JPS589793B2 - How to grow band crystals - Google Patents

Description

[Detailed description of the invention] The present invention relates to a method for growing band-shaped crystals.

Taking silicon as an example, there are two ways to grow band-shaped crystals. As shown in FIG. 1, polycrystalline silicon is melted in a quartz glass crucible 11 to form a melt 12. Two dies 13 having slits 13a made of an inorganic material that is not easily deformed by the melt and wetted by the melt are provided, and a seed crystal 14 is used to make the melt in the dies 13a made up of the two dies 13 similar to the melt. After that, slit 13
There is a method in which a band-shaped silicon 15 having a cross section substantially the same as that of the opening a is grown.

Conventionally, in the method of growing band-shaped crystals, a thin seed crystal 14 has been used as shown in FIG.

This thin seed crystal 14 is advantageous because of its small cross section, especially when providing a constriction in the growing crystal and when the difficulty of processing the seed crystal itself is compared with those of other shapes, but it is made up of two dies. It was extremely difficult to widen the width of the band-shaped crystal on the melt surface defined by the slit.

In order to obtain a band-shaped crystal equal to the width of the slit from this seed crystal, a triangular or fan-shaped portion in which the width gradually increases is required as an intermediate state.

For example, in order for a band-shaped crystal to be used as a substrate for a solar cell element, it is preferably rectangular, that is, a shape with a constant width, and the fan-shaped portion must be discarded as material loss for the band-shaped crystal.

Reducing material loss corresponds to increasing utilization efficiency, and it has been strongly desired that the number of fan-shaped or triangular portions be reduced, and that the shape extends over a shorter distance to the width of the band-shaped crystal than the thin seed crystal.

In reality, contrary to the fact that the width increases over a short distance, the apex angle of the fan shape is extremely small, and if the apex angle is increased to widen the width of the crystal band, the growing crystal will stick to the top of the die and the growth will continue. The drawback of the conventional technology was that it was impossible to do so.

Another drawback is that when thin seed crystals are used, the slit spacing increases due to capillary action, and the temperature distribution of the melt determined by the shape of the opening deviates from the uniform temperature distribution considered necessary for band-like growth. If the seed crystal is thin when there are local highs and lows, it will start growing without correcting the temperature distribution of the die.

In this case, the effect of widening the width as seen in FIG. 1 was no longer effective in localized high temperature areas, making it impossible to grow a band-shaped crystal with the same shape as the slit of the die, that is, the intended width.

The present invention has been made in view of the above-mentioned problems, and provides a method for growing band-shaped crystals that has fewer triangular or sector-shaped growing crystal parts and does not cause the grown crystals to stick to the upper end of the slit of the die. .

The present invention will be described in detail below with reference to Examples.

FIG. 2 is a configuration diagram showing a growth mode of a band-shaped crystal for explaining an embodiment of the present invention.

High purity graphite with a specific gravity of 1.8, width 30mm and height 30mm
It was cut out to a thickness of 3 mm, the tips were obliquely ground at an angle of 30 degrees, and two dies 25 having slits 25a with a spacing of 0.03 mm were formed using graphite screws of the same material with a spacer interposed therebetween.

These two dies 25 were placed in a quartz glass crucible 27 with an inner diameter of 50 mmφ.

The resistivity of the quartz glass crucible other than die 25 is 5.
40 gr of high purity raw material silicon grains of 0 Ωcm were filled, and 5 μg of As for doping was mixed.

Next, the crucible was set in a pulling furnace, and the temperature inside the furnace was raised.

After about 1 hour, the temperature inside the furnace rose to 1450° C., and the high-purity raw material silicon and doping As were melted to form a melt 26.

When the temperature inside the furnace was further increased to 1500° C., the melt 26 began to rise between the slits 25a made up of two dies, and filled the slit openings 24 after 10 minutes.

When heating is slightly lowered to create the temperature conditions necessary for growth, the temperature at the end near the center axis is low and the temperature at the end near the quartz glass crucible 27 is high, due to the arrangement of the two dies 25 that are offset from the center axis. I understand.

Measurement using an optical thermometer showed that the temperature was 1405°C on the central axis side and 1412°C on the crucible side.

A thin seed crystal of 2 mm square with a sharp tip of 0.2 mm or less is attached to the pulling shaft 21 in the seed crystal mounting part, and the seed crystal is determined by hitting the melt in the slit opening 24 near the central axis of the low temperature part. Hold the time as it is, then 1m/min.
Pulling was started at a pulling speed of m, and the heating conditions of the furnace were changed so as to spread the width of the band-shaped crystal.

After repeating the same experiment six times, he found that the maximum spread angle, or the apex of the triangle, was 3 degrees.

It has also been found that the 3 degree triangular region does not continue to the end, and the action of expanding the width a stops at a location 10 mm from the edge of the quartz glass crucible.

The final shape was a triangular part having a height of 400 mm above the seed crystal, followed by a band-shaped crystal with a width of 20 mm.

Its electrical properties were N-type 0.1 Ωcm, which was satisfactory as a substrate for solar cells.

Next, open the seed crystal attachment part, width 29mm length 100m
A plate-shaped seed crystal with a thickness of 0.2 mm was attached to the pulling shaft 21.

The pulling shaft 21 was lowered, hit the melt in the slit opening 24, and was pulled up under the same conditions as those for pulling up the thin seed crystal described above.

Although there was a temperature difference of 7° C. between the central axis side of the slit opening 24 and the crucible side as described above, the width of the ribbon crystal was not biased to one side and spread over the entire slit opening 24.

At the same time as it spread, the band-shaped crystals began to crack and the cracks progressed and did not grow.

The reason why the band-shaped crystal cracks and breaks is because the slit opening 24
This is thought to be due to insufficient mechanical strength due to the use of thinner tabular seed crystals.

Next, open the seed crystal mounting part and make a width of 30 mm or more and a length of 1.
00mm, thickness is 0.2mm at the tip and 4mm at the rear end.
A seed crystal 22 having a wedge-shaped cross section was attached to the pulling shaft 21.

For attachment, since the seed crystal 22 has a wedge-shaped cross section, a holding member whose opening at the tip becomes narrower as it approaches the tip was connected to the lower part of the pulling shaft 21.

The pulling shaft 21 was lowered, hit the slit opening 24, and was pulled up under the same conditions as those used for pulling up thin seed crystals.

As a result, even though there was a temperature difference of 7°C between the central axis side of the slit opening 24 and the crucible side, the width of the band-shaped crystal 22 was not biased to one side, but spread over the entire slit opening 24, and the length A band-shaped crystal of 800 mm could be grown with a yield of 80%.

The reason why we were able to grow with a high yield of 80% is firstly because we used a seed crystal with a width of 30mm or more in the melt of the 30mm wide slit, and secondly because when we use flat seed crystals, band-shaped crystals are easily broken. However, by increasing the thickness of the upper part and using a seed crystal with a wedge-shaped vertical cross section, the mechanical strength was increased and it became possible to grow without cracking.

The electrical characteristics were N type and 0.1 Ωcm.

Although the above embodiments have been explained using silicon as an example, it goes without saying that the present invention can also be applied to Ge, -V compound semiconductors, and -■ compound semiconductors.

In addition, although graphite, which is a substance that wets and uses capillary action, is used as an example of the material of the die having slits, the content of the invention described above can be realized even with non-wettable substances by using gas pressure, etc. in combination.

Furthermore, the die having the slit is not limited to two flat plates, and may be a hollow die.

[Brief explanation of drawings]

FIG. 1 is a diagram for explaining a conventional method for growing a band-shaped crystal, and FIG. 2 is a diagram for explaining a method for growing a band-shaped crystal according to an embodiment of the present invention. 21... Pulling shaft, 22... Wedge-shaped seed crystal, 23... Silicon band crystal, 24...
...Slit opening, 25...Slit, 26
... Silicon melt, 27 ... Quartz glass crucible.

Claims (1)

[Claims] 1. A die having a slit is disposed in a melt made of a desired crystalline substance, a seed crystal is brought into contact with the melt through the slit, and the seed crystal is pulled up to grow band-shaped crystals. A method for growing a band-shaped crystal, characterized in that the width of the lowermost end of the seed crystal is greater than the width of the slit of the die so that it contacts the entire surface of the melt in the width direction. 2. The method for growing a band-shaped crystal according to claim 1, wherein the vertical cross section of the seed crystal is wedge-shaped.