Directional Solidification Process Research Articles

Single-step directional solidification technology for solar grade polysilicon preparation

We propose a modified process for the massive production of polysilicon by directional solidification (DS) technology designed to change the currently used “two-step DS process” into a “single-step DS process”, to enable the efficient production of polysilicon with lower energy and material consumption. The temperature distribution, solid-liquid (s/l) interface shapes, and thermal stresses during the DS process using the conventional process and the modified processes were studied by three-dimensional transient simulations. The simulation results about temperature, s/l interface, and thermal stresses were validated indirectly by corresponding experiments through the test for impurity segregation effect, minority carrier lifetime, and resistivity. It was found that the ingots produced by the modified process showed ideal solidification with uniform temperature distribution, slightly convex interface shapes, and lower thermal stresses, resulting in relatively higher carrier lifetimes and uniform distribution resistivity, which indicates that the single-step DS process is feasible to produce solar grade polysilicon.

Removal of metal impurities by controlling columnar grain growth during directional solidification process

Silicon ingots with columnar grains and irregular grains as raw material for solar cells were obtained by directional solidification. The relationship between crystal morphology and concentration of metal impurities was studied. The result shows that columnar grains can improve the ratio of the high-purity area, where the concentration of metal impurities is less than 10ppmw and more consistent with the calculated value by the Scheil’s equation. Meanwhile, the metal silicide precipitates at the irregular grain boundaries. Large temperature gradient at the solid-liquid interface dedicated to the stability of solid-liquid interface, which is conductive to the growth of the columnar grains. High concentration of metal impurities introduces a large constitutional supercooling to reduce the stability of solid-liquid interface, then further form a mushy zone in front of the solid-liquid interface, which leads to the growth of irregular grains.

Effects of Precipitate Element Addition on Microstructure and Magnetic Properties in Magnetostrictive Fe83Ga17alloy

The <100> oriented <TEX>$Fe_{83}Ga_{17}$</TEX> alloys with various contents of NbC or B were prepared by directionally solidification method at the growth rate of <TEX>$720mm{\cdot}h^{-1}$</TEX>. With a small amount of precipitates, the columnar grains grew with cellular mode during directional solidification process, while like-dendrite mode of grains growth was observed in the alloys with higher contents of 0.5 at% due to the dragging effect of precipitates on the boundaries. The NbC precipitates disperse both inside grains and along the boundaries of <TEX>$Fe_{83}Ga_{17}$</TEX> alloys with NbC addition, and the Fe2B secondary phase particles preferentially distribute along the grain boundaries in B-doped alloys. Precipitates could affect grain growth and improved the <100> orientation during directional solidification process. Small amount of precipitate element addition slightly increased the magnetostrictive strain, and a high value of 335 ppm under pre-stress of 15 MPa was achieved in the alloys with 0.1 at% NbC. Despite the fact that the effect on magnetic induction density of small amount of precipitates could be negligible, the coercivity markedly increased with addition of precipitate element for <TEX>$Fe_{83}Ga_{17}$</TEX> alloy due to the retarded domain motion resulted by precipitates.

Directional melting and solidification of gallium in a traveling magnetic field as a model experiment for silicon processes

Small-scale model experiments for directional solidification processes are performed using a gallium volume with a square horizontal cross-section and dimensions of 10×10×7.5cm3. A heater at the top and a cooler at the bottom generate a vertical temperature gradient while an external coil system produces a traveling magnetic field (TMF) leading to Lorentz forces in the melt. The position and shape of the phase interface as well as the melt flow during melting and solidification processes are investigated both experimentally and with a coupled 3D numerical model. Uncertainty in various experimental parameters and appropriate methods of calibration are discussed to enable precise validation of numerical simulations. A distinct influence of the melt flow is observed, which results in a concave melting interface with an upward TMF and a convex shape with a downward TMF. In both cases, the corner region demonstrates local deflections in the opposite directions, which illustrates the challenge to obtain a smooth interface shape in silicon solidification processes. These processes can be further investigated using the validated 3D model. Additionally, direct transfer of the results between model experiments and silicon processes using scaling laws is discussed.

Comparative Investigation of the Downward and Upward Directionally Solidified Single-Crystal Blades of Superalloy CMSX-4

Single-crystal blades of Ni-base superalloys CMSX-4 have been directionally solidified using the downward directional solidification (DWDS) process. The possible benefits of the process were comparatively evaluated with respect to the Bridgman process’ results. The DWDS process exhibits good capabilities for casting the single-crystal components. The thermal gradients of this process are approximately seven times higher than those of the Bridgman process. It provides more advantages for solidifying the single-crystal superalloy blades by reducing the casting defects, refining the microstructure, decreasing the size of the γ/γ′ eutectic pools, refining the γ′ precipitates, alleviating the degree of the microsegregation, and minimizing the size and volume fraction of the micropores.

Radiation heat transfer model for complex superalloy turbine blade in directional solidification process based on finite element method

For the sake of a more accurate shell boundary and calculation of radiation heat transfer in the Directional Solidification (DS) process, a radiation heat transfer model based on the Finite Element Method (FEM) is developed in this study. Key technologies, such as distinguishing boundaries automatically, local matrix and lumped heat capacity matrix, are also stated. In order to analyze the effect of withdrawing rate on DS process, the solidification processes of a complex superalloy turbine blade in the High Rate Solidification (HRS) process with different withdrawing rates are simulated; and by comparing the simulation results, it is found that the most suitable withdrawing rate is determined to be 5.0 mm∙min-1. Finally, the accuracy and reliability of the radiation heat transfer model are verified, because of the accordance of simulation results with practical process.

BaZrO3 refractory applied to the directional solidification of TiAl alloys

Recently, much attention has been paid to the refractory used for the directional solidification process of TiAl intermetallics, the Y2O3 crucible/Y2O3 coated moulds seem to be the suitable candidate. However, the use of Y2O3 is limited by its low thermal shock resistance and high cost. In this work, a novel BaZrO3 refractory was introduced to the directional solidification of TiAl intermetallics. The melt of this alloy contained in BaZrO3 crucibles were heated for 30 minutes at 1600, 1650, 1700 °C, respectively, then cooled within the crucible for the investigation of the interface between the melt and the refractory. The scanning electron microscopy (SEM) was used to evaluate the surface topography and microstructure of the samples, and the energy dispersive spectroscopy (EDS) was used to analyse the chemical composition of the samples. The results indicated that no interfacial interaction layer and no obvious element diffusion were observed between the crucible and the metal, which may imply that the BaZrO3 is a promising candidate of refractory for the directional solidification of TiAl alloys. This work can provide a basis for the further study of directional solidification of TiAl alloys by using BaZrO3 shell mould.

Microstructure and property of directionally solidified Ni–Si hypereutectic alloy

This paper investigates the influence of the solidification rate on the microstructure, solid/liquid interface, and micro-hardness of the directionally solidified Ni–Si hypereutectic alloy. Microstructure of the Ni–Si hypereutectic alloy is refined with the increase of the solidification rate. The Ni–Si hypereutectic composite is mainly composed of α-Ni matrix, Ni–Ni3Si eutectic phase, and metastable Ni31Si12 phase. The solid/liquid interface always keeps planar interface no matter how high the solidification rate is increased. This is proved by the calculation in terms of M–S interface stability criterion. Moreover, the Ni–Si hypereutectic composites present higher micro-hardness as compared with that of the pure Ni3Si compound. This is caused by the formation of the metastable Ni31Si12 phase and NiSi phase during the directional solidification process.

Influence of thermal stabilization treatment on microstructure evolution of the mushy zone and subsequent directional solidification in Ti-43Al-3Si alloy

Thermal stabilization (TS) and directional solidification (DS) experiments were performed on Ti-43Al-3Si alloy in a Bridgman-type furnace. The effects of thermal stabilization time on the microstructure evolution of mushy zone and initial interface of directional solidification were investigated. The results show that the volume fraction of liquid in mushy zone and the length of mushy zone decrease as the thermal stabilization time increases. The concentrations of Al and Si exhibit a decreasing trend in the complete liquid zone with increasing TS time, while the solute concentrations of Al and Si in complete liquid zone are always higher than those of the original as-cast alloy. A long TS time promotes solute diffusion and provides a solute homogeneous initial interface for subsequent DS process. Isolated α grains and Ti5Si3 particles in mushy zone tend to coarsen with the increase of thermal stabilization time. Long time TS treatment also leads to growth of Ti5Si3 particles which may act as nucleation sites to terminate the continuous growth of dendrite and inheritance of the lamellar orientation. Directionally solidified samples with aligned lamellar structures can be obtained within 30min TS treatment in Ti-43Al-3Si alloy.

Microstructure of multicrystalline silicon seeded by polysilicon chips and fluidized bed reactor granules

Multicrystalline silicon displays a considerable smaller average grain size and reduced dislocation generation when being seeded by polycrystalline silicon chips or fluidized bed reactor silicon granules. A simple texture analysis shows how the initially random grain structure of the seeds develops a weak preference for near-〈111〉 and near-〈112〉 oriented grains upwards in the ingot. Closer investigations reveal a considerable coarsening of the initial microstructure of the seeds during the directional solidification process, especially for small fluidized bed reactor granules. The irregular shape of polysilicon chips allows for melt penetration into the seeding structure and potential indentation effects that may account for the increased dislocation generation observed in this case. The increased generation may, however, also be related to a higher ratio of ∑27 grain boundaries.

Effects of the hot zone design during the growth of large size multi-crystalline silicon ingots by the seeded directional solidification process

In this study, the installation of insulation blocks in the hot zone is utilized to assist in the growth of multi-crystalline silicon ingots with 800kg of silicon charge using the seeded directional solidification method. A transient global numerical simulation is carried out to investigate the heat and mass transport during growth process. At a higher solidification fraction, lower concavity of the crystal–melt interface near the crucible wall can be obtained as compared to the standard model. The lowest concavity and highest energy saving is achieved when insulation blocks are added to the side of a directional solidification block and to the low part of the side insulation. The simulation results for this design also show a reduction of the melt velocity. The average oxygen concentration is slightly higher along the crystal–melt interface, compared to the standard one.

Influence of Crucible Thermal Conductivity on Crystal Growth in an Industrial Directional Solidification Process for Silicon Ingots

We carried out transient global simulations of heating, melting, growing, annealing, and cooling stages for an industrial directional solidification (DS) process for silicon ingots. The crucible thermal conductivity is varied in a reasonable range to investigate its influence on the global heat transfer and silicon crystal growth. It is found that the crucible plays an important role in heat transfer, and therefore its thermal conductivity can influence the crystal growth significantly in the entire DS process. Increasing the crucible thermal conductivity can shorten the time for melting of silicon feedstock and growing of silicon crystal significantly, and therefore large thermal conductivity is helpful in saving both production time and power energy. However, the high temperature gradient in the silicon ingots and the locally concave melt-crystal interface shape for large crucible thermal conductivity indicate that high thermal stress and dislocation propagation are likely to occur during both growing and annealing stages. Based on the numerical simulations, some discussions on designing and choosing the crucible thermal conductivity are presented.

Simulation of liquid channel of Fe-C alloy directional solidification by phase-field method

In directional solidification, two characteristic parameters determine the dendritic growth: the thermal gradient and the pulling velocity. To achieve the suitable microstructure and improve the performance of casting, they are usually used to resize the pulling velocity or temperature gradient in directional solidification process. The structures obtained under different directional solidification conditions, and their associated properties both have been hot research points. It is difficult to observe the microstructure, which is usually on a micrometer scale, directly in experiment, and the phase-field method becomes a strong tool to understand the dendrite growth pattern. We mainly study the liquid channel formed after Fe-C alloy dendrite tip splitting under the specific condition of directional solidification and analyze the influence on liquid channel of pulling velocity in this paper. We choose the fixed thermal gradient G =20 K/mm which is on the order of the experimental value, and pulling velocity VP no more than 10 mm/s to keep the cooling rate in the range of low speed in dendrite growth, so that the interface kinetic effect can be neglected. Recent experimental results show the different interfacial energies in various compositions of Al-Zn alloy and Fe-C alloy, then we can investigate a series of directional solidification microstructures with fixed alloy Fe-0.5 wt.%C composition at different interfacial energies in our simulations. We find that the liquid channel is formed as a result of anisotropy competition between system and materials, the length and C concentration of liquid channel increase with the pulling velocity increasing, while the diameter of liquid channel is constant. It is interesting to find that there is a minimum of pulling velocity almost equal to 1 mm/s, the tip will not split and no liquid channel forms in the following steps either when the velocity is smaller than the minimum. We also compare the segregation caused by solute enrichment in liquid channel and solute segregation between dendrite arms in a series of simulations: the former is more serious than the latter. Then we point out the way to reduce the segregation caused by liquid phase channel by reducing the pulling velocity properly. It will be more practical to couple the flow field with other external field, such as magnetic field, in the simulation.

A mathematical model for distribution of calcium in silicon by vacuum directional solidification

Calcium is one of the main impurity elements in silicon. The removal of calcium strongly affects the quality of the polycrystalline silicon ingot produced by a vacuum directional solidification method. Based on the considerations of the theory of segregation, mass transfer and evaporation during vacuum directional solidification process, a mathematical model for calcium distribution in silicon was proposed and it can be used to explain the removal mechanism. In order to confirm the mathematical model, an industrial scale experiment on UMG-Si with an initial purity of 99.98 wt. % was performed. Since the reaction temperature strongly influences both the evaporation and segregation of calcium, the dependences of effective segregation coefficient (keff) and the evaporation coefficient (kE) on temperature were carefully investigated. The results showed that the proposed mathematical model was highly consistent with the experimental data and the calcium removal efficiency mainly relied on the evaporation step.

Complex banded structures in directional solidification processes

A combination of theory and numerical simulation is used to investigate impurity superstructures that form in rapid directional solidification (RDS) processes in the presence of a temperature gradient and a pulling velocity with an oscillatory component. Based on a capillary wave model, we show that the RDS processes are associated with a rich morphology of banded structures, including frequency locking and the transition to chaos.

Effect of alternating magnetic field on the removal of metal impurities in silicon ingot by directional solidification

Multicrystalline silicon ingots without and with alternating magnetic field during directional solidification process under industrial system were obtained from metallurgical grade silicon (MG-Si). The concentrations and normalized concentrations of metal impurities in the two silicon ingots were studied. The result shows that the concentrations and normalized concentrations in high-purity area of the silicon with alternating magnetic field are lower than those of the ingot without alternating magnetic field. The transport mechanism for metal atoms in the diffusion layer area has been changed due to the alternating magnetic field. Alternating magnetic field introduces a convection to reduce the thickness of diffusion layer in the molten silicon, which results in a decreased effective segregation coefficients. Enhancing transport driving force of metal atoms in molten silicon is the effective way to improve the removal rate of metal impurities.

The effects of microstructure and growth rate on microhardness, tensile strength, and electrical resistivity for directionally solidified Al–Ni–Fe alloys

Directional solidification of eutectic alloys attracts considerable attention because of microhardness, tensile strength, and electrical resistivity influenced by eutectic structures. In this research, we examined processing of Al–Ni–Fe (Al–6.5wt.%Ni–1.5wt.%Fe) eutectic by directional solidification. The alloy was prepared by vacuum furnace and directionally solidified in Bridgman-type equipment. During the directional solidification process, the growth rates utilized varied from 8.25 μm/s to 164.80 μm/s. The Al–Ni–Fe system showed a eutectic transformation, which resulted in the matrix Al, rod Al3Ni, and plates Al9NiFe phases. The eutectic spacing between (λAl3Ni−λAl3Ni,λAl9NiFe−λAl9NiFe)was measured. Additionally, the microhardness, tensile strength, and electrical resistivity of the alloy were determined using directionally solidified samples. The effects of growth rates on microhardness, tensile strength, and electrical resistivity for directionally solidified Al–Ni–Fe eutectic alloy were investigated, and the relationships between them were experimentally obtained. It was found that the microhardness, tensile strength, and electrical resistivity were affected by both eutectic spacing and the solidification parameter.

The effect of radiative heat transfer characteristics on vacuum directional solidification process of multicrystalline silicon in the vertical Bridgman system

In the vertical Bridgman system for multicrystalline silicon vacuum directional solidification process, the radiation heat transfer characteristics have a significant impact on the quality of crystal growth. In this paper, the effect of heat transfer characteristics under different pulling-down rates on the vacuum directional solidification process was investigated. Simplified radiation view factor dependent on pulling-down rate in pilot-scale vertical Bridgman system was derived. Combined with the transient simulation results, we obtained the heat transfer characteristics inside the furnace, and proposed the optimization scheme by using a rate of slowly first and then faster to pull down the crucible, which can not only satisfy thermal conditions of crystal growth, but also shorten the solidification time and effectively reduce the energy consumption. In addition, according to the characteristics of the radiative heat transfer under vacuum condition and the change of view factor, it was found that an improved design by widening the cold zone of the furnace can narrow the temperature gradient in the silicon ingot, and greatly decrease the thermal stresses, which in turn can aid in the design of future industrial equipment and process improvements.

Investigation of liquid oxide interactions with refractory substrates via sessile drop method

The wetting of refractory substrates by oxide alloys has been investigated using a high-temperature sessile drop bench. The aim of this study was to select the best crucible material for directional solidification processes such as the Bridgman, edge-defined film-fed growth or micro-pulling down methods. Contact angles and surface tension were determined via high-resolution imaging. The O2 partial pressure in the furnace atmosphere was also measured. Alumina and a eutectic ceramic called A/YAG/Z (Al2O3/Y3Al5O12/ZrO2) were studied by investigating the following systems: alumina/molybdenum, alumina/tungsten and alumina/iridium; A/YAG/Z–molybdenum, A/YAG/Z–tungsten and A/YAG/Z–iridium. The results show intermediate wetting for both oxides (30° < θ <50°). This is an original result for A/YAG/Z. Several previous studies reported good wetting (7° < θ <20°) for alumina. Accurate surface tension values could not be obtained by image processing. Therefore, the Wilhelmy method was used, giving γ = 0.63 ± 0.03 N m−1 for alumina, which is comparable to the value of 0.67 ± 0.03 N m−1 reported in the literature. The surface tension obtained for A/YAG/Z was 0.71 N m−1, but this value was influenced by the approximate density used. A metal-like layer was observed on the apex of the drops after solidification. According to an SEM–EDS analysis, most of these layers were composed of molybdenum, tungsten and oxygen. Based on thermodynamic calculations, a mechanism which takes into account the oxidation of Mo parts and the dissolution of oxides in the drop is proposed.

Crystal orientation control in TiAl–Nb alloys through a double directional solidification process

Intermetallic Ti–46Al–5Nb (at.%) alloys were directionally solidified by single directional solidification (SDS) process and double directional solidification (DDS) process using a Bridgman directional solidification furnace. Microstructure and crystal orientation neighboring the initial growth interface as well as the quenched mushy zone before starting directional solidification were investigated in the SDS and DDS processes, respectively. The results show that the dendrite morphology changes from equiaxial to columnar, but the directional dendrite segregation is almost keep the continuity across the initial growth interface after SDS and DDS processes, respectively. Crystal orientations in the regions neighboring the initial growth interface have preferential modifications along the growth direction. Compared with that in the first DS of DDS, the length of the initial mushy zone significantly decreases, but the β dendrite is well-orientated in the mushy zone in the second DS. Crystal orientation can be well-controlled in DDS by modifying the β growth morphology as well as the orientation in initial mushy zone of the two DS processes. The directional dendrite parallel to the growth direction and the preferred β orientation are considered as the main reasons to β seed crystallization. The process of the β seed crystallization is also discussed. Based on the β seeding, the DDS process can align the β dendrite orientation and promote the peritectic transformation undergoing completely, thus control the lamellar colony boundaries as well as the lamellar microstructures inside of the colonies.