当前位置：中文主页 > 我的新闻

杨建锋

同学科/专业博导同学科/专业硕导

Personal profile

个人简介

暂未填写

我的新闻

史永贵论文在Materials and Manufacturing Processes 上发表

发布时间：2011-05-04 点击次数：

发布时间：2011-05-04

文章标题：史永贵论文在Materials and Manufacturing Processes 上发表

内容：

Effects of Grain Size of Source Material on Growing 6 H-SiC Bulk Crystal by Physical Vapor Transport

immediate access to the latest key research articles

Authors: Yonggui Shi ^a; Peiyun Dai ^a; Jianfeng Yang ^a; Zhihao Jin ^a; Jikuan Cheng ^a; Hulin Liu ^a

Affiliation:

^a State Key Laboratory for Mechanical Behavior of Materials, Xi'an Jiaotong University, Xi'an, Shaanxi, China

DOI: 10.1080/10426914.2010.544829

Article Requests: Order Reprints : Request Permissions

Published in: journal

Materials and Manufacturing Processes

Publication Frequency: 12 issues per year

Accepted uncorrected manuscript posted online: 03 May 2011

Previously published as: Advanced Manufacturing Processes (0884-2558) until 1987

Previously published as: Advanced Materials & Manufacturing Processes (0898-2090) until 1989

View Related Articles

To cite this Article: Shi, Yonggui , Dai, Peiyun , Yang, Jianfeng , Jin, Zhihao , Cheng, Jikuan and Liu, Hulin (2011) 'Effects of Grain Size of Source Material on Growing 6 H-SiC Bulk Crystal by Physical Vapor Transport', Materials and Manufacturing Processes, 1, doi: 10.1080/10426914.2010.544829, First posted on: 03 May 2011 (iFirst)

Disclaimer: This is a full text version of an unedited manuscript that has been accepted for publication. As a service to authors and researchers we are providing this version of the accepted manuscript (AM). Copyediting, typesetting, and review of the resulting proof will be undertaken on this manuscript before final publication of the Version of Record (VoR). During production and pre-press, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal relate to this version also.

Abstract

In this paper, effects of grain size of source material on growing 6H-SiC bulk single crystal by physical vapor transport (PVT) have been studied by observing the experimental results using source materials of different grain sizes through optical microscope and calculative analysis and discussions. The results indicate that source materials with different grain sizes affect the growth process of SiC bulk single crystal by PVT mainly from three perspectives, i.e., the effective heat-transfer coefficient of the source material, the supersaturation and the ratio of Si/C in the growth crucible on the basis of other parameters. Furthermore, a proposed way to improve the quality of 6H-SiC bulk single crystal is optimizing the grain size of the source material, and the optimum grain size for sublimation growth of 6H-SiC bulk crystal in our lab is 120 µm.

Keywords: 6H-SiC; Bulk crystal; Grain size; Physical vapor transport; Single crystal; Source material

INTRODUCTION

Silicon carbide (SiC), as an important wide-gap semiconductor material, is suitable for fabricating electronic devices operating under high temperature, high frequency, and intensive-radiation conditions for its excellent physical and chemical properties such as wide-band gap, high thermal conductivity, high electric field breakdown strength, etc. [1, 2]. Up to now, the most successful technique of growing SiC bulk single crystal has been the physical vapor transport method (PVT) [3, 4] and some progress has been made in the diameter and the quality of SiC crystal [5].

The growth of SiC bulk single crystal by PVT is a complex process where a number of parameters have to be controlled carefully, including temperature of substrate, temperature gradient, gas phase composition and so on [6, 7]. Basing on numerical modeling and experimental results [8], the source material of SiC has great impact on the global heat transfer and gas phase composition. Although Wellmann et al.[9] have studied the impact of morphological change of the SiC source material on the growth of SiC by PVT using the X-ray imaging technique, the mechanism of source material affecting the sublimation growth process is still needed probing.

In this paper, the effects of the gain size of source material on growing 6H-SiC crystal by PVT have been studied on the basis of experimental comparison and theoretical analysis. The mechanism of different grain sizes of source materials affecting the growth process was analyzed systemically by investigating the influence of the grain sizes of source materials on the heat transfer coefficient, the supersaturation of vapor species and the Si/C ratio in the growth chamber.

EXPERIMENTAL PROCEDURE

The experiment was performed in an inductive heated graphite crucible at elevated temperature above 2300°C as presented in Fig. 1. The temperature at the top/bottom of the crucible was measured by optical pyrometers. Graphite felts served as thermal-insulation layers, and the powder of industrial 6H-SiC as source material. Three types of source materials with grain sizes of 65 µm, 120 µm and 165 µm, labeled as A, B, and C respectively, were used. The (0001)Si-face of 6H-SiC platelets prepared by Lely-method served as growth surface of seed crystal, which was fixed on graphite lid and SiC powder was packed into a graphite crucible with a distance of approximate 2 cm. FIGURE 1. —Schematic drawing of the growth furnace.

Under an argon pressure of 30 KPa, the graphite crucible was rapidly heated until the temperature of its bottom got up to 2300°C. Then vacuum pump started to bleed the gas slowly to form a gas flow between the source material region and the seed crystal, which could hinder the seed crystal from being carbonized and remove the impurity and contamination on its surface. The vacuum pump was turned off until the pressure in the furnace having reached 15 KPa. This process lasted about 10 minutes. Afterwards, the heating power was maintained the same to ensure the temperature of the bottom of crucible at 2300°C. The temperature at the top of the crucible was measured about 2200°C, and the mean temperature gradient about 10-15°C/cm. When the growth process was finished, the heating power was stopped and an argon pressure of approximately 1 atm was introduced to the furnace to prevent the grown crystal from decomposition. Optical microscopy and SEM were applied to investigate the grown crystals. The residues of source materials were also investigated.

RESULTS AND DISCUSSION

Figure 2 shows the crystals grown by the three types of source materials with different grain sizes. Crystal grown with source B is of the best-quality with smooth and bright surface as well as obvious growth habit plane, while the crystal grown with source C has many notches lacking of luster on the surface and is of grey black color, which indicated that it has been carbonized to some degree. The rough and bright surface of crystal grown with Source A implies that its growth ambient is richer in Si than the other two. And the water-wave-like pits on its surface indicate that the vapor-liquid-solid (V-L-S) crystal growth mechanism has happened. FIGURE 2. —6H-SiC crystals grown by PVT with the three types of source materials at 2300°C for 25 h: A) 65 µm, B) 120 µm and C) 165 µm.

The above grown crystals' growth parameters are the same except for their different grain sizes of the source materials, therefore, grain size has a significant influence on the growth process of the 6H-SiC crystals by PVT.

Firstly, grain size affects the effective heat-transfer coefficient of source materials. The effective heat-transfer coefficient of granular material has a relation with the thermal conductivities of gas in the pores and of the solid, the porosity of source material and temperature. At an elevated temperature, the particle-to-particle radiation of granular material will increase with the temperature increasing, especially when the grain size of source material is smaller than 100 µm, so the effective heat-transfer coefficient should contain the effective radiation coefficient.

According to Schotte's technique [10] for predicating the effective heat-transfer coefficient of packed beds where the gas phase is continuous, the effective heat-transfer coefficient, K_eff s, of the three types of source materials at the initial growth stage were calculated, as listed in Table 1.

**TABLE 1.—The calculated values involved at 2300°C with atmospheric pressure of 15KPa.**
Grain size µm	Porosity ω	K_g W/m.K	K_b W/m.K	K_r W/m.K	K_eff W/m.K
Here K_g refers to the modified gas thermal conductivity, K_b the effective thermal conductivity of the packed source materials, K_r the effective radiation conductivity of the packed raw materials, K_eff the total effective heat-transport coefficient, K_eff = K_b + K_r.
65	0.560	0.042	0.333	0.517	0.850
120	0.546	0.049	0.376	0.563	0.939
165	0.524	0.051	0.385	0.606	0.991

The calculated effective heat-transfer coefficients in Table 1 have a tendency that the effective heat-transfer coefficients decrease with the grain sizes of the source materials decreasing under subatmospheric pressure. This tendency is similar to that calculated using the model developed by Kitanin in Ref. [11], which implies that the calculated K_effs are reasonable to a certain extent.

According to Fourier's law, when temperature of the crucible bottom is fixed at 2300°C, the smaller effective heat-transfer coefficient of the source material will result in the lower temperature at the centre of upper surface of the source material. Figure 3b presents the temperature distributions along the symmetric axis of crucible on the basis of the theoretical estimation. At the initial stage of crystal growth process, the relationship of the source-growth front temperature differences ΔT_i = 2300-T_i'(I = 1, 2 and 3) is demonstrated as ΔT₁>ΔT₂>ΔT₃, and the relationship of the mean growth temperatures T_i have a relationship presented as T₁ < T₂ < T₃. FIGURE 3. —Scheme of the growth crucible system (a) and the temperature distribution along the axis of symmetry in the source center (b).

Secondly, the grain size of the source material has an impact on the supersaturation of the growth process of SiC bulk single crystal by PVT. The supersaturation σ is a function of the mean temperature T in the growth crucible and the source-growth front temperature difference ΔT [12], where ΔH presents enthalpy of sublimation and R the gas constant.

Equation (1) demonstrates that the mean temperature and the source-front growth temperature difference, on which the effective heat-transfer coefficient of source material have impacts, are two important factors determining the supersaturation in the growth of SiC bulk crystal. Combined the temperature distributions presented in Fig. 3b and Eq. (1), the supersaturations in the three growth crucibles have a relationship of σ_A > σ_B > σ_C.

In principle, the supersaturation being too high or too low is not favorable for the growth of high-quality crystals. If the supersaturation of Si vapor specie in the growth chamber is too high, the growth mechanism may transfer from the V-S to the V-L-S, while carbonization will happen to the growing crystal with the supersaturation being too low.

The experimental results in Fig. 2 also indicate that the growth mechanism of Sample A transferred from V-S to V-L-S, while that of Sample B is a typical V-S growth mechanism. And Sample C has been carbonized during the growth process. Their surface morphologies indicate that the supersaturations in the three growth crucibles have a relationship of σ_A > σ_B > σ_C, which is consistent with the calculated results.

Thirdly, the Si/C ratio is also affected by the grain size of source material in the growth process of SiC bulk single crystal. According to Tairov [13], the Si/C ratio increases quickly with the decreasing of the grain size of the source material while decreases with the temperature increasing. In sublimation growth of SiC, the C component determines the growth rate. With a suitable surpersaturation, a relative little Si/C ratio favors high growth rate.

To estimate the effect of the grain size on the Si/C ratio, the sublimation rates of the three types of source materials and recrystallization rates resulting from vapor species recrystallization on the graphite lids in the growth durations were measured, as shown in Fig. 4, which implies that with the grain size decreasing, both the sublimation rate and recrystallization rate increase. However, the recystallization rates and the sublimation rates were not in accordance with each other because the recrystallization rates were far lower (lower more than 3 times) than the sublimation rates, whose decreasing rate were in turn much higher (higher more than 6 times) than that of the recrystallization rates. The inconsistency of the recrystallization rates and the sublimation rates reflected the non-stoichiometry sublimation character of SiC source materials, which has great influence on the Si/C ratio in the vapor phase. FIGURE 4. —Sublimation rates (hollow symbols) and recrystallization rates (solid symbols) of the three types of source materials under argon pressure of 15KPa at 2300°C.

In the growth process of SiC bulk single crystal by PVT, the main vapor species for the SiC+C system at elevated temperature are Si, Si₂C and SiC₂. Their equilibrium partial pressures of vapor species are functions of temperature for the SiC+C system as following [14]: Here P_Si, P_Si2C and P_SiC2 denote the equilibrium vapor pressures of Si, Si₂C and SiC₂, respectively.

According to the above three equations, the calculated equilibrium partial pressures of Si, Si₂C and SiC₂ are 252.3 Pa, 64.3 Pa and 182.2 Pa at 2300°C, respectively. It is evident that the partial pressure of Si is much higher than that of the other two and the total amount of component Si would be much larger than the total amount of component C in the vapor phase, resulting in the Si/C ratio in the vapor phase far larger than that in the solid phase (approximately equal to 1). Therefore, in the sublimation process of SiC source material, there would be much residue of solid C powder on the surfaces of SiC source materials [6], and the residue C would hinder further sublimation of source material greatly, which might be the reason of the sublimation rates of source materials decreasing with the sublimation time lasting.

In addition, the relationship between the equilibrium vapor pressure and grain size could be expressed as where P refers to equilibrium vapor pressure of source material with a diameter of d, P₀ the bulk materials' equilibrium vapor pressure, M materials' mole mass and γ surface energy and R the gas constant.

According to Eq. (5), with the grain size of source material decreasing, the equilibrium vapor pressure would increases. Correspondingly, the difference in the partial vapor pressure among the vapor species would be increase exponentially, which not only increases the absolute amount of C component but also increase the Si/C ratio. It may be responsible for the larger decreasing ratio of sublimation rates and the higher recrystallization rates for smaller grain size of source material.

CONCLUSION

The effects of grain size of source material in the growth process of SiC bulk single crystal by PVT have been discussed by analyzing and discussing the related experimental and calculated results. The discussion results indicated that the grain size of source material could affect the effective heat-transfer coefficient of source material, the superasturation and the ratio of Si/C in the growth crucible in the growth process of SiC bulk single crystal by PVT. Therefore, optimizing the grain size of source material is an effective and convenient way to grow high-quality SiC bulk single crystal with high growth rate and less carbonization by PVT. And the optimum grain size of source material for sublimation growth of SiC crystal is 120 µm in our lab.

REFERENCES

1. Takahashi, K. , Yoshikawa, A. and Sandhu, A. (2007) Wide Bandgap Semiconductors—Fundamental Properties and Modern Photonic and Electronic Devices p. 231. Springer , Berlin
2. Wang, X. L. , Cai, D. and Zhang, H. (2007) Increase of SiC Sublimation Growth Rate by Optimizing of Powder Packaging. J. Crystal Growth 305:1 , pp. 122-132.
3. Liu, X. , Shi, E. -W. , Song, L. -X. and Zhi-Zhan, C. (2008) Effects of Graphitization of the Crucible on Silicon Carbide Crystal Growth. J. Crystal Growth 310:19 , pp. 4314-4318.
4. Gao, Y. , Hu, X. Chen, X. et al. (2010) Evolution and Structure of Low-angle Grain Boundaries in 6H-SiC Single Crystals Grown by Sublimation Method. J. Crystal Growth 312:20 , pp. 2909-2913.
5. Berkman, E. , Leonard, R. T. , Paisley, M. J. Khlebnikov, Y. et al. (2009) SiC Epitaxial Growth on Multiple 100-mm Wafers and its Application to Power-Switching Devices. Mater. Sci. Forum 600-603:3-6 , pp. 77-82.
6. Sudarshan, T. S. and Maximenko, S. I. (2006) Bulk Growth of Single Crystal Silicon Carbide. Microelectronic Engineering 83:1 , pp. 155-159.
7. Gao, B. , Chen, X. J. , Nakano, S. , Nishizawa, S. and Kakimoto, K. (2010) Analysis of SiC Crystal Sublimation Growth by Fully Coupled Compressible Multi-phase Flow Simulation. J. Crystal Growth 312:22 , pp. 3349-3355.
8. Chen, X. , Nishizawa, S. -I. and Kakimoto, K. (2010) Numerical Simulation of a New SiC Growth System by the Dual-directional Sublimation Method. J. Crystal Growth 312:10 , pp. 1697-1702.
9. Wellmann, P. J. , Hofmann, D. , Kadinski, L. , Selder, M. , Straubinger, T. L. and Winnacker, A. (2001) Impact of Source Material on Silicon Carbide Vapor Transport Growth Process. J. Crystal Growth 225:2/4 , pp. 12-316.
10. Geiger, G. H. and Poirier, D. R. (1973) Transport Phenomena in Metallurgy p. 204. Addison-Westley Publising Commpany press , New Jersey, USA
11. Kitanin, E. L. , Ramm, M. S. , Ris, V. V. and Schmidt, A. A. (1998) Heat Transfer Through Source Powder in Sublimation Growth of SiC Crystal. J. Materials Science and Engineering B 55:3 , pp. 174-183.
12. Sudarshan, T. S. (2004) SiC Power Materials, Devices and Applications pp. 1-61. Springer Series in Material Science
13. Tairov, Yu.M. (1995) Growth of Bulk SiC. J. Materials Science and Engineering B 29:1-3 , pp. 83-89.
14. Lilov, S. K. (1995) Thermodynamic Analysis of Phase Transformations at the Dissociative Evaporation of Silicon Carbide Polytypes. J. Diamond and Related Materials 4:12 , pp. 1331-1334.

List of Figures

FIGURE 1. —Schematic drawing of the growth furnace.

FIGURE 2. —6H-SiC crystals grown by PVT with the three types of source materials at 2300°C for 25 h: A) 65 µm, B) 120 µm and C) 165 µm.

FIGURE 3. —Scheme of the growth crucible system (a) and the temperature distribution along the axis of symmetry in the source center (b).

FIGURE 4. —Sublimation rates (hollow symbols) and recrystallization rates (solid symbols) of the three types of source materials under argon pressure of 15KPa at 2300°C.

List of Tables

**TABLE 1.—The calculated values involved at 2300°C with atmospheric pressure of 15KPa.**
Grain size µm	Porosity ω	K_g W/m.K	K_b W/m.K	K_r W/m.K	K_eff W/m.K
Here K_g refers to the modified gas thermal conductivity, K_b the effective thermal conductivity of the packed source materials, K_r the effective radiation conductivity of the packed raw materials, K_eff the total effective heat-transport coefficient, K_eff = K_b + K_r.