Physical vapor transport crystal growth of ZnO

Transcription

1 Vol. 35, No. 3 Journal of Semiconductors March 2014 Physical vapor transport crystal growth of ZnO Liu Yang( 刘洋 ), Ma Jianping( 马剑平 ), Liu Fuli( 刘富丽 ), Zang Yuan( 臧源 ), and Liu Yantao( 刘艳涛 ) School of Automation and Information Engineering, Xi an University of Technology, Xi an , China Abstract: Zinc oxide (ZnO) has a wide band gap, high stability and a high thermal operating range that makes it a suitable material as a semiconductor for fabricating light emitting diodes (LEDs) and laser diodes, photodiodes, power diodes and other semiconductor devices. Recently, a new crystal growth for producing ZnO crystal boules was developed, which was physical vapor transport (PVT), at temperatures exceeding 1500 ı C under a certain system pressure. ZnO crystal wafers in sizes up to 50 mm in diameter were produced. The conditions of ZnO crystal growth, growth rate and the quality of ZnO crystal were analyzed. Results from crystal growth and material characterization are presented and discussed. Our research results suggest that the novel crystal growth technique is a viable production technique for producing ZnO crystals and substrates for semiconductor device applications. Key words: ZnO; crystal boules; physical vapor transport (PVT); sublimation; impurity analysis; growth rate DOI: / /35/3/ EEACC: Introduction In recent years, zinc oxide (ZnO) has attracted the attention of many researchers with a wide band gap (3.37 ev) and a very high excitation binding energy (60 mev) Œ1, which makes ZnO superior to GaN for the fabrication of high-brightness emitters (LEDs). Semiconductor devices that can be made using the thin films of metal oxides or III-nitrides grown on ZnO single crystal substrates include, but are not limited to, light emitters, such as UV, visible light emitting diodes (LEDs) and laser diodes (LDs) Œ2; 3. The quality of the semiconductor is highly dependent on the purity and structural characteristics of the ZnO single crystals. So the study for the growth of high-quality ZnO bulk single crystal is very important. Now, the ZnO bulk single crystals of large sizes were grown with the following main three techniques: (1) the hydrothermal technique Œ4 6, (2) the high-pressure melt growth technique Œ7 and (3) the chemicalassisted vapor transport (CVT) technique Œ8 10. Other growth techniques, such as flux growth Œ11 and halide vapor phase epitaxy (HVPE) techniques Œ12, can also produce ZnO single crystals, but not large-size single crystals. Nowadays, commercially available ZnO single crystals are produced using a hydrothermal growth technique, which is conducted in an aqueous solution inside a high-pressure autoclave at a temperature between ı C. However, there are several drawbacks to the hydrothermal growth of ZnO. First, the growth rate is very low, about 4 8 m/h in the c-axis direction; another drawback is that a platinum (Pt) metal crucible is used in the hydrothermal growth of ZnO, which adds a high cost to the crystal production; the main drawback of the hydrothermal growth technique is that ZnO crystal contains large amounts of alkaline metal, such as Li and K (from LiOH and KOH used in hydrothermal growth). These alkaline metal elements are electrically active and hence can be detrimental to device performance Œ14. The high-pressure melt growth technique is a common growth technique for crystal, which has many advantages, such as a high growth rate (> 5 mm/h), the flexibility to grow in any crystal orientation and the ease of doping the crystals to achieve the desired electronic characteristic. But since the ZnO melting point is 1975 ı C, when the temperature reaches ı C, ZnO starts to sublime, which makes the melt growth difficult to manage. Another difficulty in ZnO melt growth is that only iridium (Ir) crucible, which there is a severe degradation, can be used to directly contain ZnO melt. In a CVT growth of ZnO crystals, a ZnO source material in the source zone is reacted with a chemical transport agent (H 2, H 2 O and CO) Œ17 and then transported to the growth zone with a ZnO crystal seed so that a ZnO crystal boule is grown. A CVT growth of ZnO crystals is carried out at lower temperatures ( ı C) in a quartz vessel. But a drawback of a CVT growth process is extremely low growth rates (2 3 mm/day, or 0.08 mm/h) Œ15. Since ZnO sublimes at high temperatures, particularly in the temperature range of ı C, a high-temperature physical vapor transport (PVT) growth technique (or a sublimation growth technique) may be used to grow ZnO crystals. The PVT growth technique has been used for growing many crystals, such as zinc selenide (ZnSe), silicon carbide (SiC) and aluminum nitride (AlN). In order to estimate the sublimation vapor pressures of ZnO at high temperatures and the growth rates in a ZnO PVT growth at such high temperatures, an analysis of the thermodynamics and kinetics of ZnO sublimation/condensation were carried out by the corresponding author in 2006 Œ13, which is similar to a PVT growth of AlN and SiC. The result from that analysis shows that ZnO has sufficiently high vapor pressures at temperatures higher than 1600 ı C so that ZnO crystals can be grown using a PVT technique at a growth rate of at least 1 mm/h. In this paper, we use a PVT growth technique for growing * Project supported by the Special Scientific Research Plan Project of Shaanxi Provincial Education Department, China (No. 08JK376). Corresponding author. majp@xaut.edu.cn Received 12 August 2013, revised manuscript received 27 September Chinese Institute of Electronics

2 Table 1. List of the key growth parameters and the ranges of values in ZnO PVT growth experiments. Growth parameter Range of value ZnO source temperature ( ı C) Axial thermal gradient ( ı C/cm) A mixture of O 2 and 200 : : : 50 an Ar gas (sccm) System gas pressure (Pa) Growth duration (h) 16 Fig. 1. A schematic drawing of the PVT growth method of a ZnO crystal. ZnO crystal on sapphire (Al 2 O 3 / and report the results. We also demonstrate ZnO crystal wafers in sizes up to 50 mm in diameter. The growth rate and crystalline defects of PVT grown ZnO crystals were analyzed using variety of techniques. Results from crystal growth and crystal characterization will be presented and discussed. 2. Experiment A PVT growth is a sublimation and re-condensation process, in which a source material and a seed crystal are placed inside a growth furnace in such way that the temperature of the source material is higher than that of the seed so that the source material sublimes and the vapor species diffuse and deposit onto the seed to form a single crystal Œ16. The advantages of a PVT growth for ZnO include the ease of controlling the growth process and the low cost. It should be pointed that the PVT growth technique described in this paper is different to the chemical-assisted vapor transport (CVT) technique, because CVT is useful for growing crystals at low temperatures, which rely on a chemical transport agent, such as H 2 and H 2 O. However, a PVT growth requires a suitable crucible and a thermal insulation setup. So the crucibles, such as alumina or sapphire, were used as a thermal insulation setup of the PVT growth technique, because they can be heated up to 2000 ı C. To carry out ZnO PVT crystal growth experiments, we employed induction-heated PVT growth furnaces designed and built at fairfield crystal technology. The experiment includes the following steps: (1) placing a source material (zinc oxide powder, purity of 99.9%) at the bottom of an interior crucible (a ceramic containing 99.0% 99.8% Al 2 O 3 / enclosure and placing one crucible lid (Al 2 O 3, purity of %) at the top of the interior crucible enclosure with the crucible lid and the source material separated by a predetermined distance; (2) heating the crucible to predetermined temperatures, where the temperature of the source material is higher than the temperature of the crucible lid; (3) maintaining a pressure within the interior crucible (or the system) enclosure through flowing a gas mixture; and (4) maintaining a temperature distribution within the crucible enclosure thereby causing a ZnO boule growth on the crucible lid. The PVT ZnO growth experimental apparatus are schematically shown in Fig. 1. In the experiment, the ZnO source material in the lower portion of the crucible was a sapphire seed crystal affixed onto the crucible lid; the crucible was in the growth room, which is also Al 2 O 3. Due to the fact that Al 2 O 3 -based materials have a low chemical reactivity with ZnO and a high melting point, they are used to construct the growth zone for ZnO PVT crystal growth experiments. A silicon molybdenum rod (MoSi 2 / is used for heating the furnace. The temperatures at the top and the bottom of the crucible are measured using optical pyrometers so that the ZnO crystal temperature and the ZnO source temperature can be monitored and controlled. The PVT growth furnaces are capable of reaching a temperature of over 1700 ı C in a gas pressure (in pure O 2, an inert gas or a mixture of O 2 and an inert gas) in less than 0.05 MPa, which is sufficient for ZnO PVT growth experiments. Since the ZnO powder will decompose to Zn vapor and O 2 gas at high temperatures (e.g ı C), it is critical to find thermal insulation materials compatible with ZnO powder and the vapor phase in the growth temperature regime. The powder was heated up to an operating temperature of 1500 ı C in about 5 h and kept at a constant temperature for about 16 h under a certain pressure. To obtain the ZnO single crystal boules, in this paper, we study the ZnO polycrystalline growth and explored PVT crystal growth in different conditions and under different pressures. The key parameters and the ranges of values for the PVT growth experiments are listed in Table 1. Crystal growth in PVT methods is promoted by imposing a thermal gradient between a source material (generally polycrystalline solid or powder) and the opposite extreme of a crucible, which is kept at a lower temperature. However, the sublimation and deposition processes are driven by the saturation level of the vapor above the solid ZnO; the crystal growth rate is determined by the dynamics of diffusion and the chemical reaction mechanism for determining the partial pressure of a gas phase component in any temperature. So we pass the different proportions of a mixture of O 2 and Ar gas into the system, which is intended to change the partial pressure of the system. The morphology, grain-size, grain-quality and composition of the source powder and the resulting ZnO crystalline material were characterized using a scanning electron microscope (SEM), equipped with an X-ray (EDX) spectrometer. The information of vibration or rotation of the sample molecules were measured using Raman spectroscopy

3 Fig. 2. A ZnO polycrystalline of about 50 mm in diameter. 3. Results and discussions In this part, the results from the PVT ZnO crystal growth experiments are presented and discussed. The research topics are: (1) analysis of the crystal growth rate and the polycrystalline area, (2) analysis of the crystal XRD pattern, (3) analysis of the crystal Raman spectrum and (4) SEM analysis of ZnO crystal. Details are presented in the following subsections, respectively Analysis of the crystal growth rate and polycrystalline area In the PVT ZnO crystal growth experiment, the growth rate is the most important parameter. An analysis of the average growth rates for ZnO polycrystalline was performed. The result showed that the growth rate is different under different experiments. The median growth rate was about 0.3 mm/h. Further analysis showed that the average growth rate was strongly influenced by three growth parameters, i.e. the source temperature, the axial thermal gradient and the system pressure (O 2 or Ar) in the growth chamber. If the source temperature was lower than 1500 ı C, such as 1400 ı C, only a small portion of the ZnO power sublimated, or it was even not sublimated. While the source temperature was higher than 1500 ı C, such as 1550 ı C, the growth rate increases gradually at a given axial thermal gradient and system gas pressure. Because there exists a maximum growth rate above the stoichiometic pressure for which the balance between the availability of O 2 at the growing surface and the necessary diffusivity of a Zn atom through O 2 gas is optimal. However, when the excess of O 2 partial pressure is increased beyond that point, the growth rate becomes limited by a much slower Zn diffusion in the gas phase; further increasing the pressure results in a fast decay of the growth rate. A source temperature at 1500 ı C was found to achieve a growth rate of 0.8 mm/h, where the system pressure was 480 Pa and the axial thermal gradient was 45 ı C/cm. The color of the PVT-grown ZnO crystal boules ranged from yellow to transparent, depending on the key growth parameters, such as growth temperature and system vapor pressure. Another very important aspect of ZnO crystal growth is the crystal area in crystal boules. In the PVT ZnO crystal growth experiment, ZnO crystals were grown directly onto the 50 mm diameter crucible lids, which resulted in polycrystalline ZnO and they were all 50 mm diameter. But ZnO substrates to be Fig. 3. The XRD diffraction pattern of ZnO PVT growth under different system pressures. (a) 20 Pa. (b) 480 Pa. (c) 2800 Pa. used for fabrication of semiconductor devices have to be in a specific crystal orientation. Therefore, a PVT growth has to be established in order to produce oriented ZnO crystal boules of large diameter Analysis of crystal XRD pattern The XRD patterns are shown in Fig. 3. In the XRD pattern, the diffraction peaks were (002), (102), (103), (004), which represent the polycrystalline morphology. Compared with standard ZnO card, they all had a hexagonal wurtzite structure. As the system pressure changed, the characteristic diffraction peak intensity, such as (002), strengthened and as the system pressure was further increased, the main diffraction peak intensity became weaker. That happened because as the system pressure increased, the partial pressure of O 2 was enhanced, which drove the crystal growth. So the XRD pattern intensity increased rapidly and became acute gradually, which represented that the crystalline quality of the PVT ZnO crystal growth was better and the grain sizes were larger. But when the system pressure was further increased, the XRD pattern intensity decreased and the crystalline quality reduced, because the excess of O 2 partial pressure is increased beyond the optimal growing point, which limited the ZnO crystal growth. The average grain size of a polycrystalline body can be calculated by the Scherrer equation, D m D 0:9=.ˇ cos /; (1) where D m is the average grain size, ˇ is the maximum peak FWHM, is the Bragg angle, and is the X-ray wavelength. The average grain sizes D m are nm, nm and nm in the c-axis direction (002), respectively

4 J. Semicond. 2014, 35(3) Liu Yang et al. Fig. 4. The Raman spectra of ZnO PVT growth under different system pressures. (a) 20 Pa. (b) 480 Pa. (c) 2800 Pa Analysis of crystal Raman spectrum In the Raman spectrum of PVT ZnO crystal growth, there were three peaks and their positions were cm 1, 437 cm 1 and cm 1. The Raman spectra are shown in Fig. 4. The peak positions were not changed in the different system pressure, which indicated that the stresses of PVTgrown ZnO crystals were not changed. However, the peak intensity was changed with the different system pressure. The peak intensity was strengthened with the system pressure increased, such as at Pa and then decreased at Pa. The 437 cm 1 spectrum was the characteristic peak of crystalœ18, which was produced by the E2 vibration mode and corresponded to the strong XRD peak. This phenomenon showed that the growth of ZnO crystal along the c-axis direction had sharper peaks and better crystal quality. In addition, for the crystal of a wurtzite structure, the pattern has a very strong sensitivity to stress in crystal; the change of the peak position (437 cm 1 / or peak broadening can be used to detect tiny changes of stress in the crystal and when compression stresses are present in the crystal, the peak position to the high frequency moves. When there is tensile stress, the peak position to low frequency is mobile; so researching the change of the spectral peak is of great significance to the study of stress in crystal. In Fig. 4, the patterns of the ZnO crystal peak position (437 cm 1 / did not change with the change of the experimental conditions, which shows that the stress of the ZnO crystals has not changed. The cm 1 spectrum was from multi-phonon scattering peak. The cm 1 spectrum was the characteristic spectrum of II IV compound semiconductor double phonon modes. Fig. 5. The SEM micrographs of ZnO PVT growth under different system pressures. (a) 20 Pa. (b) 480 Pa. (c) 2800 Pa SEM analysis of ZnO crystal A comparative SEM analysis of the PVT-grown ZnO revealed a dramatic change in their morphological microstructure. The crystal grain sizes were similar to the results calculated by the XRD pattern. Figure 5(a) shows that there were ZnO crystals in the crucible lids at the 20 Pa system pressure and that the grain sizes were small and the number was sparse. With the system pressure increased, the grain size became larger and they were dense in number, because the system condition was close to the optimal growing point. The crystal obtained after the experimental runs consists of micro-size,

5 polyhedral particles with clearly defined facets. This specific growth habit is characteristic of the wurzite structure and has been observed during the PVT growth of ZnO crystal. Through the analysis, we found that the concentrations of ZnO crystals were very high (99.9%). When the system pressure was 2800 Pa, the growth rate slowed down and the quality of the crystal was bad, because the optimal growing point was broken by the high system pressure. The result was consistent with the analysis using XRD and the Raman spectrum. 4. Conclusion We demonstrated ZnO crystal boules of about 50 mm in diameter, which were polycrystalline, using a PVT growth technique and analyzed what was impacting the growth rate. We also analyzed the condition of the ZnO PVT growth under different ratios of O 2 and Ar gas, and different system pressures. We found the growth rates of ZnO crystal were better with the system pressure increased. When it exceeded the optimal growing point, the growth rates were reduced. However, the growth was still faster than the CVT method, and this method is easily controlled. The quality of PVT-growing ZnO crystal was studied by the XRD diffraction pattern, Raman spectrum and SEM micrographs. We discovered that the quality was very good and they all had a hexagonal wurtzite structure. Importantly, the purity of the ZnO crystal was very high. Our research results suggest that this novel PVT crystal growth technique is a viable production technique for producing ZnO crystals and further that it has laid the foundation for the research of ZnO single crystal. References [1] Look D C. Recent advances in ZnO materials and devices. Mater Sci Eng, 2001, B80: 383 [2] Molnar R J, Maki P, Aggarwal R, et al. Gallium nitride thick films grown by hydride vapor phase epitaxy. Mat Res Soc Sympo Proc, 1996, 423: 221 [3] Paszkiewicz R, Paszkiewicz B, Korbutowicz R, et al. MOVPE GaN grown on alternative substrates. J Cryst Res Technol, 2001, 36: 971 [4] Sakagami N. Hydrothermal growth and characterization of ZnO single crystals of high purity. J Cryst Growth, 1990, 99: 905 [5] Maeda K, Sato M, Niikura I, et al. Growth of 2 inch ZnO bulk single crystals by the hydrothermal method. Semicond Sci Technol, 2005, 20(4): S49 [6] Ohshima E, Ogino H, Niikura I, et al. Growth of the 2-in-size bulk ZnO single crystals by the hydrothermal method. J Cryst Growth, 2004, 260(1/2): 166 [7] Klimm D, Ganschow S, Schulz S, et al. The growth of ZnO crystals from the melt. J Cryst Growth, 2008, 310(12): 3009 [8] Grasza K, Mycielski A. Contactless CVT growth of ZnO crystals. Phys Status Solidi, 2005, 2(3): 1115 [9] Cantwell G, Zhang J, Song J J. Vapor transport growth of ZnO substrates and homoepitaxy of ZnO device layers. In: Litton C W, Reynolds D C, Collins T C, ed. Zinc oxide materials for electronic and optoelectronic device applications. John Wiley & Sons Ltd, 2011: 171 [10] Ntep J M, Barbe M, Cohen-Solal G, et al. ZnO growth by chemically assisted sublimation. J Cryst Growth, 1998, 184/185: 1206 [11] Li X, Xu J, Jin M, et al. Growth of ZnO single crystals by an induced nucleation from a high temperature solution of the ZnO- PbF 2 system. Cryst Res Technol, 2007, 42(3): 221 [12] Fujii T, Yoshii N, Masuda R, et al. Nucleation and coalescence behavior for epitaxial ZnO layers on ZnO/sapphire templates grown by halide vapor phase epitaxy. J Cryst Growth, 2009, 311(4): 1056 [13] Rojo J C, Liang S, Chen H, et al. Physical vapor transport crystal growth of ZnO. In: Teherani F H, Litton C W, ed. Zinc oxide material and device. Proc SPIE, 2006, 6122: 61220Q1 [14] Wang S, Kopec A, Timmerman A G. Growth and characterization of large-diameter, lithium-free ZnO single crystals. In: Teherani F H, Look D C, Rogers D J, ed. Proc SPIE, 2012, 8263: 82630E [15] Wang S. Method for production of zinc oxide single crystals. USA Patent, No. 2012/ [16] Wang S. Method and apparatus for zinc oxide single crystal boule growth. USA Patent, No [17] Zhao Youwen, Dong Zhiyuan. Growth of ZnO single crystal by chemical vapor transport method. Chinese Journal of Semiconductors, 2006, 27(2): 336 [18] Zhang Fan, Zhao Youwen, Dong Zhiyuan, et al. Bulk single crystal growth and properties of In-doped ZnO. Journal of Semiconductors, 2008, 29(8):