JP5854757B2 - Single crystal ingot diameter control method - Google Patents

Description

  The present invention relates to a method for controlling the diameter of a single crystal ingot to be produced in the production of a single crystal for obtaining a crystal such as silicon used as a semiconductor material or solar cell, or silicon used as another industrial material.

  The Czochralski method (CZ method) is widely used as a silicon crystal growth method for semiconductors. In this CZ method, in order to efficiently produce a single crystal ingot (hereinafter referred to as an ingot), there are a number of factors such as the heating temperature and temperature distribution of the molten raw material, the pulling speed of the seed crystal, the crucible rising speed, and the rotating speed. It is necessary to set manufacturing conditions (parameters) appropriately. Therefore, in the CZ method, so-called feedback control for adjusting manufacturing conditions according to the quality of manufactured ingots, and program control (feed forward control) for setting temperature conditions, changing conditions, and timings in advance are performed. I came.

  As a system for performing such feedback control and feedforward control of the CZ method, for example, there is a group management system for single crystal pulling apparatus disclosed in Japanese Patent Laid-Open No. 63-2888. In this county management system, a single crystal (ingot) manufactured by each pulling device is provided with an identification mark for each pulling device that manufactures it, and along with the slide of the slide table on which the ingot is placed, The diameter of the ingot is measured over its entire length and the weight is measured. These diameters and weights are sent to the control means of the lifting device corresponding to the identification mark, and the control parameters are corrected by the control means.

Japanese Patent Laid-Open No. 63-2888

  While the CZ method requires a large amount of energy for melting raw materials, the manufactured ingot is often processed into extremely small electronic parts because it is used for precision parts and the like. When the diameter of the manufactured ingot falls below the target value, the portion cannot be used and is cut and removed as a defect. Therefore, even if the diameter of the manufactured ingot is slightly different, the manufacturing efficiency is greatly affected. However, in the conventional feedback control and feedforward control including the group management system, the diameter control accuracy of the manufactured ingot is not sufficient. For example, in the case of the group management system described above, the measurement error caused by the deviation of the focal length or the perspective magnification is caused by the fact that the crystal whose diameter is to be measured is placed obliquely due to the taper, and the perspective magnification is shifted. It was a problem.

  Then, an object of this invention is to provide the single crystal diameter control method which can control the diameter of the ingot manufactured with high precision.

  In the single crystal ingot diameter control method according to the present invention, the diameter of the ingot pulled up vertically in the furnace is measured through the in-furnace state visualizing window, and the manufacturing conditions are corrected based on the obtained diameter data.

  The in-furnace state visualizing window is provided at a position and an angle at which the molten surface of the molten raw material in the crucible can be visually recognized, and the first diameter data is obtained at the molten surface, and the first position is vertically spaced from the vicinity of the surface. Two-diameter data may be obtained, and the correction result of the manufacturing condition based on the first diameter data may be verified based on the second diameter data.

  Two imaging means are arranged outside the in-furnace state viewing window, one of the two contours extending parallel to the longitudinal direction of the ingot is photographed by one of the imaging means, and the other contour line is photographed The diameter of the image may be obtained on the basis of the interval between the imaging means at the measurement position by photographing with the other of the means, horizontally moving the imaging means to the measurement position where the outline appears at a predetermined position in the video.

  The diameter data may be obtained at a timing when no ridgeline appears in one rotation cycle centered on the axis of the ingot, starting from when the diameter data obtained based on the imaging data exceeds a predetermined threshold. The timing at which the ridge line does not appear is when the ridge lines are equally spaced in the circumferential direction, for example, <100> crystal or <111> crystal, and one ingot rotation period, diameter data exceeds a predetermined threshold. The period divided by the number of ridge lines of the crystal axis is used as a basic period, and the timing that does not overlap with this basic period can be obtained. The number of ridge lines of the crystal axis is 4 for the <100> crystal and 3 for the <111> crystal. Further, in a crystal in which the ridge lines are not equally spaced in the circumferential direction, for example, a <511> crystal or a <211> crystal, a timing that does not overlap with the ridge line can be obtained by dividing one rotation period together with the ridge line structure.

  You may link the imaging timing of the said imaging means with the raising length of the said ingot. Moreover, you may include a seed crystal, the shoulder part of the said ingot, and the terminal part of the said ingot in the imaging object site | part of the said imaging means.

  At the time of opening the furnace, two light emitting means are disposed inside the furnace, light is emitted from each of the light emitting means to the outside of the furnace through the in-furnace state viewing window, and the light from the light emitting means. One of the images is photographed by one of the imaging means, and the other light beam is photographed by the other of the imaging means. Then, the imaging unit is horizontally moved to a measurement position where the light beam appears at a predetermined position in the image of the imaging unit, the interval between the imaging unit at the measurement position is compared with the interval between the light emitting units, and the measurement unit at the measurement position A numerical value calculated based on the interval of the imaging means may be calibrated.

  According to the single crystal diameter control method according to the present invention, highly accurate diameter data can be calculated by utilizing the stable suspended state of the ingot. Since the manufacturing conditions are corrected based on highly accurate diameter data, the diameter of the ingot to be manufactured can be controlled with high accuracy. And since the diameter of an ingot can be calculated before taking out an ingot outside a furnace, the precision of correction | amendment can be improved, without making the furnace to respond | correspond. Thereby, the outer diameter grinding loss of the crystal can be reduced, and the raw material yield can be improved with the loss as the crystal length.

  In addition, when the manufacturing conditions are corrected based on the first diameter data obtained on the molten surface, the result is the position at a position vertically above the molten surface after the ingot has been pulled up by a predetermined length. It will be reflected in the part. Therefore, by verifying the correction result of the manufacturing condition based on the second diameter data obtained at that position, it is possible to confirm the suitability of the condition correction even during the manufacture of the ingot, and more precise diameter control is possible. . In addition, since the temperature of the ingot is different between the molten surface and the position where the vertical information is spaced from the molten surface, the accuracy can be further improved by correcting the diameter data with the volume expansion coefficient corresponding to the temperature. it can. As a result, grinding loss in the crystal outer shape processing step can be reduced, and productivity can be greatly improved. In some cases, the grinding process can be omitted. In this case, since the grinding process can be omitted and the lifetime can be prevented from being lowered due to peripheral processing distortion caused by grinding, the conversion efficiency of the solar battery cell can be improved.

  Furthermore, the accuracy of diameter data can be improved by using two imaging means.

  Furthermore, the period of time for which the ingot makes one rotation around the axis is divided into four periods, and the diameter data is obtained by obtaining the diameter data at a timing that does not overlap with the basic period, thereby eliminating the measurement error due to the ridgeline of the ingot. The accuracy can be further improved.

  Furthermore, by associating the imaging timing of the imaging means with the pulling length of the ingot, the diameter of the imaged part and the manufacturing conditions of the part can be related accurately and easily, and the manufacturing conditions can be more accurately determined. It becomes possible to control. Furthermore, by including the seed crystal (seed) and the shoulder of the ingot in the imaging target part of the imaging means, it is possible to accurately control the diameter of the seed crystal (seed) and the shape of the shoulder (shoulder). , Quality, crystal thermal history, and dislocation-free crystal growth. Furthermore, by including the end portion (tail) of the ingot in the imaging target portion of the imaging means, it becomes possible to accurately control the shape of the end portion (tail) of the ingot, and the tail and the residual liquid in the crucible are fixed. Etc. can be prevented.

  Furthermore, high accuracy of the diameter data obtained can be maintained by simple calibration using two light emitting means arranged inside the furnace at the time of opening the furnace.

It is a functional block diagram of the system which employ | adopted the single crystal ingot diameter control method which concerns on this invention. It is a front view which shows the general view of the single crystal pulling apparatus controlled by the system. The concept for obtaining diameter data at the timing when the ridge line does not appear is shown, (a) is a cross-sectional view showing the ridge line of the ingot, and (b) is a time chart of diameter data including the ridge line. It is a side view which shows the concept which links an imaging timing with pulling length. It is a figure which shows typically the calibration principle of the imaging means used for a diameter measurement. It is a side view which shows the concept which measures diameter data in the position which opened the space | interval from the molten surface of the fusion | melting raw material, and the perpendicular | vertical upper direction from there. It is a side view which shows the concept which associates an imaging timing with pulling-up length, when a furnace state visual recognition window is provided in the pull chamber.

An embodiment of a single crystal ingot diameter control method according to the present invention will be described with reference to FIGS.
FIG. 1 is a functional block diagram of a system that employs a single crystal ingot diameter control method according to the present invention. In this system, the controller 3 controls the heating temperature of the crucible 21, the pulling speed of the seed crystal, the rotational speed, and the like in the single crystal pulling apparatus 2 that manufactures the ingot 1. The control device 3 is a known device including a data storage, an arithmetic processing function, and a signal input / output interface, such as a PLC and a personal computer.

  A wire hoisting machine (not shown) is provided on the upper portion of the vacuum furnace 23 constituting the single crystal pulling apparatus 2, and a seed crystal is fixed to the end of the wire, and the single crystal ingot is pulled up by winding the wire. . And the diameter can be measured using the suspended state at the time of the raising.

  The vacuum furnace 22 is provided with a furnace state visualizing window 23 in the top chamber and the pull chamber. The in-furnace state visualizing window 23 provided in the top chamber portion is provided at a position and an angle at which the molten surface of the molten raw material in the crucible 21 can be viewed, and the inside of the vacuum furnace 22 is passed through the in-furnace state visualizing window 23 to the outside. Two imaging means 11 for imaging are arranged.

  The imaging means 11 is attached to a guide rail 12 provided with a thread. Then, by rotating the guide rail 12, the imaging means 11 moves horizontally along the guide rail 12. The rotational direction of the ribs provided on the guide rail 12 is opposite to the center of the guide rail 12. For this reason, when the guide rail 12 is rotated, the two imaging means 11 move in opposite directions.

  A pulse motor 13 is attached to one end of the guide rail 12, and the rotation operation of the guide rail 12 is performed via the pulse motor 13. However, there is no restriction | limiting in the method of operating rotation of the guide rail 12, For example, it is good also as what operates manually by a vernier.

  The imaging means 11 and the pulse motor 13 are connected to the measurement control device 14. The measurement control device 14 is also a known device including a data storage, an arithmetic processing function, and a signal input / output interface, such as a PLC and a personal computer, like the control device 3. The measurement control device 14 analyzes the image data from the imaging unit 11 by a known method, and outputs a rotation instruction signal to the pulse motor 13 when the contour of the ingot 1 is not reflected at a predetermined position. Then, the pulse motor 13 is operated until the contour of the ingot 1 reaches a predetermined position in the image data from the imaging unit 11. When the contour of the ingot 1 reaches a predetermined position in the image data from the imaging means 11, the interval between the imaging means 11 is calculated from the operation content of the pulse motor 13, that is, the number of revolutions, and the ingot 1 is further based on the interval between the imaging means. The diameter of is calculated.

  In the measurement control device 14, when calculating the diameter of the ingot 1, an error due to the ridgeline appearing in the ingot 1 is not caused. As shown in FIG. 3A, a ridge line 10 extending in the axial direction according to the number of ridge lines of the crystal axis appears on the outer peripheral surface of the ingot 1, and this ridge line 10 is in the ingot 1 in the image data of the imaging unit 11. The diameter R2 of the ingot 1 calculated when it is determined that the contour of the ingot 1 is larger than the actual diameter R1, resulting in a measurement error. Therefore, in the measurement control device 14, diameter data is obtained at a timing at which the ridge line 10 does not appear in one rotation cycle centered on the axis of the ingot 1.

  If it is a <100> crystal | crystallization, the ridgeline 10 will become equal intervals in the circumferential direction, and will appear in the position shown to Fig.3 (a). At this time, the transition of the diameter data in one rotation period of the ingot 1 has a peak at a constant interval as shown in FIG. Since this peak is due to the appearance of the ridge line 10, one rotation of the ingot 1 starts when the diameter data obtained based on the imaging data exceeds a predetermined threshold value Dt (T1 in FIG. 3B). A period obtained by dividing the period into four is a basic period Tb, and diameter data is obtained at a timing that does not overlap with the basic period Tb. In the <111> crystal, since the number of ridge lines of the crystal axis is 3, the basic period Tb may be obtained by dividing one rotation period of the ingot 1 into three equal parts. Further, in a crystal in which the ridge lines are not equally spaced in the circumferential direction, for example, a <511> crystal or a <211> crystal, a timing that does not overlap with the ridge line can be obtained by dividing one rotation period of the ingot 1 along with the ridge line structure. Can do.

  Pull-up length data is delivered from the control device 3 to the measurement control device 14. The pulling length data is a moving distance of a predetermined part in the vertical direction (Z direction). If it demonstrates referring FIG. 4, it will become the distance which the edge part Z0 of the seed crystal moved to the direction of the white arrow line from the horizontal position Z of the imaging | photography point of the imaging means 11. FIG. For example, at the beginning of the tail, the pulling length data Z is Z3. The diameter data calculated in the measurement control device 14 is associated with the pulling length Z of the part that is the calculation target, and is sent to the control device 3. The control device 3 corrects the manufacturing conditions such as the heating temperature of the crucible 21, the pulling speed of the seed crystal, and the rotational speed based on the diameter data and the manufacturing conditions related to the portion (Z portion) associated with the diameter data. And reflected in subsequent manufacturing. Note that the imaging target range is from Z0 to Z4 in FIG. 4, and the diameters of the seed crystal, the shoulder of the ingot, and the tail of the ingot are also calculated.

  The accuracy of measurement can be maintained by calibrating the correlation between the interval of the imaging means 11 and the diameter of the ingot 1 when the single crystal pulling apparatus 2 is inspected. In the calibration, first, two light emitting means 15 are arranged inside the vacuum furnace 2 as shown in FIG. Similar to the imaging means 11, these two light emitting means 15 are also attached to a guide rail 12 provided with a thread, and are moved horizontally along the guide rail 12 by rotating the guide rail 12. Yes.

  Next, the position of the light emitting means 15 is adjusted so that the interval between the light emitting means 15 is equal to the diameter of the ingot 1 to be manufactured. Subsequently, light is emitted from each of the light emitting means 15 to the outside of the vacuum furnace 22 through the in-furnace state visualizing window 23, and one of the light beams L from the light emitting means 15 is photographed by one of the imaging means 11, and the other The light beam L is imaged by the other of the imaging means 11. Then, the imaging unit 11 is horizontally moved to a measurement position where the light beam L appears at a predetermined position in the image of the imaging unit 11. Here, by comparing the interval of the imaging unit 11 at the measurement position with the interval of the light emitting unit 15, the numerical value calculated based on the interval of the imaging unit 11 at the measurement position can be calibrated.

  The diameter data of the ingot 1 may be measured at the molten surface of the molten raw material and at a position vertically spaced from the molten surface. In this case, the suitability of the condition correction is confirmed during the manufacture of the ingot 1, and more precise Diameter control is possible. The concept of measuring the diameter at these two positions is shown in FIG. In FIG. 6, substantially the same parts as those of the embodiment shown in FIGS.

  In the embodiment shown in FIG. 6, two sets of imaging means 11 (hereinafter, one is 11 a and the other is 11 b) are arranged outside the in-furnace state visualizing window 23. Of these two sets of imaging means 11a and 11b, one of them 11a images the molten surface (position Za) of the molten raw material, and the other 11b images the position Zb spaced vertically upward from the molten surface of the molten raw material. It has become. In the measurement control device 14, the first diameter data at the melt surface is calculated based on the measurement data of the imaging means 11a, and the second diameter data at the position Zb is calculated based on the measurement data of the imaging means 11b. Delivered to 3. The control device 3 corrects the manufacturing conditions based on the first diameter data. At this time, the controller 3 determines whether the correction conditions are appropriate based on the second diameter data, and performs correction based on the determination. That is, the manufacturing conditions at the position Zb of the ingot 1 (manufacturing conditions that are traced back by the pulling time from the melt surface Za to the position Zb) are confirmed, and for example, it is determined that the predetermined parameter adjustment amount at that time is too large. Performs correction such as slightly reducing the parameter adjustment amount. As described above, even when the ingot 1 is manufactured, it is possible to perform more precise diameter control by confirming whether or not the condition correction is appropriate.

  Since the temperature of the ingot 1 is different between the molten surface Za and the position Zb spaced from the molten surface Za to the vertical information, the measurement control device 14 corrects the diameter data with the volume expansion coefficient corresponding to the temperature. ing. In addition, since the control device 3 stores the manufacturing conditions of the ingot 1 over the entire manufacturing period, it is possible to retrospectively check the manufacturing conditions at the position Zb of the ingot 1 based on the second diameter data. It is possible.

  The two sets of imaging means 11a and 11b can obtain the first diameter data at the molten surface (position Za) of the molten raw material and the second diameter data at a position Zb spaced vertically upward from the molten surface of the molten raw material. If possible, this may be replaced with only the single imaging means 11. For example, scanning is performed from the position Za to the position Zb while moving by the moving table, and data for each position may be collected and controlled while the table is moving. Moreover, it is good also as a thing with a wide imaging range which can obtain the image data of the position Za and the position Zb as one data. In any case, it is necessary to store in advance the temperature correction at each position in the crystal growth direction and reflect the volume expansion coefficient in the measurement data.

  In calculating the diameter of the ingot 1, an in-furnace state visualizing window 23 provided on the side surface of the pull chamber can also be used. FIG. 5 shows the measurement principle when the diameter is calculated using the in-furnace state viewing window provided on the side surface of the pull chamber. The pulling length data in this case is the distance that the end portion Z0 of the seed crystal has moved in the direction of the white arrow from the position Z where it overlaps with the arrangement of the imaging means 30. For example, at the beginning of the tail, the pull-up length data Z is Z3, but this data value is the same as in the embodiment shown in FIGS. In FIG. 7, substantially the same parts as those in the embodiment shown in FIGS.

DESCRIPTION OF SYMBOLS 1 Ingot 2 Single crystal pulling apparatus 3 Control apparatus 10 Edge line 11, 11a, 11b Imaging means 12 Guide rail 13 Pulse motor 14 Measurement control apparatus 15 Light emission means 21 Crucible 22 Vacuum furnace 23 In-furnace state visualizing window

Claims (5)

At the time of opening the furnace, two light emitting means are arranged inside the furnace, and light is emitted from each of the light emitting means to the outside of the furnace through the in-furnace state visualizing window, and one of the light rays from the light emitting means Is taken by one of the two imaging means arranged outside the in-furnace state viewing window, the other light ray is taken by the other of the imaging means, and the measurement position where the light ray appears in the image of the imaging means The imaging means is moved horizontally, the interval of the imaging means at the measurement position is compared with the interval of the light emitting means, and the correlation between the interval of the imaging means and the diameter of the ingot is based on the interval of the imaging means at the measurement position. The process of calibrating
One of the two contours extending in parallel to the longitudinal direction of the ingot is photographed with one of the imaging means, and the other contour line is photographed with the other of the imaging means with respect to the diameter of the ingot pulled up vertically. And horizontally moving the imaging means to a measurement position where the contour appears in the video, and obtaining diameter data based on the interval of the imaging means at the measurement position;
Correcting the manufacturing conditions based on the obtained diameter data ;
A method for controlling the diameter of a single crystal ingot , comprising :   The in-furnace state visualizing window is provided at a position and an angle at which the molten surface of the molten raw material in the crucible can be visually confirmed, and the first diameter data is obtained at the molten surface, and at a position spaced vertically above the molten surface. 2. The single crystal ingot diameter control method according to claim 1, wherein two-diameter data is obtained, and the correction result of the manufacturing condition based on the first diameter data is verified based on the second diameter data. As a starting point when the diameter data obtained based on the imaging data exceeds a predetermined threshold, in one rotation cycle about the axis of the ingot, according to claim 1 to obtain the diameter data at the timing when the edge line does not appear or 2. The method for controlling the diameter of a single crystal ingot according to 2 . The single crystal ingot diameter control method according to any one of claims 1 to 3, wherein an imaging timing of the imaging means is associated with a pulling length of the ingot. Wherein the imaging site of the imaging means, and the shoulder of the seed crystal and the ingot, single crystal ingot diameter control method according to any one of claims 1 to 4 including the terminal portion of the ingot.

Images