JP2768499B2 - Optimal condition analysis method and control method for single crystal pulling furnace - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a single crystal pulling furnace capable of optimally controlling a temperature distribution in a furnace, a solid-liquid interface shape, a thermal stress of a crystal, and the like by simulation analysis. The optimal condition analysis method and control method.

[Conventional technology]

GaAs single crystals have attracted attention as substrates for high-speed ICs and optical devices, and have been actively studied. As a method of forming a single crystal, it is one of the main methods of forming a bulk crystal of a group III-V compound semiconductor, and a large-diameter, circular wafer is easily formed.
The LEC method (liquid sealed Czochralski method) is the mainstream.

III-V compound semiconductor bulk crystal growth
Due to the nature of the family, it is performed in a complex thermal environment under high temperature and high pressure. In this crystal growth, in order to grow crystals with lower dislocations, in order to reduce the thermal stress by relaxing the thermal environment and to increase the productivity, the contradictory requirements of increasing the temperature gradient and strengthening the directional solidification are required. There is. For this reason, since a heat insulating material is placed in the furnace or a multi-stage heater is employed, control is extremely complicated and difficult.

FIG. 9 is a diagram showing a configuration outline of a single crystal pulling apparatus of the LEC method, 31 is a water-cooled reflector, 32 is graphite, 33 is a sub upper heater, 34 is a main heater, 35 is a sub lower heater, 36 is a crucible, 37 is the crystal part, 38 is the melt part,
39 indicates B ₂ O ₃ , and 40 indicates an insulating felt. The furnace is placed in a high-pressure (10 atm) atmosphere of, for example, a non-volatile Ar (argon) gas, and is heated to a high temperature of about 1511 ° (K), the melting point of GaAs, by a heater (33 to 35). Then, the heater power, the pulling speed, the rotation speed of the crucible 36 and the crystal 37, and the like are controlled while measuring the crystal shape during the pulling and growing.
The measurement of the crystal shape during the pull-up and growth is performed by detecting the crystal weight using a load cell, and measuring the change in the crystal weight while correcting the buoyancy and the surface tension of the melt with B ₂ O ₃ 39.

There are many conditions for crystal growth by the LEC method, but the most important one is the thermal environment in the furnace. In order to grow high quality crystals and achieve high productivity, it is essential to understand the thermal environment. However, as described above, the inside of the furnace is filled with a high-temperature, high-pressure atmosphere and has a complicated hot zone structure, so that the entire inside of the furnace cannot be directly observed during crystal growth. Therefore, computer simulation is an indispensable technique to achieve a desired thermal environment, and many analyzes have been attempted so far.

For example, temperature distribution in a furnace, temperature distribution in a crystal, thermal convection in a melt, analysis of a solid-liquid interface shape coupled with a crystal / melt, analysis of a thermal stress in a crystal, and the like have been vigorously performed, and results have been obtained.

[Problems to be solved by the invention]

However, conventionally, various numerical analyzes or computer simulations have been performed as described above, but each of these analyzes is an analysis of each element alone, and is not an analysis in which each element is related, It was not directly linked to the furnace operating conditions. Therefore, there is a problem that it is difficult to determine how to change the setting in order to optimize the thermal environment in the furnace, and that it takes time to confirm the result.

In order to improve the productivity, it is necessary to increase the size of the crucible for charging a large amount of charge, that is, the raw material. Further, in order to meet the demand for large-diameter wafers, it is necessary to increase the size of the crucible. However, usually, the size of the whole furnace is fixed, and there is a situation in which the deformation cannot be performed due to various restrictions. In addition, since the heater that controls the temperature inside the furnace is located outside the crucible, if the crucible is made larger, the heat insulation part becomes thinner, and the safety of the furnace becomes a problem.
In addition, the possibility of crystal growth is unknown.

In order to actually increase the size of the crucible, it is necessary to first create a hot zone corresponding to the increase in the size of the crucible by performing the above-described analysis, and then repeat the crystal pulling experiment. This costs millions of yen,
It takes about half a year, and there is a lot of lost time and money.

The present invention has been made to solve the above problems, and an object of the present invention is to provide an optimal condition analysis method for a single crystal pulling furnace that can be analyzed efficiently and with high accuracy using a simulation. It is another object of the present invention to provide a control method of a single crystal pulling furnace capable of efficiently controlling a single crystal pulling by utilizing such an analysis result.

[Means for solving the problem]

Therefore, the present invention is an optimal condition analysis method of a single crystal pulling furnace in which a raw material is put into a crucible, and the raw material is heated to a high temperature around its melting point by a heater under a high-pressure gas atmosphere to pull up and grow a single crystal. Thermal fluid analysis means for obtaining the temperature distribution in the furnace based on the settings of the shape, boundary conditions, heater power, etc. in the furnace, taking into account the gas flow and radiation, and crystallization / melting using the temperature distribution in the furnace as the boundary conditions A solid-liquid interface shape analyzing means for performing a thermal analysis of the liquid portion to obtain a solid-liquid interface shape in the crystal / melt portion; and a thermal stress analysis for obtaining a thermal stress in the crystal portion from the solid-liquid interface shape and the temperature distribution of the crystal portion. Means for transferring the temperature distribution in the furnace from the thermal fluid analysis means to the solid-liquid interface shape analysis means, and the solid-liquid interface shape analysis means for converting the solid-liquid interface shape and the temperature distribution of the crystal part to the heat distribution. Pull to stress analysis means It characterized in that as successor. In addition, the control method of the single crystal pulling furnace using these means further includes setting means for setting the shape of the furnace, boundary conditions, heater power, etc., and operating conditions for pulling the single crystal in the furnace. Monitoring means for monitoring, and control means for performing single crystal pull-up control of the furnace while performing monitoring means by the monitoring means using the condition set by the setting means as a control signal; and The temperature distribution of the inside is taken over by the solid-liquid interface shape analysis means, and the solid-liquid interface shape and the temperature distribution of the crystal part are taken over by the thermal stress analysis means from the solid-liquid interface shape analysis means. The setting conditions of the setting means are sequentially corrected by evaluating the analysis results of the solid-liquid interface shape analysis means and the thermal stress analysis means.

[Action]

In the optimal condition analysis method of the single crystal pulling furnace according to the present invention,
Temperature and temperature of furnace and crystal, solid-liquid interface shape, heat of crystal Since the simulation of the stress is performed, the simulation conditions can be corrected by the evaluation at each step, and the scale-up of the single crystal pulling furnace and the follow-up of high quality and high production can be efficiently performed.

Further, in the control method of the single crystal pulling furnace using the analysis means, the setting conditions of the condition setting means are sequentially corrected by evaluating the analysis results, and the control means monitors the furnace by the monitoring means based on the conditions. Since the crystal pulling control is performed, a high-quality single crystal can be efficiently manufactured.

〔Example〕

 Hereinafter, embodiments will be described with reference to the drawings.

FIG. 1 is a view for explaining one embodiment of an optimum condition analysis method for a single crystal pulling furnace according to the present invention.

In the optimal condition analysis method of the single crystal pulling furnace according to the present invention,
For crystal growth, first set the crucible shape (diameter), heater length, and furnace structure, set the heater power using the hot zone outer wall temperature and gas inlet flow velocity as boundary conditions,
Thermo-fluid analysis unit 1 for obtaining temperature distribution in the furnace, focuses on the temperature of the crystal / melt part from the temperature distribution in the furnace, and takes over the temperature distribution in the furnace as a boundary condition to perform thermal analysis of the crystal / melt part A solid-liquid interface shape analysis unit 2 for obtaining a solid-liquid interface shape, and a thermal stress analysis unit 3 for obtaining a thermal stress in a crystal part from the solid-liquid interface shape obtained by the solid-liquid interface shape analysis unit 2 and a temperature distribution of the crystal part.
The simulation system is configured so as to be divided into three blocks for evaluation. The data input unit 4 inputs data such as case settings, boundary conditions, and heater power. With such a configuration, the analysis result is evaluated step by step in each block, and if the conditions at that stage cannot be satisfied, the process returns to the data input unit 4, resets, and repeats the simulation. Then, the feedback to the data input unit 4 is performed so that the optimal condition is obtained at each stage, so that the optimal control condition is finally obtained from the thermal stress analysis unit 3.

The system having the above-described configuration finally performs an optimal design of the thermal environment in the furnace, and grows a crystal with less distortion. In other words, how to derive conditions for pulling high-quality crystals without breaking the furnace. for that reason,
The temperature distribution in the furnace and the temperature distribution near the crystal are important.

For example, consider whether the crucible shape set in a given container can be adopted or how large the crucible can be, and for that purpose, examine the effect of the shape of the hot zone composed of heaters and heat insulating materials, multi-division Optimize heater arrangement and power ratio. In this case, if the obtained analysis result satisfies the set criterion, the boundary condition is met, and there is no problem in terms of temperature. For example, if the diameter of the crucible is increased by 2 inches, the thickness of the heat insulating material on the outer peripheral portion is reduced by 1 inch. Therefore, when the inside of the furnace is heated by a heater under such a change in the crucible shape, the temperature of the outer peripheral portion becomes abnormal. Check that the furnace does not rise, that is, the safety of the furnace, and determine the position of the crucible, the length of the heater, and the like.

In addition, the shape of the solid-liquid interface and the thermal stress of the crystal can be used as an index for evaluating the conditions for pulling up a high-quality crystal. In order to reduce the number of defects, it is necessary to moderate the temperature gradient in the crystal, and the temperature distribution in the crucible becomes important as such information near the crystal. The temperature distribution in the crucible is obtained based on the temperature distribution in the furnace.

 The analysis in each block will be described in detail.

FIG. 2 is a diagram showing an example of the study model, FIG. 3 is a diagram showing an example of mesh division, FIG. 4 is a diagram showing an example of gas flow in the furnace, and FIG. 5 is a diagram showing an example of temperature distribution in the furnace. FIG. 6 is a diagram showing an example of the temperature correlation between the crucible position and the central axis of the crystal / melt portion;
FIG. 7 is a diagram showing an example of the temperature distribution and the solid-liquid interface shape of the crystal / melt portion, and FIG. 8 is a diagram showing an example of the thermal stress distribution.

In FIG. 2, 11 is an Ar atmosphere, 12 is a sub upper heater, 13 is graphite, 14 and 22 are adiabatic felts, 15 is a main heater, 16 is a sub lower heater, 17 is a GaAs crystal, 18
Denotes B ₂ O ₃ , 19 denotes a GaAs melt, 20 denotes a crucible, and 21 denotes a crucible support shaft.

In the thermal fluid analysis, the interior of the furnace is divided into meshes as shown in FIG. 3 by the finite element method with respect to the study model shown in FIG.
For example, the length of the heater can be arbitrarily selected. Physical properties of various substances are defined in the mesh, and simulation is performed by changing the heater length, position, crucible position, and crystal growth amount according to the mesh. Then, the crucible shape (diameter), heater length, and furnace internal structure are set in a case, the temperature of the outer wall of the hot zone, the gas inlet flow velocity are given as boundary conditions, and the heater power is input as shown in FIG. The temperature distribution in the furnace is determined by the Ar gas flow and the isotherms 1, 2,... As shown in FIG. From this result, each factor is analyzed, and the possibility of changing the conditions, the possibility of crystal pulling, and the like are evaluated. As described above, if the given boundary condition meets the set criterion, there is no problem in terms of temperature because the boundary condition is met.

Since the inside of the furnace is under an atmosphere of high-pressure (for example, 10 atm) Ar (argon) gas, when the inside of the furnace is heated by a heater, the gas is convected and the flow becomes important. Radiation at high temperature is effective to the fourth power when the shape of the heater is changed. Therefore, the temperature distribution is solved by coupling taking into consideration the gas flow and radiation. As described above, the thermal fluid analysis is for analyzing a macroscopic temperature distribution. For heat transfer under high temperature and high pressure, analysis using radiation and gas is important. In the gas part, for example, a thermally coupled Navier-St
The calculation can be performed by applying the okes equation and applying the heat conduction equation to the structural member in consideration of radiation between the surfaces. At this time, the crystal pulling becomes difficult unless the design and setting of the multi-stage heater and the setting of the crystal pulling position are improved. The solid-liquid interface between the crystal and the melt is
It actually moves according to the heat balance state, but at this stage, it is assumed that the mesh is flat as shown in the figure.

In the solid-liquid interface shape analysis, the temperature data obtained by the thermo-fluid analysis is taken over as a boundary condition, the thermal analysis of the crystal / melt portion is performed, and the solid-liquid interface shape is analyzed from the temperature distribution as shown in FIG. The solid-liquid interface shape is obtained as an isotherm of the melting point temperature from the temperature distribution, and the heater power is set so as to have a desired solid-liquid interface shape. In order to stabilize crystal growth, it is necessary that the shape of the flat solid-liquid interface is constant in a time series. However, it is actually preferable that the temperature gradient is moderated to make the shape convex downward. However, when the crystal growth conditions change, the way of conducting heat changes, so that the method of controlling crystal pulling cannot be fixed. For the crystal part, for example, the heat conduction equation is applied, and for the melt part, by applying the thermo-fluid analysis considering the convection in the melt, the solid-liquid interface is given a limiting condition that makes the melting point free. Boundary problems can be solved.

FIG. 6 shows an example of the temperature correlation of the central axis of the crystal / melt when the crucible position is changed using the results of thermal analysis in the furnace. As is clear from this example, the crucible position is changed. Then, the temperature gradient of the crystal part and the temperature of the melt part change. From this, it is understood that it is also important to appropriately set the position of the crucible. In particular, even though the unidirectional solidification is fundamental, in the case of a profile in which the melting point below the solid-liquid interface reverses from the melting point to a temperature lower than the melting point, the system cannot be established, It is necessary to perform such an evaluation. Further, when the power of the upper sub heater, the main heater, and the lower sub heater is changed, the temperature gradient and the temperature distribution also change. FIG. 7 shows the temperature gradient and the temperature distribution by isothermal lines based on the temperature difference 1, 2,... From the solid-liquid interface with the solid-liquid interface being 0. In this way, the solid-liquid interface shape analysis shows whether the temperature gradient of the crystal in the vicinity of the solid-liquid interface is suitable for high-quality crystal pulling, the state of the melt part, and the other optimal solid-liquid interface. It can analyze and evaluate the uniform growth of shapes and crystals.

In the thermal stress analysis, thermal stress analysis is performed using the temperature distribution in the solid-liquid interface shape obtained by the solid-liquid interface shape analysis. On the other hand, in order to obtain a high-quality crystal, a uniform state with small stress is desirable. In other words, the ultimate goal is to produce a crystal with less distortion. For this purpose, it is necessary to perform thermal stress analysis based on the shear adiabatic stress distribution and thermal history as shown in FIG. 8, for example. It is necessary to reduce the stress. In this analysis, many studies have been conducted on the relationship between thermal stress and crystal defects.
It has been established as one of the evaluations to obtain high quality crystals.

As described above, if the analysis can be sequentially performed while dividing the results into three blocks and organically passing the results to the next block, the effect of the multi-split heater can be obtained by repeating the simulation up to that step by evaluating the analysis results in the middle. Can be easily analyzed. For example, looking at the temperature distribution in the crystal / melt part, it is possible to analyze how the distribution of heater power changes when the heater power is changed by repeatedly performing simulations with different heater power settings. it can.
Therefore, the solid-liquid interface should be 15001 (K)
In such a case, the distribution of the heater power to obtain the melting point of 1511 ゜ (K) can be obtained by repeating the process by returning the block unit, so that it can be easily performed without wasted time. In addition, when the power of the main heater is changed, when the power of the sub-upper heater is changed, and when the power of the sub-lower heater is changed, the change of the temperature distribution according to each can be analyzed in a relatively short time. .

The above means can also be used for controlling the single crystal pulling furnace as it is. In this case, since the optimum design of the furnace has already been made by the above analysis, the conditions in the furnace, boundary conditions,
Among the heater power conditions, the furnace shape and boundary conditions are fixed data, and the crystal pulling speed, the rotation speed of the crucible and the crystal, the heater power, and the like are variables associated with crystal pulling. Therefore, these variables may be corrected sequentially with the growth of crystal pulling.

Therefore, the overall configuration for controlling the single crystal pulling furnace is basically a condition setting section for setting the conditions in the furnace, boundary conditions, heater power, and operating conditions of the single crystal pulling furnace in the furnace. A thermo-fluid analysis unit 1, a solid-liquid interface shape analysis, wherein a monitor unit for monitoring the temperature and a control unit for controlling the single crystal pulling up of the furnace while monitoring the condition by using the condition set by the condition setting unit as a control signal. Section 2 and thermal stress analysis section 3 may be connected.

In the thermal fluid analysis unit 1, the solid-liquid interface shape analysis unit 2, and the thermal stress analysis unit 3, criteria for performing evaluation and judgment at each stage are set. For example, a reference value is set for the temperature at a specific position in the furnace, the position of the solid-liquid interface, and the like, and the evaluation and judgment are performed by comparing the analysis results with the thermo-fluid analysis unit 1 and the solid-liquid interface shape analysis unit 2 to determine a condition. The conditions of the setting unit are sequentially corrected. For example, the position of the crucible is changed according to the crystal pulling speed. As described above, for example, the monitor unit detects the weight of a crystal using a load cell, detects the temperature of each point using a thermocouple, detects the rotation speed of a crucible and a crystal using an encoder attached to a motor and a motor drive current, and the like. Do. The control section feeds back the detection data in the monitor section so as to satisfy the conditions set in the condition setting section, and controls the heater power and the rotation of the crucible and the crystal. By configuring such a control system, a high-quality crystal analyzed by the thermal fluid analysis unit 1, the solid-liquid interface shape analysis unit 2, and the thermal stress analysis unit 3 can be efficiently created. In addition, the detection data of the monitor section is also fed back to each analysis section,
Of course, the analysis result may be used as basic data for evaluation and judgment.

In addition, the present invention is characterized in that the thermofluid analysis, the solid-liquid interface shape analysis, and the thermal stress analysis are independent analysis units, and these are connected in a cascade manner. It is a thing. Therefore, the present invention is not limited to the above embodiment, and various modifications are possible. For example, in the above embodiment, the results of the thermal fluid analysis, the solid-liquid interface shape analysis, and the evaluation of the thermal stress analysis were evaluated, simulation conditions were changed, and optimal conditions for the furnace thermal environment were set. For this evaluation, it is configured to set the parameters of the evaluation and the simulation condition change in each analysis unit, or to add a learning function to automatically perform the feedback so as to automatically change the setting input of the simulation condition. Is also good. Further, it may be incorporated in a control system of a single crystal pulling furnace.

Since the temperature of the heater and the pressure in the furnace are constantly monitored, this data and the detection data of the crystal weight by the load cell described above are inputted to the present system and fed back to the temperature control of the heater. Is also good. In this case, the change in the shape of the solid-liquid interface affects the diameter of the grown crystal. That is, when the solid-liquid interface shape is convex downward, the buoyancy is increased, and the crystal weight detected by the load cell is apparently light. Therefore, the diameter of the solid-liquid interface may be determined from the change in the weight of the crystal, and the diameter of the crystal may be calculated and fed back to the control of the temperature of the heater to correct the diameter.

〔The invention's effect〕

As is apparent from the above description, according to the present invention, instead of performing each analysis individually, each block is divided and evaluated at each block stage, and the analysis result is taken over and the next block is analyzed. With such a configuration, the optimum condition analysis of the single crystal pulling furnace can be efficiently and accurately performed by the respective block evaluations. In addition, it has become possible to easily estimate a desired thermal environment by using a multi-segment heater which has been difficult in the past. This makes it possible to easily optimize the conflicting objectives of high production and high quality of the GaAs single crystal, change the crucible shape (scale up), and change the heater shape.

In addition, when the thermal fluid analysis, the solid-liquid interface shape analysis, and the thermal stress analysis are connected to form a single unit, the apparatus becomes large in size and the processing takes time, so that the analysis flexibility is also increased. However, when divided into three as in the present invention, each means can be configured compactly, and evaluation can be performed at each processing stage, and analysis can be corrected and changed quickly based on the results. Therefore, highly flexible analysis can be performed without waste, and it can be directly used for a control system of a single crystal pulling furnace.

[Brief description of the drawings]

FIG. 1 is a view for explaining one embodiment of an optimum condition analysis method for a single crystal pulling furnace according to the present invention, FIG. 2 is a view showing an example of a study model, and FIG. 3 is an example of mesh division. Fig. 4, Fig. 4 shows an example of the furnace gas flow, Fig. 5 shows an example of the furnace temperature distribution, and Fig. 6 shows an example of the temperature correlation between the crucible position and the crystal / melt central axis. FIG. 7 is a diagram showing an example of the temperature distribution of the crystal / melt portion and the shape of the solid-liquid interface, FIG. 8 is a diagram showing an example of the shear stress distribution in the cross section of the crystal, and FIG. It is a figure which shows the outline of a crystal pulling apparatus. 1 ... thermal fluid analysis part 2 ... solid-liquid interface shape analysis part 3 ...
... thermal stress analysis section, 4 ... data input section.

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Fumikazu Yajima 1000 Higashi-Anana-cho, Ushiku-shi, Ibaraki Mitsubishi Tsunami Corporation (56) References 58) Field surveyed (Int.Cl. ⁶ , DB name) C30B 1/00-35/00 H01L 21/00-23/00

Claims (2)

(57) [Claims] 1. An optimal condition analysis method for a single crystal pulling furnace in which a raw material is put into a crucible and heated by a heater under a high-pressure gas atmosphere to a high temperature near its melting point to pull up and grow a single crystal. Thermo-fluid analysis means for determining the temperature distribution in the furnace in consideration of gas flow and radiation based on the settings of the internal shape, boundary conditions, heater power, etc., and crystal / melt using the temperature distribution in the furnace as boundary conditions Solid-liquid interface shape analysis means for performing a thermal analysis of the crystal part and the solid-liquid interface shape in the crystal / melt part, and a thermal stress analysis for obtaining the thermal stress in the crystal part from the temperature distribution of the crystal part with the solid-liquid interface shape Means for transferring the temperature distribution in the furnace from the thermal fluid analysis means to the solid-liquid interface shape analysis means, and the solid-liquid interface shape analysis means converts the solid-liquid interface shape and the temperature distribution of the crystal part to the heat distribution. Pull to stress analysis means An optimal condition analysis method for a single crystal pulling furnace characterized by being connected. 2. A method for controlling a single crystal pulling furnace in which a raw material is put into a crucible and heated in a high-pressure gas atmosphere with a heater to a high temperature around its melting point to pull up and grow a single crystal. Setting means for setting the shape, boundary conditions, heater power, etc., and the temperature distribution in the furnace in consideration of gas flow and radiation based on the shape, boundary conditions, heater power in the furnace set by the setting means. A thermal-fluid analysis means for obtaining; a solid-liquid interface shape analysis means for performing a thermal analysis of a crystal / melt portion with the temperature distribution in the furnace as a boundary condition to obtain a solid-liquid interface shape in the crystal / melt portion; Thermal stress analysis means for determining the thermal stress in the crystal part from the liquid interface shape and the temperature distribution of the crystal part; monitor means for monitoring the operating condition of pulling the single crystal in the furnace; and controlling the conditions set by the setting means Monitor as a signal Control means for controlling the pulling of the single crystal of the furnace while monitoring by means of a step, wherein the thermal fluid analysis means takes over the temperature distribution in the furnace to the solid-liquid interface shape analysis means, and the solid-liquid interface shape analysis Means to carry over the solid-liquid interface shape and the temperature distribution of the crystal part to the thermal stress analysis means, and evaluate the analysis results in the thermal fluid analysis means, the solid-liquid interface shape analysis means, and the thermal stress analysis means. A method for controlling a single crystal pulling furnace, wherein a setting condition of a setting means is sequentially corrected.