JP2975948B2 - Crystal growth method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a crystal growing method used for controlling a growing condition in producing a single crystal by a Czochralski method (hereinafter referred to as a CZ method).

[0002]

2. Description of the Related Art The CZ method is a method conventionally used for producing a single crystal, and its outline is as follows.
First, a crucible is provided inside a single crystal growing furnace, and raw materials are put into the crucible. The single crystal growing furnace is provided with a high-frequency heating coil on the periphery thereof, thereby heating the raw material with high frequency to form a melt. On the other hand, a single crystal is manufactured by immersing a seed crystal in the melt and pulling the seed crystal at a predetermined speed.
The heating output of the crucible is controlled by the variation of the weight of the grown crystal, and the temperature was detected by measuring the temperature variation of the bottom of the crucible and the surface of the melt with a thermocouple or a radiation thermometer.

[0003]

As is apparent from the above description, the accuracy of temperature measurement is low in the conventional method. Therefore, these measurement results are used as one condition for starting the growth, and are mostly measured as reference data for output control during crystal growth. However, in order to realize the growth of high quality crystals, it is necessary to evaluate and monitor and control the melt state that changes with the growth of the crystals. The present invention has been made in view of the above circumstances, and has as its object to provide a crystal growing method capable of growing even higher quality crystals.

[0004]

In the present invention, the main point is to evaluate the structure (referred to as a cluster) of a melt using the magnitude of fluctuations such as temperature and concentration. That is, in the case of a component such as a melt and a material which causes a composition fluctuation, the temperature and concentration thereof fluctuate with respect to their absolute values. It is considered that the magnitude of this fluctuation can be detected as a potential difference, and by evaluating the cluster state of the melt from the magnitude of the fluctuation, a high quality of It grows a single crystal. FIG. 1 is a diagram showing a system for evaluating the state of a melt cluster, in which a pair of electrodes 1 is inserted into a melt 3 filled in a crucible 2. 4 is a voltage probe and 5 is a current probe, which are connected to oscilloscopes 6 and 7, respectively. 8 is a high frequency oscillation coil. Then, the voltage amplitude value generated between the electrodes is measured by an oscilloscope via the voltage probe 4 and the current amplitude value is also measured by the oscilloscope 7 via the current probe 5 to evaluate the melt structure. Things. In the present invention, although no external input is applied between the measurement electrodes,
When changing the signal level between the electrodes or analyzing the signal, it is necessary to add a constant voltage or constant current application circuit.

[0005] In the above structure, showing an experimental example using L _i N _b O ₃ melt. First, the behavior of the voltage and current between the electrodes in the melt according to the change in the high-frequency oscillation output was as follows. In the case of the voltage, the amplitude increased when the high-frequency oscillation output was lowered, and particularly, when the melt surface was solidified, a sharp amplitude fluctuation was observed. Also, in the case of current, it was recognized that when the high-frequency oscillation output was lowered, the amplitude was reduced. Therefore, the voltage and current generated between the electrodes are detected, and the cluster state of the melt can be evaluated from the amplitude value. In addition, since these detections can be performed during the growth of the crystal, it is possible to feed back the detection results to the growth control.

[0006]

An embodiment will be described below with reference to the drawings. FIG.
1 is a growth control system diagram of an embodiment for explaining a crystal growth method according to the present invention. 2, the same reference numerals are given to the same functional portions as those in FIG. Reference numeral 10 denotes a melt cluster state evaluation system, which is as described in FIG. Reference numeral 11 denotes a load cell which pulls up the pulling shaft 12 and raises the growing crystal 13 at the tip thereof. Data from the load cell 11 is input to a first automatic diameter control system 14. Reference numeral 15 denotes a second automatic diameter control system, which inputs data from the melt cluster state evaluation system 10 and the first automatic diameter control system 14. Reference numeral 16 denotes a high-frequency oscillating device, which inputs to the high-frequency oscillating coil 8 based on an input from the second automatic diameter control system 15.

[0007] In the above system, the crucible 2 is _Pt φ
Used as a 50 mm, filled with 85% by melt L _i N _b O _3. As the electrode 1 P _t - a _{(P t + R h 30%} ) of a pair away distance 5mm is inserted into the surface center of the melt 3, the voltage generated between the line, measures the variation width of the current (amplitude) did. A high frequency induction heating type CZ furnace was used for heating the crucible, and the oscillation frequency was about 200 kHz. Therefore, a device with a large measurement frequency band capable of evaluating a signal of 200 kHz was used for the measurement. An AC voltage probe that converts the value to 1/1000 was used for voltage measurement, and the signal was measured with an oscilloscope. An AC current probe was used for current measurement, and the signal was measured with an oscilloscope. First, in order to confirm the operation, the signal behavior was examined in three cases where the tip of the electrode was positioned above the melt, right above the melt surface, and inside the melt.

The change in the voltage signal is as shown in FIG. 3, and the change in the current signal is as shown in FIG. In each figure, (a) shows the case above the melt, (b) shows just above the melt surface, and (c) shows the case inside the melt. In the case of FIGS. 3A and 3C, the vertical axis is 10 V / unit, and the horizontal axis is 5 μs / unit. In the case of FIG. 3B, the vertical axis is 50 V / unit.
The horizontal axis is 1 ms / unit. In FIG. 4, the vertical axis is 40 mA / unit and the horizontal axis is 10 μs / unit for all of (a), (b) and (c). As can be seen from the figures, for the signal above the melt, almost no change in the signal amplitude due to the distance from the liquid surface at the electrode tip and the magnitude of the high-frequency oscillation output was observed for both the voltage and current ( 3 (a) and 4 (a)). Next, in the case directly above the melt surface, as can be seen from the figure, a very large amplitude is shown, and the signal becomes very unstable. When observing the electrode tip, the time when such a signal is shown coincides with the time when the electrode tip emits light, which is considered to be due to discharge occurring between the melt surface and the electrode (FIG. 3 (b) , Fig. 4 (b)
). The behavior of the signal when the electrode tip is inserted into the melt will be described. The moment the electrode tip touches the melt,
As shown in FIG. 3 (c) and FIG. 4 (c), the signal associated with the discharge immediately above the melt surface changed, the amplitude of the voltage decreased, and the amplitude of the current increased. It was confirmed that as the high-frequency oscillation output was lowered, the amplitude value of the voltage was increased, and the rate of increase was particularly large when the melt surface was solidified. In addition, it is recognized that the amplitude value of the current decreases as the high-frequency oscillation output decreases, and these results are summarized in FIG.

The behavior of the amplitude of the voltage and current due to the change of the high-frequency oscillation output is interpreted as follows. The amplitude value of the voltage is the difference between the potential fluctuations at both electrodes,
This is considered to be determined by the fluctuation of the magnitude of the electromotive force between the two electrodes depending on the electric field and the degree of relaxation of the potential difference due to the current transfer between the two electrodes. Although the electromotive force between both electrodes seems to depend on the magnitude of the high-frequency oscillation output and the distance between the electrodes, the result that the amplitude value of the potential difference between the two electrodes above the melt did not depend on the magnitude of the high-frequency oscillation output Considering
It seems that it hardly depends on the value of the high-frequency oscillation output. Also, the electric mobility between the two electrodes is considered to be higher as the melt temperature is higher, and thus the electric mobility is higher as the high-frequency oscillation output is higher, and the potential difference between the two electrodes is more buffered. It can be interpreted that it becomes smaller.

Next, an embodiment in which a crystal is actually grown by using this method will be described. Is shown in this embodiment as the single crystal growth example L _i T _a O _3. In this case, a set of electrodes is inserted into the surface of the melt, the amplitude of the potential difference between the electrodes is measured, and the effectiveness of growing the crystal while suppressing the fluctuation of the amplitude is confirmed. the purpose, L _i T _a O ₃ employed the method in the single crystal growing, L _i already known
I tried to suppress the bending phenomenon of T _a O ₃ crystal. First, the crucible 2 of _{φ = 50mm L i T a O} 3 raw material to 400g
Filling was performed and crystal growth was performed using the hot zone shown in FIG. 6, the same functional portions as those in FIG. 2 are denoted by the same reference numerals. 9 is an after heater, and 17 is a heat insulating bubble.

Here, a growth experiment was carried out while controlling the rotation speed and the pulling speed of the crystal pulling shaft at 20 rpm and 3 mm / h and controlling the high frequency oscillation output based on an automatic diameter control (ADC) system. The two I _r is the potential measurement of the melt surface
Using a wire, the electrode was placed at a position of about 5 mm from the crucible wall and at a distance of about 5 mm. The amplitude of the electromotive force generated between both electrodes (hereinafter referred to as k value) is measured, and the rate of change of this k value is
When it exceeded 10%, the crystal rotation speed was increased by 10%, and the growth was performed using the ADC system while repeating this operation. FIG. 7 shows the change in the k value depending on the crystal pulling length when the crystal rotation speed is controlled based on the k value using the ADC system and when it is not. In the figure,
Indicates a case where crystal rotation control is not performed, and indicates a case where crystal rotation control is performed.

When the number of rotations of the crystal is not controlled, the k value increases from the point where the crystal pulling length is about 4 cm and becomes about twice when the crystal is grown about 7 cm. Was shown. K when crystal rotation is controlled
The rate of change of the value is small for the growth length of the crystal,
The k value at the time of pulling up by 10 cm was about 1.5 times that at the time of the crystal length of 4 cm, and the bending phenomenon occurred at about 11 cm. in this way
It was found that the bending phenomenon, which was difficult to suppress under the ADC system, can be suppressed by controlling the k value on the melt surface.

[0013]

As described above, according to the present invention, with respect to a material such as a melt or the like which causes a composition variation, the voltage and current generated between the electrodes are detected, and the structure of the melt is determined from the amplitude value. Was evaluated, so that the melt structure, which could not be obtained by measuring only the temperature, could be kept constant, and a high-quality single crystal became possible.

[Brief description of the drawings]

FIG. 1 is a system diagram of a melt cluster state evaluation system.

FIG. 2 is a growth control system diagram of an embodiment for explaining a crystal growth method according to the present invention.

FIG. 3 is a change diagram of a voltage signal between electrodes.

FIG. 4 is a change diagram of a current signal between electrodes.

FIG. 5 is a diagram illustrating a relationship between a high-frequency oscillation output and an amplitude value.

FIG. 6 is a schematic diagram of a hot zone.

FIG. 7 is a relationship diagram between a crystal growth length and a potential difference amplitude value.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 electrode 2 crucible 3 melt 4 voltage probe 5 current probe 6,7 oscilloscope 8 high frequency oscillation coil 9 after heater 10 melt cluster state evaluation system 11 load cell 12 pull-up shaft 13 growing crystal 14,15 automatic diameter control system 16 high frequency Oscillator

Claims (2)

(57) [Claims] 1. A crystal growing method for producing a single crystal from a raw material melt by the Czochralski method, wherein a predetermined interval is provided on a surface of the melt heated by a high-frequency oscillator.
A method for growing a crystal, comprising inserting a pair of electrodes, detecting respective amplitude values of a voltage and a current generated between the electrodes, and feeding back the amplitude values to a high-frequency oscillator. 2. The crystal growing method according to claim 1, wherein the control by feeding back to the high-frequency oscillator is performed so as to keep the melt structure constant.