films is essential to their protective performance. So the Vickers hardness of the GaP film and the uncoated ZnS was tested and the results are shown in Fig. 4. The hardness of the GaP film decreased with increasing load, but the variation of 96 Y.P. Li et al.: J. Mater. Sci. Technol., 2010, 26(1), 93–96. phases existed in the film. Although a small amount of IR transmission loss was caused by scattering, the GaP film had good IR transmission properties. (3) The hardness of the GaP film was much higher than that of the ZnS substrate; hence the film is anticipated to give good protection to ZnS. Acknowledgement This work was supported by the Aviation Science Foundation of China under grant No. 2008ZE53043. Fig. 4 Vickers hardness of the GaP thick film the hardness became moderate at relatively high load. This phenomenon can be explained as follows. When the load was relatively large, the ZnS substrate was also indented to some extent despite the large thickness of the GaP film. Hence the hardness tested at relatively high load reflected the integrative hardness of the GaP film and the ZnS substrate and such an integrative hardness tended to become constant with increasing load. Yet when the load was relatively low, the substrate was not indented. Thus the hardness tested at small load was close to the real hardness of the GaP film. After all, the hardness of the GaP film was much higher than that of the ZnS substrate according to Fig. 4. Furthermore, the GaP film could not be peeled from the ZnS substrate in the qualitative adhesion test with peeling method, suggesting the film well adhered to the substrate. Therefore, the GaP film could provide good protection to the ZnS substrate. 4. Conclusions (1) Stoichiometric GaP thick films were successfully prepared on ZnS substrates by pure RF magnetron sputtering. (2) Both amorphous and zinc-blende crystalline Volume 26 Number 1 pp 1-96 2010 REFERENCES [1 ] E.S. Kelly, R.J. Ondercin, J.A. Detrio and P.R. Greason: Proc. SPIE, 1997, 3060, 68. [2 ] C.C. Clark, A.H. Lettington, S.J. Wakeham, P.S. Jones and D. Waterman: Proc. SPIE, 2001, 4375, 266. [3 ] S. Joseph, O. Marcovitch, Y. Yadin, A. Steimberg and H. Zipin: Proc. SPIE, 2005, 5786, 373. [4 ] E.M. Waddell, D.R. Gibson and J. Meredith: Proc. SPIE, 1994, 2286, 364. [5 ] D.C. Harris: Proc. SPIE, 1997, 3060, 17. [6 ] J. Askinazi and A. Narayanan: Proc. SPIE, 1997, 3060, 356. [7 ] S. Joseph, O. Marcovitch, Y. Yadin, D. Klaiman, N. Koren and H. Zipin: Proc. SPIE, 2007, 6545, 6545OT. [8 ] D.R. Gibson, E.M. Waddell, S.A.D. Wilson and K.L. Lewis: Proc. SPIE, 1992, 1760, 178. [9 ] P. Klocek, J.T. Hoggins and M. Wilson: Proc. SPIE, 1992, 1760, 210. [10] M. Wilson, M. Thomas, I.M. Perez and D. Price: Proc. SPIE, 1994, 2286, 108. [11] D.R. Gibson, E.M. Waddell and K.L. Lewis: Proc. SPIE, 1994, 2286, 335. [12] L.H. Dong, Y.J. Sun and X.Z. He: Proc. SPIE, 2007, 6722, 672201. [13] Y.P. Li, Z.T. Liu, H.L. Zhao, W.T. Liu and F.Yan: Acta Phys. Sin., 2007, 56(5), 2937. (in Chinese) [14] Y.P. Li, Z.T. Liu, H.L. Zhao and Q. Li: Acta Opt. Sin., 2006, 26(10), 1589. (in Chinese)