- Email: [email protected]
Controlling nucleation during unseeded THM growth of CdZnTe crystal
Journal Pre-proofs Controlling nucleation during unseeded THM growth of CdZnTe crystal Bangzhao Hong, Song Zhang, Lili Zheng, Hui Zhang, Cheng Wang, Bo Zhao PII: DOI: Reference:
S0022-0248(20)30005-1 https://doi.org/10.1016/j.jcrysgro.2020.125482 CRYS 125482
To appear in:
Journal of Crystal Growth
Received Date: Revised Date: Accepted Date:
11 October 2019 11 December 2019 4 January 2020
Please cite this article as: B. Hong, S. Zhang, L. Zheng, H. Zhang, C. Wang, B. Zhao, Controlling nucleation during unseeded THM growth of CdZnTe crystal, Journal of Crystal Growth (2020), doi: https://doi.org/10.1016/ j.jcrysgro.2020.125482
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier B.V.
Controlling nucleation during unseeded THM growth of CdZnTe crystal Bangzhao Hong1, Song Zhang2*, Lili Zheng3, Hui Zhang1 Cheng Wang4, Bo Zhao4 1Department
of Engineering Physics, Tsinghua University, Beijing, China for Aero Engine, Tsinghua University, Beijing, China 3School of Aerospace Engineering, Tsinghua University, Beijing, China 4Ruiyan Technology Co. Ltd., Hangzhou, Zhejiang, China 2Institute
There are always undesired grains formed at the beginning of unseeded THM growth which deteriorate the crystal quality greatly. In this paper, several ingots with the diameter of 27mm were grown by unseeded THM to investigate the nucleation process. The growth temperature in the experiments was 750°C and the growth rate was 5mm/day. The influence of crucible bottom geometry and temperature gradient on nucleation are studied. Numerical simulations of thermal and concentration fields were also performed for the THM process to study the growth process. The simulation results show that square cavity flow's structure and parabolic wall temperature profile are the main reason for secondary nucleation in the radial direction. Dummy crystal with cold finger was introduced in order to stabilize the thermal field and modify the lateral temperature gradient. Subsequent experimental results indicate that a large longitudinal temperature gradient with small lateral temperature gradient is more favor to suppress secondary nucleation and maintain stable growth. Keywords: B2. CdZnTe; A2. Traveling heater method; A1. Nucleation; A2. Single crystal growth; A1. Mass transfer, heat transfer; A1. Computer simulation
1. Introduction CdxZn1-xTe, commonly known as CZT, has been recognized as the best material for room temperature nuclear radiation detector due to its excellent optoelectronic properties, such as suitable band gap, high average atomic number, high mobility and high value of μτ product electrons [1-4]. Despite decades of research on crystal growth of CZT, some of the difficulties in growth, including Te inclusions/precipitates, crystal size, compositional segregation and low growth rate, have limited the yield of CZT. In recent years, benefiting from lower growth temperature and better compositional homogeneity, Traveling Heater Method (THM) as one of solution methods, is becoming the most practical approach to grow high-quality CZT crystals [5, 6]. In order to obtain large-volume single crystal, the number of grains in ingot must be limited. For unseeded THM growth, controlling the nucleation position and reducing nucleation number at the initial stage are required for improving crystal size. So far, there has been limited work on nucleation control during unseeded THM growth of CZT. Roy et al. [7, 8] studied Te inclusions/precipitates and optimization of growth interface using unseeded THM. They used conical crucible to control nucleation, however, there was not much description on nucleation control. Surface energy plays a critical role in heterogeneous nucleation. Since good adhesion between molten CdTe and uncoated crucible, the nucleation in uncoated crucible is expected to occur with small undercooling and inhibit the formation of new grains after the first nucleation [9]. However the sticking between crystal and quartz often leads to movement and multiplication of dislocations even crystal cracking during cooling. Considering the relation between supercooling and nucleation, accurate prediction of supercooling region is also of great help for nucleation control [10, 11]. In experiment, controlling supercooling position can be achieved by enhancing heat release at the center of crucible bottom. Scheel[12] proposed an experimental scheme for controlling nucleation by combination of accelerated crucible rotation technique (ACRT) with localized cooling, and succeed in solution growth of GdAlO3. Such process can be used in CZT growth. In general, controlling the formation of supercooling region and undercooling during nucleation is the main part of nucleation control in unseeded growth. In the present work, the effects of crucible bottom geometry and temperature gradient including longitudinal and lateral ones on nucleation are studied. Both experimental and simulation methods are performed to understand and optimize the nucleation process.
2. Experimental Fig. 1 shows the schematic diagram of THM growth system. This furnace is homemade electric furnace. Only one heater controlled by PID controller is used to provide heat for crystal growth. The movement of crucible is achieved by lowering the crucible support which is driven by high precision stepping motor with repositioning precision of 2 microns. There are four thermocouples set at several different places of crucible wall to monitor the growth condition, and the distance between the upper and lower thermocouples is the designed length of molten zone. A dummy crystal with cold finger was introduced at the bottom of crucible. The main purpose of introducing dummy crystal with same thermal prosperities as CZT is to stabilize the thermal field during crystal growth [13]. All the polycrystalline CZT ingots were synthesized from 7 N metals by using the traditional Vertical Bridgman method. Prior to loading the charge, the quartz crucibles were carefully cleaned by ultrasonic cleaner, then immersed in nitric acid and rinsed repeatedly with ultrapure water, and finally coated with graphite layer by cracking high purity acetone. The excessive tellurium and polycrystalline CZT, used as solvent and feed respectively, were loaded into carbon-coated quartz crucible and sealed under dynamic vacuum of 5 × 10−4 Pa. At the initial state of growth, a relative low pull-down velocity of 3 mm/day was used to control nucleation, and then increased to 5 mm/day for crystal growth. There are six crystal growth experiments conducted in order to study the effects of different growth conditions on nucleation. The growth parameters used are listed in Table. 1. CZT-1 and CZT-2 are set to study the influence of conical crucible on nucleation. Double shoulders design with different angle is used. The first sharp shoulder is expected to provide a large undercooling within limited space and thus cold finger is not used in CZT-1 and CZT-2. The second shoulder with large cone angle is designed for grain enlargement. CZT-3 to CZT-6 focus on optimization of nucleation with different combination of longitudinal and lateral temperature gradients. The longitudinal temperature gradient refers to the wall temperature gradient at the bottom of crucible, and the lateral temperature gradient is the temperature difference between the center and wall of crucible bottom. The modification of longitudinal and lateral temperature gradients in the experiments are realized by changing the thickness of insulation at bottom and the material of cold finger respectively.
Fig. 1. Schematic of one-zone THM furnace and its temperature profile.
Table 1 The growth process parameters used in experiments. Longitudinal
Lateral
Bottom
temperature
temperature
Shape of
gradient (°C/cm)
gradient (°C/cm)
Crucible
85
35.0
/
See Fig. 2(a)
3~5
70
35.0
/
See Fig. 2(b)
CZT-3
3~5
45
42.1
26.7
CZT-4
3~5
45
75.2
26.7
CZT-5
3~5
45
75.2
44.2
CZT-6
3~5
45
75.2
18.3
Growth rate
Length of crystal
(mm/day)
(mm)
CZT-1
3~5
CZT-2
Case#
Fig. 2. The different geometric shape of crucible bottom.
See Fig. 2(c)
3. Results and discussion Fig .1 shows the startup status of unseeded THM growth of CZT while no grain forms on the crucible bottom. With the down-movement of crucible, the solution at the crucible bottom begins to be supercooled and nucleates after reaching the critical undercooling. In this paper, the formation and development of supercooling region are calculated to predict the nucleation. For the sake of simplifying the simulation, it is assumed that there is no grains formed on the crucible bottom as the crucible moves down. Besides, considering the computational cost in solving the unsteady equations, the dynamic growth process is approximated by translating the temperature profile on crucible surface. The complete description of model and material properties used in simulation are presented in our previous work[13] while the “single crystal” in original computational domain is removed and no-mass flux condition is set on crucible bottom.
3.1. Geometry Fig. 3 shows the as-grown ingots of 27mm diam grown in conical crucible. Those two ingots were cut perpendicular to the growth direction with a thickness of 5mm after finishing the shouldering process. It can be seen that the shoulder design plays a very important role in crystal selection. Only one grain is obtained both in CZT-1 and CZT-2 after finishing the first shouldering process. Unfortunately, the initial selected grain disappears soon, and new grains are formed at second shouldering process and show better enlargement tendency. Fig. 4 shows the evolution and development of supercooling regions in the melt with different cone angle as the crucible moves down. The wall temperature profile of crucible used in simulation is collected from experimental data of CZT-1 while the longitudinal temperature gradient is 35.0°C/cm. Simulation results reveal that the shoulder design ensures that supercooling region is preferentially formed in the center (see Fig. 4(b) and Fig. 4(c)). However, as the crucible continues to move down, a new supercooling region is formed in the radial direction, which is attributed to the square cavity flow's structure and parabolic wall temperature profile applied in THM. In the absence of flow, a convex isotherm is formed at the crucible bottom due to the parabolic wall temperature profile. However, buoyancy-driven natural convection transports heat from the wall to interior and heats the interior melt, which makes the isotherm concave near the center and forms a hump-shaped isotherm at the crucible bottom. As seen in Fig. 4(a), when the cone angle of crucible is very large, the supercooling region firstly appears in the radial direction of crucible instead of in the center. As is well known, the melt undercooling in solution growth is determined by local solute concentration and local temperature. As shown in Fig. 5, it is clear that the concentration alone the bottom of crucible is relative uniform with small increase from the center to the wall. There are two minimum points of temperature in the radial direction. The low temperature at the center is a result of conical crucible while the formation of the other one located at about 1/2R, where it is exactly the location of post-formed supercooling region, is caused by hump-shaped isotherm as analyzed above. Compared with the position of new grains formed at the second shoulder in the experiments, this hump-shaped isotherm distribution caused by the coupling of square cavity flow's structure and parabolic wall temperature profile is responsible for the secondary nucleation in the radial direction. Despite wide application of conical crucible in crystal growth, there is an unneglectable disadvantage of irregular crucible that is the changed length of molten zone during the shouldering
process. Since THM is zone melting method, the changed length of molten zone will affect the convection intensity of melt and result in redistribution of concentration and temperature fields.
Fig. 3. Longitudinal and lateral cross section of as grown crystals in different conical crucible, (a) from the bottom and (d) form the top.
Fig. 4. The evolution and development of supercooling regions in the melt of 1-inch crucible with different cone angle of (a) 165°, (b) 150°and (c) 135°. The right side of each case is streamline and the red area on the left represents the supercooling region. The numbers below each picture is the descending distance of crucible, i.e., +0mm represents the initial start position and +1mm is the distance that the crucible moves down from the initial start position.
Fig. 5. The temperature and concentration distribution along the radial direction of crucible bottom of third case in Fig. 4(b).
3.2 Temperature gradient Four as-grown CZT ingots under different combination of longitudinal and lateral temperature gradients are shown in Fig. 6. All the ingots were cut perpendicular to the growth direction with a thickness of 5mm. The number of grains of four ingots in the “first” wafer are listed in Table. 2. Due to crystal cracking, the wafer closest to the bottom and capable of counting the number of grains is used to represent the nucleation of as-grown ingot. In comparisons between CZT-3 and CZT-4, when the longitudinal temperature gradient rises from 42.1°C/cm to 75.2°C/cm with the same lateral temperature gradient of 26.7°C/cm, the number of initial grains decreases from 34 to 23. More importantly, the grains of CZT-4 exhibit typical columnar growth pattern after finishing nucleation (see in Fig. 6(b)). Increasing the longitudinal temperature gradient of growth interface shows a positive effect on nucleation. However, the large longitudinal temperature gradient used in subsequent experiments is mainly for maintaining the stable growth of crystal. According to the classical theory of constitutional supercooling, a large longitudinal temperature gradient is believed to suppress constitutional supercooling and allow increased growth rate. Moreover, a large longitudinal temperature gradient is easier to be realized and minimizes the effects of temperature fluctuation on crystal growth. Crystal growth experiments under different lateral temperature gradients by changing the material of cold finger were conducted while the longitudinal temperature gradient remains at 75.2°C/cm. From CZT-5, CZT-4 and CZT-6, it is obvious that the number of initial grains decreases rapidly from 55 to 14 when the lateral temperature gradient decreases from 44.2°C/cm to 18.3°C/cm. Besides that, it is worth mentioning that all grains of CZT-4, CZT-5 and CZT-6 are enlarging in subsequent crystal growth, and CZT-4 and CZT-6 show the obvious columnar growth pattern. In order to understand nucleation under different lateral temperature gradients, a series of simulation experiments were carried out. In this part, the wall temperature profile of crucible was approximated by Gauss distribution with the longitudinal temperature gradient of 70°C/cm, and three groups of lateral temperature gradients, 10°C/cm, 30°C/cm and 50°C/cm, were set for comparison. Since a slow speed of 3mm/day was used in the experiments, it is assumed that the quasi-steady state is reached. Then the evolution of supercooling region with time is obtained by solving the flow field at different positions in the furnace. As seen in Fig. 7, when the lateral temperature gradient decreases, the supercooling region gradually moves outward, which is also the reason for the radial nucleation in the conical crucible. The main reason for the fluctuation of undercooling under large lateral temperature gradient of 50°C/cm shown in Fig. 7 is the redistribution of concentration and temperature fields caused by dynamic change of flow structure. Fig. 8 shows the width of “nucleation” area at the crucible bottom under assumed nucleation undercooling of 10°C. Although the exact value of nucleation undercooling is unknown, it does not change the trend of curve in Fig. 8. It is clear from the curve that when the lateral temperature gradient is 50°C/cm and 30°C/cm, the width of the “nucleation” area shows a rapid increase at the initial stage, and then slowly increases to the maximum. Meanwhile the change is linear with smaller slope under lateral temperature gradient of 10°C/cm. Since the supercooling point under large lateral temperature gradient firstly appears in the center, a large supercooling area will be formed at the center in a short time, which may result in large-scale random nucleation. On the contrary, a ring-shape supercooling region with smaller linear increase in size is more suitable for the formation and growth of single grain. Because the supercooling point lies in the radial
direction of the crucible under small lateral temperature gradient, it is hard to form a single grain on crucible bottom.
Fig. 6. The as-grown CZT wafers under the different combination of longitudinal and lateral temperature gradients, (a) from the bottom and following letters closer to top.
Table 2 Nucleation number under the different combination of longitudinal and lateral temperature gradients. Growth rate
Longitudinal temperature
Lateral temperature
(mm/day)
gradient (°C/cm)
gradient (°C/cm)
CZT-3
3
42.1
26.7
34
CZT-4
3
75.2
26.7
~23
CZT-5
3
75.2
44.2
~55
CZT-6
3
75.2
18.3
14
Case#
Number of grains
Fig. 7. Development of undercooling of crucible bottom with time while the longitudinal temperature gradient is 70°C/cm.
Fig. 8. Width of “nucleation” area at the crucible bottom with assumed nucleation undercooling of 10°C.
4. Conclusions In this work, nucleation during unseeded THM growth of CZT is studied. Both experiment and simulation are adopted to study the influence of crucible bottom geometry and temperature gradient including longitudinal and lateral temperature gradients on nucleation. The experiment results show that the conical crucible is helpful for initial crystal selection and only one grain was obtained both in conical crucibles after finishing the first shouldering process. However, secondary nucleation inevitably occurs in the radial direction during the second shouldering process and the selected grain disappears soon. Simulation results reveal that the square cavity flow's structure and parabolic wall temperature profile are the main reason for the secondary nucleation. Subsequent experiments in flat-bottom crucible show that the large longitudinal temperature gradient and small radial temperature gradient are better for growth and nucleation. Under the combination of large longitudinal temperature gradient of 75.2°C/cm and small lateral temperature gradient of 18.3°C/cm, the nucleation number decreases obviously from 55 to 14, more importantly, the typical columnar pattern is observed in subsequent crystal growth.
Acknowledgements This work is supported by National Key R&D Program of China (2016YFB1102300), National Science Foundation of China (Grant Nos. 91646201, U1633203) and Beijing Key Laboratory of City Integrated Emergency Response Science.
Reference [1] H. Chen, S.A. Awadalla, K. Iniewski, P.H. Lu, F. Harris, J. Mackenzie, T. Hasanen, W. Chen, R. Redden, G. Bindley, I. Kuvvetli, C. Budtz-Jorgensen, P. Luke, M. Amman, J.S. Lee, A.E. Bolotnikov, G.S. Camarda, Y. Cui, A. Hossain, R.B. James, Characterization of large cadmium zinc telluride crystals grown by traveling heater method, J Appl Phys, 103 (2008). [2] C. Szeles, Advances in the crystal growth and device fabrication technology of CdZnTe room temperature radiation detectors, Ieee T Nucl Sci, 51 (2004) 1242-1249. [3] T.E. Schlesinger, J.E. Toney, H. Yoon, E.Y. Lee, B.A. Brunett, L. Franks, R.B. James, Cadmium zinc telluride and its use as a nuclear radiation detector material, Materials Science and Engineering: R: Reports, 32 (2001) 103-189. [4] M. Schieber, R.B. James, H. Hermon, A. Vilensky, I. Baydjanov, M. Goorsky, T. Lam, E. Meerson, H.W. Yao, J. Erickson, E. Cross, A. Burger, J.O. Ndap, G. Wright, M. Fiederle, Comparison of cadmium zinc telluride crystals grown by horizontal and vertical Bridgman and from the vapor phase, J Cryst Growth, 231 (2001) 235-241. [5] H. Chen, S.A. Awadalla, R. Redden, G. Bindley, A.E. Bolotnikov, C.G. Carini, R.B. James, High-Performance, Large-Volume THM CdZnTe Detectors for Medical Imaging and Homeland Security Applications, Ieee Nucl Sci Conf R, (2006) 3629-3637. [6] H. Shiraki, M. Funaki, Y. Ando, A. Tachibana, S. Kominami, R. Ohno, THM Growth and Characterization of 100 mm Diameter CdTe Single Crystals, Ieee T Nucl Sci, 56 (2009) 1717-1723. [7] U.N. Roy, A. Gueorguiev, S. Weiller, J. Stein, Growth of spectroscopic grade Cd0.9Zn0.1Te:In by THM technique, J Cryst Growth, 312 (2009) 33-36. [8] U.N. Roy, S. Weiler, J. Stein, Growth and interface study of 2 in diameter CdZnTe by THM
technique, J Cryst Growth, 312 (2010) 2840-2845. [9] R. Shetty, W.R. Wilcox, L.L. Regel, Influence of Ampoule Coatings on Cadmium Telluride Solidification, J Cryst Growth, 153 (1995) 103-109. [10] J.H. Peterson, M. Fiederle, J.J. Derby, Analysis of the traveling heater method for the growth of cadmium telluride, J Cryst Growth, 454 (2016) 45-58. [11] C.W. Lan, D.T. Yang, A computer simulation of crystal growth by the traveling-solvent method (TSM): pseudo-steady-state calculations, Model Simul Mater Sc, 3 (1995) 71-92. [12]
H.J.
Scheel,
Accelerated
Crucible
Rotation
-
Novel
Stirring
Technique
in
High-Temperature Solution Growth, J Cryst Growth, 13 (1971) 560-&. [13] B. Hong, S. Zhang, L. Zheng, H. Zhang, C. Wang, B. Zhao, Towards stabilizing thermal field and controlling growth interface shape in THM growth system of CdZnTe crystals: Proceedings of the 19th International Conference on Crystal Growth and Epitaxy (ICCGE-19), J Cryst Growth, (2019).
Highlights 1. Nucleation on unseeded THM growth of CZT crystals was studied. 2. The flow structure and parabolic temperature profile are reasons to secondary nucleation. 3. Large longitudinal and small temperature gradients can reduce the nucleation number 4. The typical columnar growth pattern was observed
Abstract
There are always undesired grains formed at the beginning of unseeded THM growth which deteriorate the crystal quality greatly. In this paper, several ingots with the diameter of 27mm were grown by unseeded THM to investigate the nucleation process. The growth temperature in the experiments was 750°C and the growth rate was 5mm/day. The influence of crucible bottom geometry and temperature gradient on nucleation are studied. Numerical simulations of thermal and concentration fields were also performed for the THM process to study the growth process. The simulation results show that square cavity flow's structure and parabolic wall temperature profile are the main reason for secondary nucleation in the radial direction. Dummy crystal with cold finger was introduced in order to stabilize the thermal field and modify the lateral temperature gradient. Subsequent experimental results indicate that a large longitudinal temperature gradient with small lateral temperature gradient is more favor to suppress secondary nucleation and maintain stable growth.
The authors declare no conflict of interest.
Declaration of interests
☒ 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.
☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
S0022-0248(20)30005-1 https://doi.org/10.1016/j.jcrysgro.2020.125482 CRYS 125482
To appear in:
Journal of Crystal Growth
Received Date: Revised Date: Accepted Date:
11 October 2019 11 December 2019 4 January 2020
Please cite this article as: B. Hong, S. Zhang, L. Zheng, H. Zhang, C. Wang, B. Zhao, Controlling nucleation during unseeded THM growth of CdZnTe crystal, Journal of Crystal Growth (2020), doi: https://doi.org/10.1016/ j.jcrysgro.2020.125482
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier B.V.
Controlling nucleation during unseeded THM growth of CdZnTe crystal Bangzhao Hong1, Song Zhang2*, Lili Zheng3, Hui Zhang1 Cheng Wang4, Bo Zhao4 1Department
of Engineering Physics, Tsinghua University, Beijing, China for Aero Engine, Tsinghua University, Beijing, China 3School of Aerospace Engineering, Tsinghua University, Beijing, China 4Ruiyan Technology Co. Ltd., Hangzhou, Zhejiang, China 2Institute
There are always undesired grains formed at the beginning of unseeded THM growth which deteriorate the crystal quality greatly. In this paper, several ingots with the diameter of 27mm were grown by unseeded THM to investigate the nucleation process. The growth temperature in the experiments was 750°C and the growth rate was 5mm/day. The influence of crucible bottom geometry and temperature gradient on nucleation are studied. Numerical simulations of thermal and concentration fields were also performed for the THM process to study the growth process. The simulation results show that square cavity flow's structure and parabolic wall temperature profile are the main reason for secondary nucleation in the radial direction. Dummy crystal with cold finger was introduced in order to stabilize the thermal field and modify the lateral temperature gradient. Subsequent experimental results indicate that a large longitudinal temperature gradient with small lateral temperature gradient is more favor to suppress secondary nucleation and maintain stable growth. Keywords: B2. CdZnTe; A2. Traveling heater method; A1. Nucleation; A2. Single crystal growth; A1. Mass transfer, heat transfer; A1. Computer simulation
1. Introduction CdxZn1-xTe, commonly known as CZT, has been recognized as the best material for room temperature nuclear radiation detector due to its excellent optoelectronic properties, such as suitable band gap, high average atomic number, high mobility and high value of μτ product electrons [1-4]. Despite decades of research on crystal growth of CZT, some of the difficulties in growth, including Te inclusions/precipitates, crystal size, compositional segregation and low growth rate, have limited the yield of CZT. In recent years, benefiting from lower growth temperature and better compositional homogeneity, Traveling Heater Method (THM) as one of solution methods, is becoming the most practical approach to grow high-quality CZT crystals [5, 6]. In order to obtain large-volume single crystal, the number of grains in ingot must be limited. For unseeded THM growth, controlling the nucleation position and reducing nucleation number at the initial stage are required for improving crystal size. So far, there has been limited work on nucleation control during unseeded THM growth of CZT. Roy et al. [7, 8] studied Te inclusions/precipitates and optimization of growth interface using unseeded THM. They used conical crucible to control nucleation, however, there was not much description on nucleation control. Surface energy plays a critical role in heterogeneous nucleation. Since good adhesion between molten CdTe and uncoated crucible, the nucleation in uncoated crucible is expected to occur with small undercooling and inhibit the formation of new grains after the first nucleation [9]. However the sticking between crystal and quartz often leads to movement and multiplication of dislocations even crystal cracking during cooling. Considering the relation between supercooling and nucleation, accurate prediction of supercooling region is also of great help for nucleation control [10, 11]. In experiment, controlling supercooling position can be achieved by enhancing heat release at the center of crucible bottom. Scheel[12] proposed an experimental scheme for controlling nucleation by combination of accelerated crucible rotation technique (ACRT) with localized cooling, and succeed in solution growth of GdAlO3. Such process can be used in CZT growth. In general, controlling the formation of supercooling region and undercooling during nucleation is the main part of nucleation control in unseeded growth. In the present work, the effects of crucible bottom geometry and temperature gradient including longitudinal and lateral ones on nucleation are studied. Both experimental and simulation methods are performed to understand and optimize the nucleation process.
2. Experimental Fig. 1 shows the schematic diagram of THM growth system. This furnace is homemade electric furnace. Only one heater controlled by PID controller is used to provide heat for crystal growth. The movement of crucible is achieved by lowering the crucible support which is driven by high precision stepping motor with repositioning precision of 2 microns. There are four thermocouples set at several different places of crucible wall to monitor the growth condition, and the distance between the upper and lower thermocouples is the designed length of molten zone. A dummy crystal with cold finger was introduced at the bottom of crucible. The main purpose of introducing dummy crystal with same thermal prosperities as CZT is to stabilize the thermal field during crystal growth [13]. All the polycrystalline CZT ingots were synthesized from 7 N metals by using the traditional Vertical Bridgman method. Prior to loading the charge, the quartz crucibles were carefully cleaned by ultrasonic cleaner, then immersed in nitric acid and rinsed repeatedly with ultrapure water, and finally coated with graphite layer by cracking high purity acetone. The excessive tellurium and polycrystalline CZT, used as solvent and feed respectively, were loaded into carbon-coated quartz crucible and sealed under dynamic vacuum of 5 × 10−4 Pa. At the initial state of growth, a relative low pull-down velocity of 3 mm/day was used to control nucleation, and then increased to 5 mm/day for crystal growth. There are six crystal growth experiments conducted in order to study the effects of different growth conditions on nucleation. The growth parameters used are listed in Table. 1. CZT-1 and CZT-2 are set to study the influence of conical crucible on nucleation. Double shoulders design with different angle is used. The first sharp shoulder is expected to provide a large undercooling within limited space and thus cold finger is not used in CZT-1 and CZT-2. The second shoulder with large cone angle is designed for grain enlargement. CZT-3 to CZT-6 focus on optimization of nucleation with different combination of longitudinal and lateral temperature gradients. The longitudinal temperature gradient refers to the wall temperature gradient at the bottom of crucible, and the lateral temperature gradient is the temperature difference between the center and wall of crucible bottom. The modification of longitudinal and lateral temperature gradients in the experiments are realized by changing the thickness of insulation at bottom and the material of cold finger respectively.
Fig. 1. Schematic of one-zone THM furnace and its temperature profile.
Table 1 The growth process parameters used in experiments. Longitudinal
Lateral
Bottom
temperature
temperature
Shape of
gradient (°C/cm)
gradient (°C/cm)
Crucible
85
35.0
/
See Fig. 2(a)
3~5
70
35.0
/
See Fig. 2(b)
CZT-3
3~5
45
42.1
26.7
CZT-4
3~5
45
75.2
26.7
CZT-5
3~5
45
75.2
44.2
CZT-6
3~5
45
75.2
18.3
Growth rate
Length of crystal
(mm/day)
(mm)
CZT-1
3~5
CZT-2
Case#
Fig. 2. The different geometric shape of crucible bottom.
See Fig. 2(c)
3. Results and discussion Fig .1 shows the startup status of unseeded THM growth of CZT while no grain forms on the crucible bottom. With the down-movement of crucible, the solution at the crucible bottom begins to be supercooled and nucleates after reaching the critical undercooling. In this paper, the formation and development of supercooling region are calculated to predict the nucleation. For the sake of simplifying the simulation, it is assumed that there is no grains formed on the crucible bottom as the crucible moves down. Besides, considering the computational cost in solving the unsteady equations, the dynamic growth process is approximated by translating the temperature profile on crucible surface. The complete description of model and material properties used in simulation are presented in our previous work[13] while the “single crystal” in original computational domain is removed and no-mass flux condition is set on crucible bottom.
3.1. Geometry Fig. 3 shows the as-grown ingots of 27mm diam grown in conical crucible. Those two ingots were cut perpendicular to the growth direction with a thickness of 5mm after finishing the shouldering process. It can be seen that the shoulder design plays a very important role in crystal selection. Only one grain is obtained both in CZT-1 and CZT-2 after finishing the first shouldering process. Unfortunately, the initial selected grain disappears soon, and new grains are formed at second shouldering process and show better enlargement tendency. Fig. 4 shows the evolution and development of supercooling regions in the melt with different cone angle as the crucible moves down. The wall temperature profile of crucible used in simulation is collected from experimental data of CZT-1 while the longitudinal temperature gradient is 35.0°C/cm. Simulation results reveal that the shoulder design ensures that supercooling region is preferentially formed in the center (see Fig. 4(b) and Fig. 4(c)). However, as the crucible continues to move down, a new supercooling region is formed in the radial direction, which is attributed to the square cavity flow's structure and parabolic wall temperature profile applied in THM. In the absence of flow, a convex isotherm is formed at the crucible bottom due to the parabolic wall temperature profile. However, buoyancy-driven natural convection transports heat from the wall to interior and heats the interior melt, which makes the isotherm concave near the center and forms a hump-shaped isotherm at the crucible bottom. As seen in Fig. 4(a), when the cone angle of crucible is very large, the supercooling region firstly appears in the radial direction of crucible instead of in the center. As is well known, the melt undercooling in solution growth is determined by local solute concentration and local temperature. As shown in Fig. 5, it is clear that the concentration alone the bottom of crucible is relative uniform with small increase from the center to the wall. There are two minimum points of temperature in the radial direction. The low temperature at the center is a result of conical crucible while the formation of the other one located at about 1/2R, where it is exactly the location of post-formed supercooling region, is caused by hump-shaped isotherm as analyzed above. Compared with the position of new grains formed at the second shoulder in the experiments, this hump-shaped isotherm distribution caused by the coupling of square cavity flow's structure and parabolic wall temperature profile is responsible for the secondary nucleation in the radial direction. Despite wide application of conical crucible in crystal growth, there is an unneglectable disadvantage of irregular crucible that is the changed length of molten zone during the shouldering
process. Since THM is zone melting method, the changed length of molten zone will affect the convection intensity of melt and result in redistribution of concentration and temperature fields.
Fig. 3. Longitudinal and lateral cross section of as grown crystals in different conical crucible, (a) from the bottom and (d) form the top.
Fig. 4. The evolution and development of supercooling regions in the melt of 1-inch crucible with different cone angle of (a) 165°, (b) 150°and (c) 135°. The right side of each case is streamline and the red area on the left represents the supercooling region. The numbers below each picture is the descending distance of crucible, i.e., +0mm represents the initial start position and +1mm is the distance that the crucible moves down from the initial start position.
Fig. 5. The temperature and concentration distribution along the radial direction of crucible bottom of third case in Fig. 4(b).
3.2 Temperature gradient Four as-grown CZT ingots under different combination of longitudinal and lateral temperature gradients are shown in Fig. 6. All the ingots were cut perpendicular to the growth direction with a thickness of 5mm. The number of grains of four ingots in the “first” wafer are listed in Table. 2. Due to crystal cracking, the wafer closest to the bottom and capable of counting the number of grains is used to represent the nucleation of as-grown ingot. In comparisons between CZT-3 and CZT-4, when the longitudinal temperature gradient rises from 42.1°C/cm to 75.2°C/cm with the same lateral temperature gradient of 26.7°C/cm, the number of initial grains decreases from 34 to 23. More importantly, the grains of CZT-4 exhibit typical columnar growth pattern after finishing nucleation (see in Fig. 6(b)). Increasing the longitudinal temperature gradient of growth interface shows a positive effect on nucleation. However, the large longitudinal temperature gradient used in subsequent experiments is mainly for maintaining the stable growth of crystal. According to the classical theory of constitutional supercooling, a large longitudinal temperature gradient is believed to suppress constitutional supercooling and allow increased growth rate. Moreover, a large longitudinal temperature gradient is easier to be realized and minimizes the effects of temperature fluctuation on crystal growth. Crystal growth experiments under different lateral temperature gradients by changing the material of cold finger were conducted while the longitudinal temperature gradient remains at 75.2°C/cm. From CZT-5, CZT-4 and CZT-6, it is obvious that the number of initial grains decreases rapidly from 55 to 14 when the lateral temperature gradient decreases from 44.2°C/cm to 18.3°C/cm. Besides that, it is worth mentioning that all grains of CZT-4, CZT-5 and CZT-6 are enlarging in subsequent crystal growth, and CZT-4 and CZT-6 show the obvious columnar growth pattern. In order to understand nucleation under different lateral temperature gradients, a series of simulation experiments were carried out. In this part, the wall temperature profile of crucible was approximated by Gauss distribution with the longitudinal temperature gradient of 70°C/cm, and three groups of lateral temperature gradients, 10°C/cm, 30°C/cm and 50°C/cm, were set for comparison. Since a slow speed of 3mm/day was used in the experiments, it is assumed that the quasi-steady state is reached. Then the evolution of supercooling region with time is obtained by solving the flow field at different positions in the furnace. As seen in Fig. 7, when the lateral temperature gradient decreases, the supercooling region gradually moves outward, which is also the reason for the radial nucleation in the conical crucible. The main reason for the fluctuation of undercooling under large lateral temperature gradient of 50°C/cm shown in Fig. 7 is the redistribution of concentration and temperature fields caused by dynamic change of flow structure. Fig. 8 shows the width of “nucleation” area at the crucible bottom under assumed nucleation undercooling of 10°C. Although the exact value of nucleation undercooling is unknown, it does not change the trend of curve in Fig. 8. It is clear from the curve that when the lateral temperature gradient is 50°C/cm and 30°C/cm, the width of the “nucleation” area shows a rapid increase at the initial stage, and then slowly increases to the maximum. Meanwhile the change is linear with smaller slope under lateral temperature gradient of 10°C/cm. Since the supercooling point under large lateral temperature gradient firstly appears in the center, a large supercooling area will be formed at the center in a short time, which may result in large-scale random nucleation. On the contrary, a ring-shape supercooling region with smaller linear increase in size is more suitable for the formation and growth of single grain. Because the supercooling point lies in the radial
direction of the crucible under small lateral temperature gradient, it is hard to form a single grain on crucible bottom.
Fig. 6. The as-grown CZT wafers under the different combination of longitudinal and lateral temperature gradients, (a) from the bottom and following letters closer to top.
Table 2 Nucleation number under the different combination of longitudinal and lateral temperature gradients. Growth rate
Longitudinal temperature
Lateral temperature
(mm/day)
gradient (°C/cm)
gradient (°C/cm)
CZT-3
3
42.1
26.7
34
CZT-4
3
75.2
26.7
~23
CZT-5
3
75.2
44.2
~55
CZT-6
3
75.2
18.3
14
Case#
Number of grains
Fig. 7. Development of undercooling of crucible bottom with time while the longitudinal temperature gradient is 70°C/cm.
Fig. 8. Width of “nucleation” area at the crucible bottom with assumed nucleation undercooling of 10°C.
4. Conclusions In this work, nucleation during unseeded THM growth of CZT is studied. Both experiment and simulation are adopted to study the influence of crucible bottom geometry and temperature gradient including longitudinal and lateral temperature gradients on nucleation. The experiment results show that the conical crucible is helpful for initial crystal selection and only one grain was obtained both in conical crucibles after finishing the first shouldering process. However, secondary nucleation inevitably occurs in the radial direction during the second shouldering process and the selected grain disappears soon. Simulation results reveal that the square cavity flow's structure and parabolic wall temperature profile are the main reason for the secondary nucleation. Subsequent experiments in flat-bottom crucible show that the large longitudinal temperature gradient and small radial temperature gradient are better for growth and nucleation. Under the combination of large longitudinal temperature gradient of 75.2°C/cm and small lateral temperature gradient of 18.3°C/cm, the nucleation number decreases obviously from 55 to 14, more importantly, the typical columnar pattern is observed in subsequent crystal growth.
Acknowledgements This work is supported by National Key R&D Program of China (2016YFB1102300), National Science Foundation of China (Grant Nos. 91646201, U1633203) and Beijing Key Laboratory of City Integrated Emergency Response Science.
Reference [1] H. Chen, S.A. Awadalla, K. Iniewski, P.H. Lu, F. Harris, J. Mackenzie, T. Hasanen, W. Chen, R. Redden, G. Bindley, I. Kuvvetli, C. Budtz-Jorgensen, P. Luke, M. Amman, J.S. Lee, A.E. Bolotnikov, G.S. Camarda, Y. Cui, A. Hossain, R.B. James, Characterization of large cadmium zinc telluride crystals grown by traveling heater method, J Appl Phys, 103 (2008). [2] C. Szeles, Advances in the crystal growth and device fabrication technology of CdZnTe room temperature radiation detectors, Ieee T Nucl Sci, 51 (2004) 1242-1249. [3] T.E. Schlesinger, J.E. Toney, H. Yoon, E.Y. Lee, B.A. Brunett, L. Franks, R.B. James, Cadmium zinc telluride and its use as a nuclear radiation detector material, Materials Science and Engineering: R: Reports, 32 (2001) 103-189. [4] M. Schieber, R.B. James, H. Hermon, A. Vilensky, I. Baydjanov, M. Goorsky, T. Lam, E. Meerson, H.W. Yao, J. Erickson, E. Cross, A. Burger, J.O. Ndap, G. Wright, M. Fiederle, Comparison of cadmium zinc telluride crystals grown by horizontal and vertical Bridgman and from the vapor phase, J Cryst Growth, 231 (2001) 235-241. [5] H. Chen, S.A. Awadalla, R. Redden, G. Bindley, A.E. Bolotnikov, C.G. Carini, R.B. James, High-Performance, Large-Volume THM CdZnTe Detectors for Medical Imaging and Homeland Security Applications, Ieee Nucl Sci Conf R, (2006) 3629-3637. [6] H. Shiraki, M. Funaki, Y. Ando, A. Tachibana, S. Kominami, R. Ohno, THM Growth and Characterization of 100 mm Diameter CdTe Single Crystals, Ieee T Nucl Sci, 56 (2009) 1717-1723. [7] U.N. Roy, A. Gueorguiev, S. Weiller, J. Stein, Growth of spectroscopic grade Cd0.9Zn0.1Te:In by THM technique, J Cryst Growth, 312 (2009) 33-36. [8] U.N. Roy, S. Weiler, J. Stein, Growth and interface study of 2 in diameter CdZnTe by THM
technique, J Cryst Growth, 312 (2010) 2840-2845. [9] R. Shetty, W.R. Wilcox, L.L. Regel, Influence of Ampoule Coatings on Cadmium Telluride Solidification, J Cryst Growth, 153 (1995) 103-109. [10] J.H. Peterson, M. Fiederle, J.J. Derby, Analysis of the traveling heater method for the growth of cadmium telluride, J Cryst Growth, 454 (2016) 45-58. [11] C.W. Lan, D.T. Yang, A computer simulation of crystal growth by the traveling-solvent method (TSM): pseudo-steady-state calculations, Model Simul Mater Sc, 3 (1995) 71-92. [12]
H.J.
Scheel,
Accelerated
Crucible
Rotation
-
Novel
Stirring
Technique
in
High-Temperature Solution Growth, J Cryst Growth, 13 (1971) 560-&. [13] B. Hong, S. Zhang, L. Zheng, H. Zhang, C. Wang, B. Zhao, Towards stabilizing thermal field and controlling growth interface shape in THM growth system of CdZnTe crystals: Proceedings of the 19th International Conference on Crystal Growth and Epitaxy (ICCGE-19), J Cryst Growth, (2019).
Highlights 1. Nucleation on unseeded THM growth of CZT crystals was studied. 2. The flow structure and parabolic temperature profile are reasons to secondary nucleation. 3. Large longitudinal and small temperature gradients can reduce the nucleation number 4. The typical columnar growth pattern was observed
Abstract
There are always undesired grains formed at the beginning of unseeded THM growth which deteriorate the crystal quality greatly. In this paper, several ingots with the diameter of 27mm were grown by unseeded THM to investigate the nucleation process. The growth temperature in the experiments was 750°C and the growth rate was 5mm/day. The influence of crucible bottom geometry and temperature gradient on nucleation are studied. Numerical simulations of thermal and concentration fields were also performed for the THM process to study the growth process. The simulation results show that square cavity flow's structure and parabolic wall temperature profile are the main reason for secondary nucleation in the radial direction. Dummy crystal with cold finger was introduced in order to stabilize the thermal field and modify the lateral temperature gradient. Subsequent experimental results indicate that a large longitudinal temperature gradient with small lateral temperature gradient is more favor to suppress secondary nucleation and maintain stable growth.
The authors declare no conflict of interest.
Declaration of interests
☒ 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.
☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: