Elsevier

Journal of Crystal Growth

Volume 546, 15 September 2020, 125776
Journal of Crystal Growth

Studies on thermal and interface optimization for CdZnTe crystals by unseeded Traveling Heater Method

https://doi.org/10.1016/j.jcrysgro.2020.125776Get rights and content

Highlights

  • Two-dimensional simulation model of THM growth of CdZnTe was developed.

  • A stable thermal field was obtained by introducing a dummy crystal.

  • Convex interfaces were obtained by using low and different constant rotation rates.

Abstract

A stable thermal field and a controlled convex growth interface are essential for the stable growth of large-size CdZnTe crystals by the Traveling Heater Method (THM). However, the crucible wall temperature always increases due to the deposition of CdZnTe crystals with low thermal conductivity at the bottom of the crucible. A comprehensive two-dimensional simulation model for THM growth of CdZnTe is developed and implemented to study the heat and mass transfer. The effect of increasing wall temperature on interface shape and concentration distribution is discussed. Experimental results show that a dummy crystal with a thermal conductivity similar to that of the CdZnTe crystal helps to stabilize the wall temperature. In a stable thermal field, low constant rotation is assumed to be sufficient to weaken natural convection flow and to make the interface convex in THM. Simulation results show that low constant rotation rates of 2.5 RPM for 2-in. crystal and of 1.25 RPM for 3-in. crystal are effective in interface optimization. Several ingots with a diameter of 2 in. and 3 in. were grown by unseeded THM. Convex interfaces were observed both in 2-in. and 3-in. crystals by applying simulated rotation rates after fast cooling.

Introduction

CdZnTe is a potential material for 3rd generation nuclear detectors due to its advantages in material properties and wide applications in homeland security and medical imaging [1], [2], [3], [4]. To date, the mass production of CdZnTe single crystals lager than 20 mm × 20 mm × 15 mm with low defects and uniform composition is still a big challenge and suffers from a low yield rate. Compared to melt growth, the traveling heater method (THM), as a solution growth method, is a suitable solution for growing crystals with a high melting point or high vapor pressure and has its intrinsic advantages, such as low growth temperature, charge purification, and longitudinal homogeneity [5], [6], [7], [8].

THM is quite similar to zone melting, where pre-synthesized polycrystalline feed dissolves into solution through diffusion and convective transport and finally deposits as a single crystal at the growth interface. Due to buoyancy effects, a vortex with an upward flow near the ampoule wall and a downward flow along the center occupies the main area of the molten zone, while diffusion works only in thin boundary layers. Such convection, arising from a parabolic temperature profile, plays a significant role in enhancing mixing, but, on the other hand, is considered as the main cause of concave growth interfaces [8], [9]. It is well known that a slightly convex growth interface is essential for grain enlargement and preventing polycrystalline growth through post-formed grains on the ampoule wall. In order to obtain a convex interface, many researches start by weakening natural convection using a shorter molten zone, crucible rotation, an additional magnetic field, and microgravity growth [10], [11], [12], [13]. Crucible rotation is a simple and convenient technology in experiments. There are two main rotation regimes used in THM growth of CdZnTe, high rotation rate with time-varying fluctuation, namely the accelerated crucible rotation technique (ACRT), and low constant rotation. The former regime is used most commonly. The Ekman flow occurs during acceleration and deceleration time and causes strong mixing near the solid/fluid interfaces. Due to the stable thermal structure of cold below hot in the Vertical Bridgeman (VB) method, a high rotation rate with a high frequency is necessary to destroy the stable density stratification and facilitate mixing. For THM growth of GaAs, Lan et al. [14] first reported that a concave growth front due to the buoyancy convection can easily be inverted into a convex one by applying a constant rotation of 3–5 RPM when the diameter is 2.4 cm. Dost et al. [15] also found that a constant rotation of 5.0 RPM for a 1-in. CdTe crystal is effective in optimizing the interface. However, none of them has further experimental verifications. Besides, there are some potential dangers of using ACRT with a high rotation rate, including mechanical instabilities during ampoule rotation, transition from laminar to turbulent flow, and severe re-melting [14], [16].

For growing large grains, it is essential not only to create optimal crystal growth conditions, but also to ensure the stability of thermal and flow fields during crystal growth. So far, there has been little discussion on maintaining a stable thermal environment throughout the growth process. Martı́nez-Tomás et al. [17] considered the whole growth system, including the furnace, ampoule, and charge, to study the thermal conditions in THM growth of HgTe, and found that the heat flux from the furnace to the ampoule is not constant for the entire growth cycle.

In this study, a comprehensive two-dimensional simulation model of heat and mass transfer during THM growth of CdZnTe is developed. A well-selected dummy crystal is introduced to stabilize the thermal environment in the furnace. Two 1-in. THM experiments were carried out with/without a dummy crystal. The effect of increasing wall temperature on interface shape and concentration distribution is discussed. To obtain a convex interface, three ingots, one 2-in. and two 3-in. were grown using unseeded THM at low and different constant rotation rates and compared with simulation results.

Section snippets

Experiments

A schematic diagram of the THM growth system and a simplified model are shown in Fig. 1. The furnace is a self-designed single-zone electric furnace with a precision of ±0.1 °C. Before growth, polycrystalline CdZnTe feed was synthesized from 7 N metals using the traditional VB method with a composition of Cd:Zn = 9:1. Then, pure tellurium as solvent and pre-synthesized CdZnTe polycrystalline material were loaded into a carbon-coated quartz crucible and vacuum sealed at 5 × 10−4 Pa. The designed

Governing equations

Fig. 1b shows the simplified computational domain. For simplicity, it is assumed that the fluid is an incompressible viscous liquid, and the primary species of solution is CdTe with excess Te. The system is assumed axisymmetric, and the flow in the liquid is laminar. The Boussinesq approximation is also adopted. The solubility of the solvent in the solid and the solution buoyancy-driven convection were neglected, since a low growth temperature and a small concentration differences in the

Thermal optimization

Fig. 2 shows the temperature of the crucible wall measured by thermocouples (as shown in Fig. 1a) in a 1-in. THM experiment without a dummy crystal. It is evident that the crucible wall temperature gradually increases with the pulling process. For instance, the temperature difference between the upper and lower thermocouples reaches 37 °C in the same position. This is due to the deposition of low-thermal-conductivity CdZnTe at the bottom of the crucible, which weakens the heat removal from the

Conclusion

Both experimental and numerical simulation studies were carried out for CdZnTe crystal growth from a Te-rich solution by Traveling Heater Method. A two-dimensional simulation model for THM growth of CdZnTe is developed to study the heat and mass transfers, and double-interface tracking is achieved. The major conclusions are:

  • (1)

    The 1-in. experiments show that the temperature of the crucible wall increases with crystal growth due to the deposition of low-thermal-conductivity CdZnTe at the bottom of

CRediT authorship contribution statement

Bangzhao Hong: Conceptualization, Software, Data curation, Writing - original draft. Song Zhang: Methodology, Investigation, Supervision, Writing - review & editing. Lili Zheng: Software, Supervision. Hui Zhang: Methodology, Software. Cheng Wang: Validation, Resources. Bo Zhao: Validation, Resources.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work is supported by National Key R&D Program of China (2016YFB1102300), National Nature Science Foundation of China (Grant Nos. 91646201, U1633203) and Beijing Key Laboratory of City Integrated Emergency Response Science.

References (29)

Cited by (0)

View full text