Method of growing single crystal and apparatus therefor

Abstract

To make shading work efficient by improving prediction accuracy for a crucible bottom temperature at which the shading is started.SOLUTION: A method of growing single crystal comprises: an acquisition process 11 of acquiring a crucible bottom temperature in shading operation for each number of times of growth of the single crystal; a prediction process 12 of predicting a crucible bottom temperature at which the shading operation is started as a predicted crucible bottom temperature; and an implementation process 13 of starting the shading operation at the timing when the predicted crucible bottom temperature is reached, wherein the prediction process 12 includes: a first process 12a of selecting a group of most recent n numbers Xn of times of growth of single crystal after the number X of times of growth and a group of crucible bottom temperatures Yn in the shading operation; a second process 12b of finding relational expression (I) Y=aX+b between the number X of times of growth of single crystal and a crucible bottom temperature Y in the shading operation from the selected information groups by a least square method; and a third process 12c of determining, as a predicted crucible bottom temperature Yp, the crucible bottom temperature Y calculated by the number X=p of times of growth to be performed from now in the number X of times of growth of single crystal of the relational expression (I).SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method and apparatus for growing a single crystal, and more particularly to a method and apparatus for growing a single crystal that is effective in simplifying the seeding operation, which is important in growing a single crystal. [Background technology]

[0002] In general, oxide single crystals such as lithium tantalate (LiTaO3: hereafter abbreviated as LT when necessary) or lithium niobate (LiNbO3: hereafter abbreviated as LN when necessary) are artificial ferroelectric crystals, and single crystal substrates processed from these single crystals are used as materials for surface acoustic wave elements (SAW filters) that remove electrical signal noise, mainly in mobile communication devices. Oxide single crystals such as LT and LN, which are used as materials for SAW filters, are mainly produced industrially by a rotational pulling method such as the Czochralski method (hereinafter abbreviated as the Cz method) (see, for example, Patent Document 1).

[0003] To grow a single crystal by this type of rotational pulling method, first, the raw material for the single crystal is charged into a crucible and filled therein, and then the raw material in the crucible is heated to a temperature above its melting point to melt it. Next, a single crystal piece that will become the seed crystal for the single crystal is brought into contact with the center of the molten raw material, and the raw material melt is gradually solidified while the seed crystal is rotated and raised. By continuously performing these operations, a substantially cylindrical single crystal can be grown. When growing a single crystal by this type of rotational pulling method, optimizing the seeding operation of bringing a seed crystal into contact with the raw material melt is an important factor that determines the quality of the grown single crystal. To perform an optimal seeding operation, it is necessary to bring the seed crystal into contact with the raw material melt at an appropriate temperature and to stabilize the meniscus state of the crystal grown at the tip of the seed crystal.

[0004] Furthermore, when growing single crystals, a method of repeatedly growing single crystals using a crucible in the same furnace is often adopted. However, when single crystals are repeatedly grown using a crucible in the same furnace, the apparent temperature at the bottom of the crucible tends to decrease due to deterioration of the refractory material (which functions as a heat insulator, etc.) that makes up the furnace body and deterioration of the thermocouple built into the bottom of the crucible. In contrast, in the past, during seeding operations, skilled workers would observe the conditions inside the furnace (the meniscus state of the crystal being grown at the tip of the seed crystal) through an observation window in the growth furnace and rely on experience and intuition to keep the conditions inside the furnace constant, so the surface temperature of the raw material melt was kept approximately constant even after repeated growth operations. Here, if the temperature at the bottom of the crucible during the seeding operation is referred to as the "seeding temperature," and the temperature at the bottom of the crucible when the seeding operation is performed at the appropriate timing is referred to as the "appropriate seeding temperature," the relationship between the "surface temperature of the raw material melt" and the "seeding temperature" during the seeding operation will shift with the number of growth operations. Therefore, even if the temperature at the bottom of the crucible (seeding temperature) is kept constant, the surface temperature of the raw material melt varies with each growth cycle, so the operator must adjust this by intuition, but this adjustment tends to vary between operators. As a result, there is a tendency for inconsistency to occur in determining the "appropriate seeding temperature," and the success rate of single crystal growth was only around 80%, which was not sufficient.

[0005] Furthermore, when single crystals are repeatedly grown continuously using the same crucible in the same furnace, seeding operations are sometimes performed by referring to the most recent "optimal seeding temperature" data without observing the state inside the furnace for each growth. However, when single crystals are repeatedly grown using the same crucible in the same furnace, the apparent temperature at the bottom of the crucible decreases due to deterioration of the refractory material that makes up the furnace body and the thermocouple built into the bottom of the crucible, raising concerns that seeding operations based on the most recent "optimal seeding temperature" data may lack accuracy.

[0006] As a prior art technique for solving such problems, a seeding operation method described in Patent Document 2 has already been proposed. Patent Document 2 discloses a method for growing oxide single crystals, which includes the following steps: a first step of repeating single crystal growth multiple times using a crucible in the same furnace, measuring the crucible bottom temperature Xn during raw material melting and the crucible bottom temperature Yn during the seeding operation for each growth, and determining the relational expression (Y = aX + b) between the crucible bottom temperature X during raw material melting and the crucible bottom temperature Y during the seeding operation from the data set of the measured Xn and Yn; a second step of measuring the crucible bottom temperature Xp during raw material melting and substituting the temperature Xp into the relational expression (Y = aX + b) to calculate the crucible bottom temperature Yp during the seeding operation; and a third step of performing the seeding operation when the crucible bottom temperature reaches the determined crucible bottom temperature Yp. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Japanese Patent Application Laid-Open No. 10-338596 (Embodiment of the Invention, Figure 4) [Patent Document 2] JP 2017-202955 A (Mode for carrying out the invention, Figure 2) Summary of the Invention [Problem to be solved by the invention]

[0008] The seeding operation method described in the aforementioned Patent Document 2 involves repeatedly growing single crystals using a crucible in the same furnace, focusing on the relationship between the crucible bottom temperature Xn during raw material melting in each previous growth run and the crucible bottom temperature Yn during the seeding operation, and predicting the crucible bottom temperature during the seeding operation in the growth run to be carried out in the future. However, in Patent Document 2, since Xn is the temperature of the crucible bottom when the raw material is melted, not only is the temperature data for Xn prone to variation, but since not only the temperature Yn of the crucible bottom during the seeding operation but also the temperature Xn of the crucible bottom when the raw material is melted is obtained, the number of times temperature data is obtained inevitably increases. Furthermore, when predicting the crucible bottom temperature Yp during the seeding operation in the growth cycle to be carried out, it is not possible to predict the crucible bottom temperature Yp during the seeding operation until the raw material for the single crystal is melted. Furthermore, in the first step of predicting the crucible bottom temperature during the seeding operation, a relational expression was created using all data on the crucible bottom temperature Xn during raw material melting for each previous growth run and the crucible bottom temperature Yn during the seeding operation. Roughly speaking, the decrease in apparent crucible bottom temperature due to the deterioration of the refractory material that makes up the furnace body and the deterioration of the thermocouple is a linear relationship. However, the degree of deterioration of the refractory material that makes up the furnace body and the thermocouple is not necessarily constant, and it was found that there are subtle changes in the tendency for the decrease in apparent crucible bottom temperature depending on the number of growth runs.

[0009] The technical problem that the present invention aims to solve is to provide a method and apparatus for growing single crystals that can improve the accuracy of predicting the crucible bottom temperature at which seeding begins and make the seeding process more efficient. [Means for solving the problem]

[0010] The first technical feature of the present invention is to heat and melt the raw material for single crystal input into the crucible in the furnace, and at the timing when the crucible bottom temperature, which is the temperature at the bottom of the crucible, reaches an appropriate seeding temperature, perform a seeding operation of contacting a seed crystal with the molten raw material obtained by heating and melting the raw material for single crystal, and grow a single crystal by a rotation pulling method of pulling up the seed crystal while rotating it, which is a method for growing a single crystal that repeats multiple times in the same furnace. The method includes an acquisition step of acquiring the crucible bottom temperature during the seeding operation for each number of times of growing the single crystal, a prediction step of predicting, when the number of times of growing the single crystal exceeds a predetermined number of times n, the crucible bottom temperature at which the seeding operation starts as a predicted crucible bottom temperature, and an implementation step of starting the seeding operation at the timing when the crucible bottom temperature reaches the predicted crucible bottom temperature predicted in the prediction step. The prediction step sets the number of times of growing the single crystal repeatedly performed in the same furnace as X, the crucible bottom temperature during the seeding operation as Y, and the predicted crucible bottom temperature Yp at the time of the seeding operation when the number of times of growing the single crystal X = p to be performed next. When the number of times of growing the single crystal X exceeds a predetermined number of times n (n < p), from the information acquired in the acquisition step, a first step of selecting a group Xn of the number of times of growing the single crystal for the most recent past n times and a group Yn of the crucible bottom temperatures during the seeding operation from the number of times of growing the single crystal X, a second step of obtaining a relational expression (I): Y = aX + b between the number of times of growing the single crystal X and the crucible bottom temperature Y during the seeding operation using the least squares method from the group Xn of the number of times of growing the single crystal and the group Yn of the crucible bottom temperatures selected in the first step, and a third step of determining the crucible bottom temperature Y calculated by substituting the number of times of growing the single crystal X = p to be performed next into the relational expression (I): Y = aX + b as the predicted crucible bottom temperature Yp. It is a method for growing a single crystal characterized by having these steps.

[0011] The second technical feature of the present invention is a method for growing a single crystal, which is characterized in that in the method for growing a single crystal having the first technical feature, the value of n is 4 ≤ n ≤ 10. The third technical feature of the present invention is a method for growing a single crystal, which is characterized in that in the method for growing a single crystal having the second technical feature, the value of n is 5 ≤ n ≤ 7. A fourth technical feature of the present invention is a method for growing a single crystal having the first technical feature, characterized in that the acquisition step uses information detected by a temperature detection means incorporated in the bottom of the crucible. A fifth technical feature of the present invention is the method for growing a single crystal having the fourth technical feature, characterized in that the temperature detection means is a type B thermocouple. A sixth technical feature of the present invention is a method for growing a single crystal having the first technical feature, characterized in that the first step includes a correction process for selecting the group of Xn times of single crystal growth over the most recent n times and the group of crucible bottom temperatures Yn during the seeding operation, and if the i-th Yn(i) in the group of crucible bottom temperatures Yn during the seeding operation includes a seeding operation performed using the predicted crucible bottom temperature Yp, excluding the crucible bottom temperature Yn(i) from the data candidates for the group of single crystal growth times Xn if the difference between the crucible bottom temperature Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is outside an allowable range. A seventh technical feature of the present invention is a method for growing a single crystal having the sixth technical feature, characterized in that, in the second step, when the crucible bottom temperature Yn(i) is excluded in the correction process, the relationship formula (I) is determined using a group of the crucible bottom temperatures Yn other than the excluded one. An eighth technical feature of the present invention is a method for growing a single crystal having the sixth technical feature, characterized in that, when the crucible bottom temperature Yn(i) is excluded in the correction process, the second step determines the relational formula (I) using the most recent data from the most recent n data sets as a data candidate different from the excluded data. A ninth technical feature of the present invention is a method for growing a single crystal having the first technical feature, characterized in that the single crystal is either a lithium tantalate single crystal or a lithium niobate single crystal.

[0012] The tenth technical feature of the present invention includes a furnace for blocking external influences, a crucible provided in the furnace into which a raw material for a single crystal is charged, heating means for heating and melting the raw material for a single crystal charged into the crucible to form a raw material melt, a seeding operation for bringing a seed crystal into contact with the raw material melt in the crucible, lifting means for lifting the seed crystal while rotating it to grow a single crystal, temperature detection means provided at the bottom of the crucible for detecting the temperature of the crucible bottom, and control means for controlling the seeding operation and the lifting operation by the lifting means based on the temperature information detected by the temperature detection means. The control means includes acquisition means for acquiring the temperature of the crucible bottom during the seeding operation for each number of single crystal growths by the temperature detection means, prediction means for predicting, as a predicted crucible bottom temperature, the crucible bottom temperature at which the seeding operation starts when the number of single crystal growths exceeds a predetermined number n, and execution means for starting the seeding operation at the timing when the crucible bottom temperature reaches the predicted crucible bottom temperature predicted by the prediction means. The prediction means sets the number of single crystal growths repeatedly performed in the same furnace as X, the temperature of the crucible bottom during the seeding operation as Y, and the predicted crucible bottom temperature during the seeding operation at the number of growths X = p to be performed next as Yp. When the number of growths X exceeds a predetermined number n (n < p), an information selection unit selects a group Xn of the number of single crystal growths for the most recent past n times and a group Yn of the temperature of the crucible bottom during the seeding operation from the information acquired by the acquisition means. A relational expression creation unit obtains a relational expression (I): Y = aX + b for the relationship between the number of single crystal growths X and the temperature of the crucible bottom Y during the seeding operation using the least squares method from the selected group Xn of the number of single crystal growths and the selected group Yn of the temperature of the crucible bottom. An arithmetic unit determines the calculated crucible bottom temperature Y as the predicted crucible bottom temperature Yp by substituting the number of growths X = p to be performed next into the number of single crystal growths X in the relational expression (I): Y = aX + b. It is a single crystal growth apparatus characterized by having these components. An eleventh technical feature of the present invention is a single crystal growth apparatus having the tenth technical feature, characterized in that, when the information selection unit selects the group of Xn times of single crystal growth for the most recent n times and the group of crucible bottom temperatures Yn during the seeding operation, if the ith Yn(i) in the group of crucible bottom temperatures Yn during the seeding operation includes one in which the seeding operation was performed using the predicted crucible bottom temperature Yp, the information selection unit includes a correction processing unit that excludes the i-th Yn(i) in the group of crucible bottom temperatures Yn during the seeding operation from the data candidates for the group of single crystal growth times when the difference between the crucible bottom temperature Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is outside an allowable range. [Effects of the Invention]

[0013] According to the first technical means of the present invention, it is possible to improve the accuracy of predicting the temperature of the bottom of the crucible at which seeding starts, and to improve the efficiency of the seeding work. According to the second technical means of the present invention, it is possible to more easily detect the latest change in the temperature of the bottom of the crucible at which seeding is started, compared to a case where this configuration is not provided. According to the third technical means of the present invention, it is possible to more easily and accurately detect the immediate change in the temperature of the bottom of the crucible at which seeding starts, compared to a case where this configuration is not provided. According to the fourth technical means of the present invention, the temperature of the bottom of the crucible at which seeding starts can be more appropriately obtained than when the temperature of the raw material melt in the crucible is detected. According to the fifth technical means of the present invention, even if the temperature of the crucible bottom at which seeding is started is in a high temperature range, the temperature of the crucible bottom can be appropriately acquired. According to the sixth technical means of the present invention, by selecting appropriate data from the data associated with the most recent n single crystal growths, a more accurate relational equation (I) can be created compared to when the second step does not involve correction processing. According to the seventh technical means of the present invention, the temperature at the bottom of the crucible during the seeding operation can be predicted more accurately than when it is affected by data excluded by the correction process. According to the eighth technical means of the present invention, the crucible bottom temperature during the seeding operation can be predicted more accurately without reducing the number of data used to determine the relational formula (I), compared to when the influence of data excluded by the correction process is present. According to the ninth technical means of the present invention, in growing a lithium tantalate single crystal or a lithium niobate single crystal, the temperature of the bottom of the crucible during the seeding operation can be predicted with high accuracy. According to the tenth technical means of the present invention, it is possible to appropriately realize a single crystal growth method that can improve the prediction accuracy of the crucible bottom temperature at which seeding begins and make the seeding work more efficient. According to the eleventh technical means of the present invention, by selecting appropriate data from the data of the most recent n single crystal growths, it is possible to create a more accurate relational expression (I) than in the case where no correction processing unit is provided, and therefore it is possible to further improve the accuracy of predicting the crucible bottom temperature at which seeding starts. [Brief explanation of the drawings]

[0014] [Figure 1] FIG. 1( a ) is an explanatory diagram showing an outline of an embodiment of a single crystal growth method to which the present invention is applied, and FIG. 1( b ) is an explanatory diagram showing an outline of an embodiment of a single crystal growth apparatus to which the present invention is applied. [Figure 2] 1A and 1B are explanatory diagrams showing the meniscus state of a crystal grown at the tip of a seed crystal by a seeding operation, where (a) shows the case where the seeding temperature is lower than the appropriate temperature, (b) shows the case where the seeding temperature is the appropriate temperature, and (c) shows the case where the seeding temperature is higher than the appropriate temperature. [Figure 3] 1 is an explanatory diagram showing the overall configuration of a single crystal growth apparatus according to a first embodiment. [Figure 4] 4 is a flowchart showing the processing steps during the seeding operation of the single crystal growth apparatus according to the first embodiment. [Figure 5] 5 is a flowchart showing details of the prediction step in FIG. 4. [Figure 6]10 is a flowchart showing details of a prediction step in the processing steps during the seeding operation of the single crystal growth apparatus according to the second embodiment. [Figure 7] FIG. 1 is a graph showing the relationship between the number of times a single crystal is grown and the predicted crucible bottom temperature during the seeding operation when a single crystal is repeatedly grown using the single crystal growth apparatus of Example 1, as well as the relationship between the number of times a single crystal is grown and the crucible bottom temperature during the seeding operation. [Figure 8] FIG. 1 is a graph showing the relationship between the number of times a single crystal is grown and the predicted crucible bottom temperature during the seeding operation when a single crystal is repeatedly grown using the single crystal growth apparatus of Example 2, as well as the relationship between the number of times a single crystal is grown and the crucible bottom temperature during the seeding operation. DETAILED DESCRIPTION OF THE INVENTION

[0015] Overview of the implementation form FIG. 1(a) is an explanatory diagram showing an outline of an embodiment of a method for growing a single crystal to which the present invention is applied. In the figure, the single crystal growth method is a single crystal growth method that repeatedly performs the growth of a single crystal by a rotational pulling method in which a seed crystal S is brought into contact with a raw material melt M obtained by heating and melting a single crystal raw material charged into a crucible in a furnace, and the seed crystal S is pulled up while being rotated, a plurality of times in the same furnace. The method includes an acquisition step 11 of acquiring the crucible bottom temperature at the time of the seeding operation for each number of single crystal growths, a prediction step 12 of predicting, as a predicted crucible bottom temperature, the crucible bottom temperature at which the seeding operation is started when the number of single crystal growths exceeds a predetermined number n, and an implementation step 13 of starting the seeding operation at the timing when the crucible bottom temperature reaches the predicted crucible bottom temperature predicted in the prediction step 12. The prediction step 12 uses X as the number of single crystal growths repeatedly performed in the same furnace, Y as the crucible bottom temperature at the time of the seeding operation, and Yp as the predicted crucible bottom temperature at the time of the seeding operation for the number of growths X = p to be performed hereafter. When the number of single crystal growths X exceeds a predetermined number n (n < p), from the information acquired in the acquisition step 11, a first step 12a of selecting a group Xn of the number of single crystal growths for the last n times and a group Yn of the crucible bottom temperatures at the time of the seeding operation from the number of single crystal growths X, a second step 12b of obtaining a relational expression (I): Y = aX + b between the number of single crystal growths X and the crucible bottom temperature Y at the time of the seeding operation using the least squares method from the selected group Xn of the number of single crystal growths and the group Yn of the crucible bottom temperatures, and a third step 12c of determining the calculated crucible bottom temperature Y as the predicted crucible bottom temperature Yp by substituting the number of growths X = p to be performed hereafter into the number of single crystal growths X in the relational expression (I): Y = aX + b. Here, "seeding" refers to a technique of bringing a seed crystal S into contact with a raw material melt M and gradually solidifying the raw material melt M around the seed crystal S to promote crystal growth. Further, the "seeding operation" means an operation of bringing the seed crystal S into contact with the raw material melt M. Furthermore, the "appropriate seeding temperature" means the crucible bottom temperature when the seeding operation is performed at an appropriate timing.

[0016] In this technical means, the present embodiment targets a single crystal that can be grown by a rotational pulling method such as the Cz method, etc. Examples of the single crystal to be grown include, but are not limited to, oxide single crystals such as lithium tantalate single crystal, lithium niobate single crystal, and sapphire single crystal (Al2O3). Furthermore, this embodiment is characterized by the prediction method, in which the seeding operation, which is important in the single crystal growth method, is performed based on the predicted value by predicting the crucible bottom temperature. Therefore, in this embodiment, the prediction step 12, as described above, must include a first step 12a (corresponding to a step of selecting a group of Xn times of single crystal growth for the most recent n previous times and a group of Yn crucible bottom temperatures during the seeding operation, which are necessary for creating the relational expression), a second step 12b (corresponding to a step of creating the relational expression), and a third step 12c (corresponding to a step of calculating the predicted crucible bottom temperature). Furthermore, the prediction step 12 and the implementation step 13 of this embodiment are performed in the seeding operation when the number of times X of growth is n+1 or later. For this reason, when growing a single crystal up to n times, existing methods may be used for the seeding operation, for example, a method in which an operator observes the state of the raw material melt M in the furnace through an observation window (not shown) in the growth furnace and performs the seeding at the appropriate timing based on the operator's judgment.

[0017] In this embodiment, the seeding operation must be started at an appropriate seeding temperature for the following reasons. Specifically, as shown in Figure 1(a), a crystal is grown at the tip of a seed crystal S by seeding. If the seeding temperature is lower than the appropriate seeding temperature, the crystal begins to grow rapidly immediately after the seed crystal S comes into contact with the raw material melt M. This causes the meniscus diameter to expand, resulting in a deterioration in crystal quality and even polycrystallization, as shown in Figure 2(a). Conversely, if the seeding temperature is higher than the appropriate seeding temperature, the seed crystal S melts before coming into contact with the raw material melt M. This causes the meniscus diameter to shrink or disappear, as shown in Figure 2(c), potentially preventing crystal growth. In contrast, if the seeding temperature is appropriate, the meniscus diameter of the crystal is stable, as shown in Figure 2(b), enabling the growth of high-quality single crystals.

[0018] Next, a typical or preferred embodiment of the method for growing a single crystal according to this embodiment will be described. In this example, a typical embodiment of the acquisition step 11 is to use information detected by a temperature detection means incorporated in the bottom of the crucible. This embodiment employs a temperature detection means that is preferable for acquiring the temperature at the bottom of the crucible. Here, a type B thermocouple can be used as the temperature detection means, because a type B thermocouple with high heat resistance is preferable as the temperature detection means in the region where the raw material melt M is used when seeding is started.

[0019] In this example, the prediction step 12 is performed using the Xn group of the most recent n single crystal growth times and the Yn group of the crucible bottom temperatures during the seeding operation, where the value of n is preferably 4≦n≦10, and more preferably 5≦n≦7. Here, the number of most recent n times can be selected as appropriate, but if n ≦ 3, the predicted temperature will vary due to insufficient data. If n > 10, it will be difficult to capture changes due to recent deterioration of the refractory material that makes up the furnace body or deterioration of the thermocouple. Therefore, the preferred range is 4 ≦ n ≦ 10. In particular, if n is set in the range of 5 ≦ n ≦ 7, it is possible to further reduce the difference between the predicted and measured crucible bottom temperatures compared to other cases.

[0020] Furthermore, in the first step 12a of the prediction step 12, the Xn group of the most recent n number of single crystal growths and the Yn group of crucible bottom temperatures during seeding operations are selected each time to create the relational formula (I), but this is not limited to this, and it is preferable to adopt a method of excluding any inappropriate data groups selected (Xn group, Yn group). When such a method is adopted, in the first step 12a, when selecting the group of Xn times of single crystal growth and the group of Yn times of crucible bottom temperatures during seeding operations for the most recent n times, if the i-th Yn(i) in the group of Yn times of crucible bottom temperatures during seeding operations includes a seeding operation performed using a predicted crucible bottom temperature Yp, a correction process may be included to exclude the i-th Yn(i) from the data candidates for the group of Xn times of single crystal growth when the difference between the crucible bottom temperature Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is outside the allowable range.

[0021] Furthermore, preferred embodiments of the second step 12b when the correction process is included in the first step 12a include an embodiment in which, when the crucible bottom temperature Yn(i) is excluded in the correction process, the relationship formula (I) is determined using a group of crucible bottom temperatures Yn other than the excluded one, or an embodiment in which, when the crucible bottom temperature Yn(i) is excluded in the correction process, the relationship formula (I) is determined using the most recent data from the most recent data group of the past n data sets as a data candidate other than the excluded one.

[0022] Furthermore, when the above-described single crystal growing method is embodied, a single crystal growing apparatus as shown in FIG. 1(b) can be obtained. In Fig. 1(b), the single crystal growth apparatus includes a furnace 10 for blocking external influences, a crucible 1 provided in the furnace 10 into which a raw material for a single crystal is charged, heating means 2 for heating and melting the raw material for a single crystal charged into the crucible 1 to form a raw material melt M, lifting means 3 for performing a seeding operation of bringing a seed crystal S into contact with the raw material melt M in the crucible 1 and lifting the seed crystal S while rotating it to grow a single crystal, temperature detection means 4 provided at the bottom of the crucible 1 for detecting the temperature of the crucible bottom, and control means 5 for controlling the seeding operation and the lifting operation by the lifting means 3 based on the temperature information detected by the temperature detection means 4. The control means 5 includes acquisition means 7 for acquiring the temperature of the crucible bottom during the seeding operation for each number of single crystal growths by the temperature detection means 4, prediction means 8 for predicting, as a predicted crucible bottom temperature, the crucible bottom temperature at which the seeding operation starts when the number of single crystal growths exceeds a predetermined number n, and execution means 9 for starting the seeding operation at the timing when the crucible bottom temperature reaches the predicted crucible bottom temperature predicted by the prediction means 8. The prediction means 8 sets the number of single crystal growths repeatedly performed in the same furnace 10 as X, the temperature of the crucible bottom during the seeding operation as Y, and the predicted crucible bottom temperature during the seeding operation at the number of growths X = p to be performed hereafter as Yp. When the number of single crystal growths X exceeds a predetermined number n (n < p), an information selection unit 8a for selecting, from the information acquired by the acquisition means 7, a group Xn of the number of single crystal growths X for the most recent past n times and a group Yn of the temperature of the crucible bottom during the seeding operation, a relational expression creation unit 8b for obtaining a relational expression (I): Y = aX + b for the relationship between the number of single crystal growths X and the temperature of the crucible bottom during the seeding operation Y using the least squares method from the group Xn of the number of single crystal growths and the group Yn of the temperature of the crucible bottom selected by the information selection unit 8a, and a calculation unit 8c for determining, as the predicted crucible bottom temperature Yp, the temperature of the crucible bottom calculated by substituting the number of growths X = p to be performed hereafter into the number of single crystal growths X in the relational expression (I): Y = aX + b.

[0023] In this example, the acquisition means 7 of the control means 5 embodies the acquisition step 11. Furthermore, the prediction means 8 of the control means 5 embodies the prediction step 12, and the information selection unit 8a, the relational equation creation unit 8b, and the calculation unit 8c embody the first step 12a, the second step 12b, and the third step 12c of the prediction step 12. Furthermore, the execution means 9 of the control means 5 embodies the implementation step 13. Furthermore, a preferred embodiment of the single crystal growth apparatus according to this example is one in which the information selection unit 8a, when selecting the group of Xn times of single crystal growth for the most recent n times and the group of crucible bottom temperatures Yn during the seeding operation, includes a correction processing unit (not shown) that, when the ith Yn(i) among the group of crucible bottom temperatures Yn during the seeding operation includes a seeding operation performed using a predicted crucible bottom temperature Yp, excludes the i-th Yn(i) from the data candidates for the group of single crystal growth times Xn when the difference between the crucible bottom temperature Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is outside the allowable range.

[0024] The present invention will be described in more detail below based on the embodiments shown in the accompanying drawings. Embodiment 1 -Overall configuration of single crystal growth equipment- FIG. 3 shows the overall configuration of a single crystal growth apparatus according to the first embodiment. In the figure, a single crystal growth apparatus 20 grows single crystals of, for example, LT or LN. Here, single crystals such as LT or LN are used, for example, as materials for surface acoustic wave devices (SAW filters). This type of SAW filter has a structure in which comb-shaped electrodes made of a metal thin film such as an AlCu alloy are formed on a substrate made of a piezoelectric material such as LT or LN single crystal. The pitch of the comb-shaped electrodes and the sound velocity of the piezoelectric material play an important role in determining the frequency of the device. The comb-shaped electrodes are formed by depositing a metal thin film on a substrate made of a piezoelectric material using a method such as sputtering, and then etching away unnecessary portions of the metal thin film using photolithography to form the comb-shaped pattern.

[0025] In this example, the single crystal growth apparatus 20 has a structure that employs the Cz method as a rotational pulling method. In other words, the single crystal growth apparatus 20 comprises a crucible 1 into which the raw material for the single crystal is charged, a heat-insulating and insulating material 22 that surrounds the outer periphery and top of the crucible 1 to form a growth furnace 21 capable of keeping the temperature and heat insulating, a heating coil 23 that is installed so as to surround the heat-insulating and insulating material 22 of the crucible 1 and serves as a heating means for heating the crucible 1, and a lifting mechanism 24 that has a lifting shaft 25 that can be raised and lowered and serves as a lifting means for bringing a seed crystal S held at the tip of the lifting shaft 25 into contact with the surface of the molten raw material melt M, and then gradually lifting the seed crystal S while rotating the lifting shaft 25 to grow the single crystal.

[0026] In this example, the material of the crucible 1 is appropriately selected depending on the melting point of the growing crystal and the atmosphere inside the growth furnace 21. When growing an oxide single crystal, the crucible 1 must be made of a material that is difficult to oxidize because the inside of the growth furnace 21 is an oxidizing atmosphere. For example, iridium is used for LT single crystals, and platinum is used for LN single crystals. The heating coil 23 is configured, for example, as a high-frequency induction coil, and by applying high-frequency current from the high-frequency heating power supply 26, the crucible 1 and the single crystal raw material are heated, and the single crystal raw material is maintained at a high temperature to form a melt. Furthermore, in this example, a temperature sensor 30 is provided at approximately the center of the bottom of the crucible 1 as a temperature detection means for detecting the temperature at the bottom of the crucible. In this example, a B-type thermocouple is used as the temperature sensor 30, which is suitable for measuring temperatures of 1000°C or higher and suitable for long-term use in an oxidizing atmosphere. However, the type of temperature sensor 30 is not particularly limited, and various thermocouples (such as an R-type thermocouple), or optical fiber sensors or pyrometers may also be used.

[0027] In this example, the lifting mechanism 24 is connected to a control device 40 as a control means. Furthermore, the temperature sensor 30 is connected to the control device 40 via a digital thermometer 31. Therefore, in this example, the detected value output from the temperature sensor 30 is read as temperature information by the digital thermometer 31 and then input into the control device 40. Furthermore, a high-frequency output regulator 27 is connected to the high-frequency heating power supply 26, and this high-frequency output regulator 27 is connected to the control device 40. The control device 40 determines the growth state of the single crystal based on the elevation position of the pulling shaft 25 and the detection value of the temperature sensor 30, and controls the driving of the pulling shaft 25 by the elevation mechanism 24 (driving for rotation and elevation movements) and the heating operation by the heating coil 23 in accordance with a predetermined growth drive program. In this example, the growth drive program includes a seeding operation program (see Figures 4 and 5) used during the seeding operation of bringing the seed crystal S into contact with the raw material melt M, and further includes a growth adjustment program (temperature conditions, rotation conditions, pulling speed) that adjusts the single crystal growth process after seeding.

[0028] -Operation of single crystal growth equipment- Next, the basic operation of the single crystal growth apparatus according to this embodiment will be described. (a) In an oxidizing atmosphere in the growth furnace 21, the crucible 1 is heated by the heating coil 23, the raw material for the single crystal in the crucible 1 is melted, and the temperature is maintained at a temperature near the melting point. (b) Thereafter, the seeding operation is carried out at an appropriate timing. That is, the lifting mechanism 24 lowers the pulling shaft 25, and the seed crystal S held at the tip of the pulling shaft 25 is brought into contact with the surface of the molten raw material melt M. Details of the seeding operation will be described later. (c) Then, when the precipitation of the single crystal begins, the seed crystal S and the single crystal are rotated by the pulling shaft 25 of the lifting mechanism 24 and pulled up vertically at a constant speed to grow the single crystal. (d) After that, when the growth of the single crystal has progressed to the designated effective diameter and the designated effective length, the lifting mechanism 24 further raises the pulling shaft 25 to separate the single crystal from the raw material melt M, and the single crystal is moved to a predetermined height. (e) After this, the output of the high frequency heating power supply 26 starts to decrease, and cooling of the single crystal begins. Then, when the single crystal has completely cooled, the single crystal is taken out of the growth furnace 21.

[0029] - Seeding operation processing - FIG. 4 is a flowchart showing the process steps during the seeding operation of the single crystal growth apparatus according to this embodiment. First, in this example, it is assumed that single crystals are repeatedly grown using the crucible 1 of the same growth furnace 21, and therefore it is necessary to input the number of times X that single crystals have been grown in the same furnace into the control device 40. In this case, the number of times X that single crystals have been grown may be input by an operator each time, but if a counter is set up to automatically count the number of times single crystals have been grown, the count value of the counter may be used as is.

[0030] Next, in order to perform the prediction step of predicting the crucible bottom temperature Y at which the seeding operation is to be started as the predicted crucible bottom temperature Yp, it is necessary to select temperature data from the most recent n single crystal growth processes, where n is preferably selected within the range of 4≦n≦10, more preferably 5≦n≦7. In this example, the control device 40 performs the above-mentioned prediction step when the number of times X of single crystal growth is n+1 or later. On the other hand, when the number of times X of single crystal growth is equal to or less than n, the prediction step by the control device 40 is not performed, and the previous seeding operation is performed. The "conventional seeding operation" referred to here is, for example, a method in which an operator observes the state of the raw material melt M in the furnace through an observation window in the growth furnace 21, determines the "appropriate seeding temperature" based on the operator's judgment, and starts the seeding operation at the appropriate timing. Note that, if the surface state of the raw material melt M is displayed on a monitor in the growth furnace 21, the operator may of course determine the "appropriate seeding temperature" while looking at the monitor image.

[0031] <Acquisition process> In this example, the control device 40 performs an acquisition process to acquire the crucible bottom temperature Y during the seeding operation for each single crystal growth, regardless of whether a prediction process is performed or not, and as shown in Figure 4, the crucible bottom temperature Y during the seeding operation for each single crystal growth count X is measured by the temperature sensor 30 and stored in a memory unit such as a RAM (not shown).

[0032] <Prediction process> In this example, the prediction step is performed according to the flowchart shown in FIG. 5 when the number of times of single crystal growth X>n, as shown in FIG. In this example, the prediction process consists of first to third processes. First, the control device 40 performs a first step and selects the most recent n number of single crystal growths Xn group and the crucible bottom temperature Yn group during the seeding operation from the data group of the crucible bottom temperature Y during the seeding operation stored in a memory unit such as RAM. Specifically, if the number of single crystal growths X is p (p>n), the data group of the most recent n number of single crystal growths Xn group is p (i = 1 to n) times, and the crucible bottom temperature Yn group during the seeding operation is the data group of p (i = 1 to n) times of the crucible bottom temperature Y(p i ) (i = 1 to n). For example, if n = 5 is selected and p = 6, the Xn group is the data group of X = 1 to 5 times, and the Yn group is the data group of Y(1) to Y(5). In addition, when p = 7, the Xn group is a data group of X = 2 to 6 times, and the Yn group is a data group of Y(2) to Y(6). Furthermore, when p = 8, the Xn group is a data group of X = 3 to 7 times, and the Yn group is a data group of Y(3) to Y(7).

[0033] Thereafter, the control device 40 performs a second step, in which the least squares method is applied to the data group selected in the first step to calculate an approximate relational expression (I). Y=aX+b ……(I) where a is the slope coefficient and b is the initial constant when X is set to 0. The reason for calculating relational expression (I) by focusing on the data from the most recent n runs is as follows. Specifically, the crucible bottom temperature Yn group used during the seeding operation in the most recent n runs of single crystal growth appears to decrease as the number of single crystal growth runs X increases due to deterioration of the refractory material constituting the furnace body and the thermocouple installed in the crucible bottom. However, this decrease is not necessarily constant, and it was found that the degree of decrease in the apparent crucible bottom temperature varies slightly depending on the number of growth runs. In this case, if n is too small, there is insufficient data for least-squares approximation. On the other hand, if n is too large, there is a concern that the least-squares approximation will be affected by too old data. For this reason, in this example, n is selected within the range of 4≦n≦10, as mentioned above.

[0034] Furthermore, the control device 40 performs the third step, and substitutes the number of growth rounds X=p to be performed into the relational expression (I) to calculate Y. Then, the control device 40 determines the calculated Y as the predicted crucible bottom temperature Yp.

[0035] <Implementation process> Then, when the single crystal is about to be grown (growth count X=p), the control device 40 performs the seeding operation when the crucible bottom temperature Y reaches the predicted crucible bottom temperature Yp predicted in the prediction step. When performing the seeding operation, the meniscus diameter is stable within a range of approximately ±0.5°C from the appropriate seeding temperature, but if the meniscus diameter is not stable after starting the seeding operation, the operator performs the seeding operation while fine-tuning the temperature. To evaluate the seeding operation in this example, we compared the predicted crucible bottom temperature Yp with the actual crucible bottom temperature Y during the seeding operation for each single crystal growth run from the n+1th run onward. We found that the difference between the two was extremely small, staying within ±1.6°C when the seeding temperature was around 1500°C. Therefore, even if the meniscus diameter is unstable after starting the seeding operation in this example, the operator can easily find the appropriate seeding temperature by fine-tuning the temperature based on the predicted crucible bottom temperature Yp. While the smaller the difference between the appropriate seeding temperature and the predicted crucible bottom temperature Yp, a difference within ±2°C significantly reduces the operator's burden during the seeding operation. Thus, it can be seen that the prediction accuracy of the predicted crucible bottom temperature Yp in this example is sufficiently high. This will be described in detail in the examples below.

[0036] Thus, according to this embodiment, from the n+1th single crystal growth iteration X onwards, the seeding operation can be carried out at the appropriate timing by using the predicted crucible bottom temperature Yp calculated in advance based on the data of the most recent n previous growths (Xn group, Yn group), without the need for the operator to observe the state of the raw material melt M through the observation window of the growth furnace 21. If the predicted crucible bottom temperature Yp were calculated based on data from all past growth batches instead of the data from the most recent n growth batches, there is a concern that the predicted crucible bottom temperature Yp would be more likely to vary due to variations in the number of data, and the difference between the predicted crucible bottom temperature Yp and the actual crucible bottom temperature Y during the seeding operation would become larger. In particular, once the number of growth batches exceeds about 10, it becomes difficult to capture recent changes when formulating relational expression (I). Furthermore, in this embodiment, when performing the seeding operation, there is no need to measure the crucible bottom temperature when the raw material is melted, as in the method of Patent Document 2. This not only reduces the number of times temperature data is acquired by the temperature sensor 30, but also makes it possible to predict the predicted crucible bottom temperature Yp in advance before the raw material for the single crystal is melted.

[0037] -Relationship between the number of times single crystals are grown and maintenance- Single crystal growth is carried out several tens of times in succession in the same growth furnace 21, and maintenance of the growth furnace 21 and the temperature sensor 30 is carried out every several tens of times. The number of times growth is carried out consecutively varies depending on the type of single crystal, but for LT, for example, it is about 30 to 60 times. Here, maintenance of the growth furnace 21 includes, for example, reassembling the heat-insulating material 22, which is a refractory material. In addition, maintenance of the temperature sensor 30 is also performed in conjunction with the maintenance of the growth furnace 21. When a thermocouple is used as the temperature sensor 30, the maintenance of the thermocouple includes, for example, removing the thermocouple from the bottom of the crucible 1, removing any deteriorated portion of the tip of the thermocouple, reconnecting the electrodes, and then reattaching the thermocouple to the bottom of the crucible 1. In this way, once maintenance is performed, the conditions of the growth furnace 21 and the temperature sensor 30 (for example, a thermocouple) are completely changed from the conditions before the maintenance. Therefore, after the maintenance, the single crystal growth is started under new conditions. That is, the number of growth times X is reset using the crucible 1 of the growth furnace 21 after the maintenance, and the single crystal growth is repeated sequentially from the first time. The thermocouple serving as the temperature sensor 30 is replaced with a new thermocouple after the above-described maintenance is repeated several times (for example, about 3 to 5 times).

[0038] Embodiment 2 The basic configuration of the single crystal growth apparatus according to the second embodiment is substantially the same as that of the first embodiment, but the prediction process of the seeding operation by the control device 40 is different from that of the first embodiment. Note that the same components as those in the first embodiment are denoted by the same reference numerals as those in the first embodiment, and detailed description thereof will be omitted here. FIG. 6 shows a flowchart of the prediction process used in the second embodiment. In the same figure, the control device 40 performs the first step, as in embodiment 1, and selects the Xn group of the most recent n single crystal growth times and the Yn group of crucible bottom temperatures during seeding operations from the data group of crucible bottom temperatures Y during seeding operations stored in a memory unit such as a RAM. Thereafter, in the first step, the control device 40 checks whether the selected data groups (Xn group, Yn group) are appropriate for creating a relational expression.

[0039] Specifically, if the number of times X to grow a single crystal to be performed this time is p (p>n), and the ith crucible bottom temperature Yn(i) in the group of crucible bottom temperatures Yn during the seeding operation includes one in which the seeding operation was performed using the predicted crucible bottom temperature Yp, the difference ΔT(i) between the crucible bottom temperature Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is calculated, and it is checked whether the difference ΔT(i) is within the allowable range U. ΔT(i)=|Yn(i)-Yp(i)|≦U In this case, the allowable range U may be determined in advance depending on the prediction accuracy of the predicted crucible bottom temperature Yp. The allowable range U may be set to, for example, about 2 to 3°C. If ΔT(i) is equal to or less than the allowable range U, the crucible bottom temperature Yn(i) is selected as a data candidate for the group of single crystal growth times Xn. On the other hand, if ΔT(i) is outside the allowable range, a correction process is performed in which the crucible bottom temperature Yn(i) is excluded from the data candidates for the group of single crystal growth times Xn.

[0040] Thereafter, if the control device 40 determines that all of the data candidates for the group of single crystal growth times Xn are appropriate, it carries out the second step, as in embodiment 1, and calculates an approximate relational expression (I) by using the least squares method on the data group selected in the first step. On the other hand, when the crucible bottom temperature Yn(i) is excluded from the data candidates of the group of single crystal growth times Xn in the correction process, whether or not to supplement the data candidates is designated in advance. Here, if it is specified that no data candidates are to be added, the control device 40 determines the data group (Xn group, Yn group) other than the crucible bottom temperature Yn(i) for the excluded single crystal growth number Xn(i) as the data group selected in the first step, and calculates an approximated relational expression (I) using the least squares method for the determined data group. In this example, if the number of data items in the crucible bottom temperature Yn group for the number of single crystal growths Xn group becomes 3 or less as a result of excluding data candidates, it is preferable to process the data items to be supplemented with the data candidates described below.

[0041] Furthermore, when specifying to replenish data candidates, the control device 40 may replenish the most recent data of the crucible bottom temperature Yn group of the most recent n number of single crystal growths Xn group as data candidates, that is, the data (X(j), Y(j): where j=pn-1) from the n+1th time prior, based on the number of single crystal growths X=p to be performed this time. In this case, the control device 40 determines the data group selected in the first step as the data group including the data candidates added to the data group (Xn group, Yn group) excluding the excluded data, and calculates an approximated relational equation (I) using the least squares method on the determined data group. It is preferable to select data candidates to be supplemented such that ΔT(i) falls within the allowable range. If the ΔT(i) of a supplemented data candidate is outside the allowable range, the data candidate can be supplemented again.

[0042] Thereafter, the control device 40 performs the third step, and substitutes the number of growth rounds X=p to be performed into the relational expression (I) to calculate Y. Then, the control device 40 determines the calculated Y as the predicted crucible bottom temperature Yp. Thus, in this embodiment, in the first step, when the control device 40 selects the appropriate data group (Xn group, Yn group) for creating the relational equation (I), it performs the correction process described above to select the case where the predicted crucible bottom temperature Yp(i) used in the previous single crystal growth cycle i is a value that is extremely close (ΔT(i)≦U) to the actual crucible bottom temperature Y(i) during the seeding operation. Therefore, when creating the relational expression (I) for calculating the predicted crucible bottom temperature Yp, selecting a crucible bottom temperature Y(i) with a small difference ΔT(i) from the predicted crucible bottom temperature Yp(i) makes it easier to capture recent changes in the relational expression (I). Therefore, it is possible to improve the prediction accuracy of the predicted crucible bottom temperature Yp compared to when the correction process described above is not performed in the first step. [Example]

[0043] Example 1 In Example 1, a single crystal was grown using an embodiment of the single crystal growth apparatus according to the first embodiment. In this example, an LT raw material was used to grow an LT single crystal by the Cz method using a single crystal growth apparatus 20 shown in Figure 3. The growth atmosphere in the growth furnace 21 was a nitrogen-oxygen mixed gas with an oxygen concentration of 2%. First, the LT raw material was placed in an iridium crucible 1, and a B-type thermocouple was installed as a temperature sensor 30 near the center of the bottom of the crucible 1. In order to regulate the temperature environment during growth, alumina and zirconia heat-insulating materials 22 were assembled around and above the crucible 1 to form a growth furnace 21, and the temperature was then raised to melt the LT raw material.

[0044] Thereafter, a seeding operation is performed at an appropriate timing to bring the tip of the seed crystal S into contact with the raw material melt M. Then, the seed crystal S and the single crystal are rotated by the pulling shaft 25 of the lifting mechanism 24 while being pulled up vertically at a constant speed, thereby growing the single crystal. This time, the growth of LT single crystals was repeated multiple times (11 times in this example) using the crucible 1 in the same growth furnace 21, and the crucible bottom temperature Y (corresponding to the seeding temperature) during the seeding operation was measured for each growth. Furthermore, until the number of times X of LT single crystal growth reached n, the timing to start the seeding operation was determined by an operator by observing the state of the melt inside the furnace through an observation window in the growth furnace 21. On the other hand, when the number of times X of LT single crystal growth reached n+1 or later, the above-mentioned prediction step (see FIG. 5) was carried out for the seeding operation, the predicted crucible bottom temperature Yp was predicted, and the seeding operation was started with reference to this predicted crucible bottom temperature Yp.

[0045] Here, n was selected from 4 to 9, and for each case, the relational expression (I) = aX + b was created to calculate the predicted crucible bottom temperature Yp. In addition, the difference (prediction - start) between the predicted crucible bottom temperature Yp calculated for each LT single crystal growth run and the actual crucible bottom temperature Y (seeding temperature) during the seeding operation was calculated. Here, [Table 1] shows the case where n = 4, [Table 2] shows the case where n = 5, [Table 3] shows the case where n = 6, [Table 4] shows the case where n = 7, [Table 5] shows the case where n = 8, and [Table 6] shows the case where n = 9. In [Table 2] to [Table 6], the crucible bottom temperature Y (seeding temperature) is omitted until the number of growth times X reaches n.

[0046] [Table 1]

[0047] [Table 2]

[0048] [Table 3]

[0049] [Table 4]

[0050] [Table 5]

[0051] [Table 6]

[0052] 7 is a graph plotting the crucible bottom temperature Y and the predicted crucible bottom temperature Yp for each growth run, for example, in Table 2. In this figure, the difference between the plotted points for each growth run corresponds to the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y. From this figure, it can be seen that the predicted crucible bottom temperature Yp during the seeding operation is predicted to be close to the actual crucible bottom temperature Y (seeding temperature). According to Tables 1 to 6, in all cases, in a molten state around 1480°C, the difference (predicted - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is kept to 1.2°C or less, which indicates that the prediction accuracy of the predicted crucible bottom temperature Yp is sufficiently high. Although Tables 1 to 6 show the case where the number of times of growth X is 11 or less, it was confirmed that the prediction accuracy of the predicted crucible bottom temperature Yp was maintained at a sufficiently high level when the number of times of growth X was repeated up to 50 times. Furthermore, when n = 10, the above-mentioned prediction process was carried out in the single crystal growth cycles from n+1 onwards, and it was confirmed that it was possible to predict a predicted crucible bottom temperature Yp that approximates the actual crucible bottom temperature Y (seeding temperature) during the seeding operation.

[0053] Furthermore, in the cases of Tables 2 to 4 (n = 5 to 7), the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is kept to a maximum of 1.1°C or less, and the variation (± fluctuation range) of the difference (prediction - start) for each growth run is also kept to a maximum of 1.4°C, which is a sufficiently small range. In contrast, for example, in the case of [Table 1] (n=4), the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is kept to 0.8°C or less, but the variation (± fluctuation) of the difference (prediction - start) for each growth run is up to 1.5°C. Also, for example, in the case of [Table 5] (n=8), the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is up to 1.2°C. Based on these results, it is clear that it is preferable to select n=5 to 7 from the viewpoint of maintaining both the difference (prediction - start) between each rearing run and the variability of the difference (prediction - start) between each rearing run in a good condition.

[0054] Example 2 In Example 2, a single crystal was repeatedly grown using a crucible 1 of a growth furnace 21 different from that of Example 1, which was an embodiment of the single crystal growth apparatus according to the first embodiment. In this example, an LT raw material was used and an LT single crystal was grown by the Cz method using a single crystal growth apparatus 20 shown in Figure 3. The growth atmosphere and growth conditions in the growth furnace 21 were approximately the same as in Example 1. As in Example 1, n was selected from 4 to 9, and for each case, the relational expression (I) = aX + b was created to calculate the predicted crucible bottom temperature Yp. In addition, the difference (prediction - start) between the predicted crucible bottom temperature Yp calculated for each LT single crystal growth run and the actual crucible bottom temperature Y (seeding temperature) during the seeding operation was calculated. Here, Table 7 shows the case where n = 4, Table 8 shows the case where n = 5, Table 9 shows the case where n = 6, Table 10 shows the case where n = 7, Table 11 shows the case where n = 8, and Table 12 shows the case where n = 9. In Tables 8 to 12, the crucible bottom temperature Y (seeding temperature) is omitted until the number of growth cycles X reaches n.

[0055] [Table 7]

[0056] [Table 8]

[0057] [Table 9]

[0058] [Table 10]

[0059] [Table 11]

[0060] [Table 12]

[0061] Fig. 8 is a graph plotting the crucible bottom temperature Y and the predicted crucible bottom temperature Yp for each growth run, for example, in Table 8. In this figure, the difference between the plotted points for each growth run corresponds to the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y. From this figure, it can be seen that the predicted crucible bottom temperature Yp during the seeding operation is predicted to be close to the actual crucible bottom temperature Y (seeding temperature). According to Tables 7 to 12, in all cases, in a molten state around 1530°C, the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is kept to 1.6°C or less, which shows that the prediction accuracy of the predicted crucible bottom temperature Yp is sufficiently high. Although Tables 7 to 12 show the case where the number of times of growth X is up to 11, it was confirmed that the prediction accuracy of the predicted crucible bottom temperature Yp was maintained at a sufficiently high level when the number of times of growth X was repeated up to 50 times. Furthermore, when n = 10, the above-mentioned prediction process was carried out in the single crystal growth cycles from n+1 onwards, and it was confirmed that it was possible to predict a predicted crucible bottom temperature Yp that approximates the actual crucible bottom temperature Y (seeding temperature) during the seeding operation.

[0062] Furthermore, in the cases of Tables 8 to 10 (n = 5 to 7), the difference (prediction - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is kept to a maximum of 1.2°C or less, and the variation in the difference (± fluctuation range) from one growth run to the next is also kept to a sufficiently small range of a maximum of 0.9°C. In contrast, for example, in the case of Table 7 (n=4), the difference (predicted - start) between the predicted crucible bottom temperature Yp during the seeding operation and the actual crucible bottom temperature Y is kept to 1.6°C or less, but the variation in the difference (± fluctuation range) for each growth run is up to 2.5°C. Based on these results, it can be seen that, similarly to Example 1, it is preferable to select n=5 to 7 from the viewpoint of maintaining good balance between the difference between each rearing run and the variation of the difference between each rearing run.

[0063] Comparative Example 1 Fifty LT single crystals were grown under the same conditions as in Example 1, except that the operator determined the "appropriate seeding temperature" from the melt surface state observed on the monitor image inside the growth furnace. Comparative Example 2 Fifty LT single crystals were grown under the same conditions as in Example 2, except that the operator determined the "appropriate seeding temperature" from the melt surface state observed on the monitor image inside the growth furnace.

[0064] In both Comparative Examples 1 and 2, the seeding operation was initiated at the discretion of the operator in every single crystal growth run, and so the success rate of single crystal growth tended to decrease depending on the operator's level of proficiency. However, in Examples 1 and 2, a method was adopted in which the seeding temperature was predicted using the predicted crucible bottom temperature Yp from the n+1th single crystal growth run onwards, which not only significantly increased the success rate of single crystal growth but also simplified the seeding operation for the operator. [Industrial Applicability]

[0065] The single crystal growth method according to the present invention can improve the success rate of single crystal growth by a rotational pulling method such as the Cz method without relying on the skill of the operator, and can also shorten the seeding time, making it highly industrially applicable to the growth of single crystals such as LT single crystals and LN single crystals by the Cz method. [Explanation of symbols]

[0066] 1...crucible, 2...heating means, 3...lifting means, 4...temperature detection means, 5...control means, 7...acquisition means, 8...prediction means, 8a...information selection unit, 8b...relational equation creation unit, 8c...calculation unit, 9...execution means, 10...furnace, 11...acquisition process, 12...prediction process, 12a...first process, 12b...second process, 12c...third process, 13...implementation process, 20...growth apparatus, 21...growth furnace, 22... Heat insulation material, 23...heating coil, 24...lifting mechanism, 25...pulling shaft, 26...high frequency heating power supply, 27...high frequency output regulator, 30...temperature sensor, 31...digital thermometer, 40...control device, M...raw material melt, S...seed crystal, X...number of growths, Xn...number of most recent n growths, Y...bottom temperature of crucible, Yn...bottom temperature of crucible for most recent n growths, Yp...predicted bottom temperature of crucible

Claims

1. A method for growing a single crystal, comprising the steps of: heating and melting a raw material for a single crystal placed in a crucible in a furnace; performing a seeding operation in which a seed crystal is brought into contact with the raw material melt obtained by heating and melting the raw material for the single crystal when the temperature of the bottom of the crucible, which is the temperature at the bottom of the crucible, reaches an appropriate seeding temperature; and growing a single crystal by a rotational pulling method in which the seed crystal is pulled up while rotating, in the same furnace, an acquiring step of acquiring the temperature of the bottom of the crucible during the seeding operation for each number of times of growing the single crystal; a prediction step of predicting the crucible bottom temperature at which the seeding operation is started when the number of times the single crystal has been grown exceeds a predetermined number n; an implementation step of starting the seeding operation when the crucible bottom temperature reaches the predicted crucible bottom temperature predicted in the prediction step; Equipped with In the prediction step, the number of times of single crystal growth repeated in the same furnace is X, the crucible bottom temperature during the seeding operation is Y, and the predicted crucible bottom temperature during the seeding operation when the number of times of growth to be performed next is X=p is Yp, a first step of selecting, when the number of times X of growth exceeds a predetermined number n (n<p), a group of Xn of single crystal growth times immediately preceding the number of times X of growth and a group of Yn of crucible bottom temperatures during seeding operations from the information acquired in the acquisition step; a second step of determining a relational expression (I): Y = aX + b between the number of times X to grow the single crystal and the crucible bottom temperature Y during the seeding operation by using a least squares method from the group of number of times Xn to grow the single crystal and the group of crucible bottom temperatures Yn selected in the first step; a third step of determining the crucible bottom temperature Y, calculated by substituting the number of times X (=p) of single crystal growth to be performed into the number of times X (=aX+b) of single crystal growth in the relational expression (I): Y=aX+b, as a predicted crucible bottom temperature Yp; A method for growing a single crystal, comprising:

2. 2. The method for growing a single crystal according to claim 1, A method for growing a single crystal, wherein the value of n is 4≦n≦10.

3. 3. The method for growing a single crystal according to claim 2, A method for growing a single crystal, wherein the value of n is 5≦n≦7.

4. 2. The method for growing a single crystal according to claim 1, A method for growing a single crystal, wherein the acquiring step uses information detected by a temperature detecting means incorporated in the bottom of the crucible.

5. 5. The method for growing a single crystal according to claim 4, A method for growing a single crystal, wherein the temperature detecting means is a type B thermocouple.

6. 2. The method for growing a single crystal according to claim 1, The first step is a method for growing a single crystal, characterized in that when selecting the Xn group of the most recent n number of single crystal growths and the Yn group of crucible bottom temperatures during the seeding operation, if the i-th Yn(i) in the Yn group of crucible bottom temperatures during the seeding operation includes a seeding operation performed using the predicted crucible bottom temperature Yp, the i-th Yn(i) in the Yn group of crucible bottom temperatures during the seeding operation is excluded from the data candidates for the Xn group of single crystal growths if the difference between the Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is outside the allowable range.

7. 7. The method for growing a single crystal according to claim 6, The second step is a method for growing a single crystal, characterized in that when the crucible bottom temperature Yn(i) is excluded in the correction process, the relationship formula (I) is determined using the group of crucible bottom temperatures Yn other than the excluded one.

8. 7. The method for growing a single crystal according to claim 6, The second step is a method for growing a single crystal, characterized in that when the crucible bottom temperature Yn(i) is excluded in the correction process, the relational formula (I) is obtained using the most recent data from the most recent n data sets as a data candidate other than the excluded data.

9. 2. The method for growing a single crystal according to claim 1, A method for growing a single crystal, wherein the single crystal is either a lithium tantalate single crystal or a lithium niobate single crystal.

10. A furnace to isolate it from external influences; a crucible provided in the furnace and into which a single crystal raw material is charged; a heating means for heating and melting the single crystal raw material placed in the crucible to form a raw material melt; a lifting means for lifting and lowering the seed crystal so as to carry out a seeding operation of bringing the seed crystal into contact with the raw material melt in the crucible and to grow a single crystal by pulling up the seed crystal while rotating it; a temperature detection means provided at the bottom of the crucible for detecting the temperature of the bottom of the crucible; a control means for controlling the seeding operation and the lifting operation of the lifting means based on the temperature information detected by the temperature detection means; Equipped with the control means is an acquisition means for acquiring the temperature of the bottom of the crucible during the seeding operation for each number of times of growing the single crystal by the temperature detection means; a prediction means for predicting a crucible bottom temperature at which the seeding operation is to be started when the number of times the single crystal has been grown exceeds a predetermined number n; an execution means for starting the seeding operation when the crucible bottom temperature reaches the predicted crucible bottom temperature predicted by the prediction means; Equipped with the prediction means defines X as the number of times the single crystal is grown repeatedly in the same furnace, Y as the temperature at the bottom of the crucible during the seeding operation, and Yp as the predicted temperature at the bottom of the crucible during the seeding operation when the number of times X=p of growth is to be performed; an information selection unit that, when the number of times X of growth exceeds a predetermined number n (n<p), selects, from the information acquired by the acquisition means, a group of Xn times of single crystal growth immediately preceding the number of times X of growth and a group of Yn times of crucible bottom temperatures during seeding operations; a relational equation creating unit that calculates a relational equation (I): Y=aX+b between the number of times X to grow the single crystal and the crucible bottom temperature Y during the seeding operation using a least squares method from the group of number of times X to grow the single crystal and the group of crucible bottom temperatures Y selected by the information selecting unit; a calculation unit that determines the crucible bottom temperature Y, calculated by substituting the number of times X (=p) of single crystal growth to be performed into the number of times X (=aX+b) of single crystal growth in the relational expression (I): Y=aX+b, as a predicted crucible bottom temperature Yp; A single crystal growth apparatus comprising:

11. The single crystal growth apparatus according to claim 10, The information selection unit, when selecting the Xn group of the most recent n number of single crystal growths and the Yn group of crucible bottom temperatures during the seeding operation, includes a correction processing unit that, when the ith Yn(i) among the Yn group of crucible bottom temperatures during the seeding operation includes one in which the seeding operation was performed using the predicted crucible bottom temperature Yp, excludes the i-th Yn(i) from the data candidates for the Xn group of single crystal growths if the difference between the Yn(i) and the predicted crucible bottom temperature Yp(i) used at that time is outside the allowable range.

Images