JP7823596B2 - Quartz glass crucible and its manufacturing method - Google Patents

Description

The present invention relates to a quartz glass crucible used for pulling silicon single crystals using the Czochralski method (CZ method) and a method for manufacturing the same.

Most silicon single crystals used as substrate materials for semiconductor devices are manufactured using the CZ method. In the CZ method, polycrystalline silicon raw material is melted in a quartz glass crucible to produce a silicon melt. A seed crystal is then immersed in the silicon melt and gradually pulled up while rotating the quartz glass crucible and seed crystal, growing a large single crystal at the bottom of the seed crystal. The CZ method can increase the yield of large-diameter silicon single crystals.

A quartz glass crucible (silica glass crucible) is a container made of silica glass that holds molten silicon. Therefore, quartz glass crucibles must be highly durable enough to withstand long-term use and not deform at high temperatures above the melting point of silicon. They must also be resistant to cracking and chipping even at room temperature.

Regarding technology for increasing the strength of quartz glass crucibles, for example, Patent Document 1 describes a method for preventing inward collapse of the crucible wall due to differences in thermal expansion coefficient and glass density by rapidly cooling the inner surface of the crucible to increase the fictive temperature of the inner surface and by keeping the outer surface warm using residual heat after arc melting to lower the fictive temperature of the outer surface.

Patent Document 2 also describes a silica glass crucible in which a first region, a second region, and a third region are provided in that order from the inner surface of the crucible toward the outer surface, and the internal residual stress in the first and third regions is compressive stress, while the internal residual stress in the second region is tensile stress.

Japanese Patent Application Laid-Open No. 2008-297154 Japanese Patent Application Laid-Open No. 2019-151494

A large amount of polycrystalline silicon raw material is filled into a quartz glass crucible, and a large load is placed on the inner surface of the crucible, causing damage to the inner surface. In particular, because the individual polycrystalline silicon chunks are crushed into small pieces during the manufacturing process, resulting in sharp corners, the load is concentrated on the sharp tips of the polycrystalline silicon chunks, making it easy for cracks and breakage to occur on the inner surface of the crucible.

In addition, when the crucible is being transported, there is a possibility that something may hit the outer surface of the crucible, and if the outer surface of the crucible is not strong enough, cracks or breaks may occur on the outer surface of the crucible.

Therefore, an object of the present invention is to provide a quartz glass crucible with increased strength at least on its inner surface, and a method for manufacturing the same.

To solve the above problems, the quartz glass crucible of the present invention has a cylindrical side wall, a bottom, and a corner portion between the side wall and the bottom that has a greater curvature than the bottom, and is characterized in that the fictive temperature of the inner surface of the side wall is at least 30°C lower than the fictive temperature of the interior at a depth of at least 3 mm from the inner surface.

This invention stabilizes the Si-O structure in the glass matrix near the inner surface of the sidewall, increasing the strength of the crucible's inner surface. This prevents the crucible from cracking when filled with polycrystalline silicon.

In the present invention, the fictive temperature of the inner surface of the side wall portion is preferably at least 50°C lower than the fictive temperature of the inner portion. This further increases the strength of the inner surface of the crucible.

In the present invention, the OH group concentration on the inner surface of the crucible is preferably 15 ppm to 70 ppm. This makes it possible to realize a quartz glass crucible in which the fictive temperature on the inner surface of the side wall is sufficiently lower than the fictive temperature inside the crucible.

In the present invention, the fictive temperature of the outer surface of the side wall is preferably at least 30°C lower than the fictive temperature of the outer interior at a depth of at least 3 mm from the outer surface. This stabilizes the Si-O structure in the glass matrix near the outer surface of the side wall, thereby increasing the strength of the crucible outer surface.

In the present invention, the fictive temperature of the outer surface of the crucible is preferably at least 50°C lower than the fictive temperature of the inner exterior. This further increases the strength of the outer surface of the crucible.

The method for manufacturing a quartz glass crucible according to the present invention comprises the steps of: forming a deposition layer of raw silica powder on the inner surface of a carbon mold that matches the outer shape of the quartz glass crucible; the carbon mold having a cylindrical side wall, a bottom provided below the side wall, and a corner provided between the side wall and the bottom and having a greater curvature than the bottom; arc-melting the deposition layer of raw silica powder from inside the carbon mold; and cooling the fused quartz glass crucible formed within the carbon mold; wherein the cooling step includes the steps of: immediately after completion of arc-melting of the raw silica powder, holding a high-temperature arc electrode within the carbon mold for a certain period of time to slowly cool the fused quartz glass crucible within the carbon mold; and, after the slow cooling, retracting the arc electrode above the carbon mold to increase the cooling rate of the quartz glass crucible.

The present invention makes it possible to manufacture a quartz glass crucible with a reinforced inner surface.

In the present invention, the process of arc-melting the raw silica powder preferably includes a process of cooling the carbon mold, and the process of cooling the quartz glass crucible preferably includes a process of stopping cooling of the carbon mold for a certain period of time and a process of restarting cooling of the carbon mold. This makes it possible to manufacture a quartz glass crucible with a reinforced outer surface.

In the present invention, the step of cooling the quartz glass crucible preferably includes a step of humidifying the chamber in time with the start of the slow cooling. This makes it possible to produce a quartz glass crucible in which the fictive temperatures of the inner and outer surfaces of the crucible are sufficiently lower than the inner and outer interiors.

The present invention provides a quartz glass crucible with increased strength at least on its inner surface, as well as a method for manufacturing the same.

FIG. 1 is a schematic perspective view showing the configuration of a quartz glass crucible according to an embodiment of the present invention. FIG. 2 is a schematic cross-sectional side view of the quartz glass crucible shown in FIG. FIG. 3 is a schematic diagram for explaining the fictive temperature distribution of the crucible wall of the vitreous silica crucible according to this embodiment. FIG. 4 is a schematic diagram showing a method for manufacturing a quartz glass crucible by the rotating mold method. FIG. 5 is a sequence diagram showing the details of the process from the arc melting step to the cooling step. FIG. 6 is a graph showing the measurement results of the fictive temperature distribution of the quartz glass crucible according to Example A1. FIG. 7 is a graph showing the measurement results of the fictive temperature distribution of the quartz glass crucible according to Example A2.

A preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings.

Figure 1 is a schematic perspective view showing the configuration of a quartz glass crucible according to an embodiment of the present invention. Figure 2 is a schematic side cross-sectional view of the quartz glass crucible shown in Figure 1.

As shown in Figures 1 and 2, quartz glass crucible 1 is a container made of silica glass for holding silicon melt. It has a cylindrical side wall 10a, a bottom 10b located below side wall 10a, and a corner 10c located between side wall 10a and bottom 10b. Bottom 10b is preferably a gently curved, so-called rounded bottom, but may also be a so-called flat bottom. Corner 10c has a greater curvature than bottom 10b. The boundary between side wall 10a and corner 10c and the boundary between bottom 10b and corner 10c can be determined from the point where the curvature of the outer surface changes from a small curvature to a large curvature.

The bore (diameter) of the quartz glass crucible 1 varies depending on the diameter of the silicon single crystal ingot being pulled from the silicon melt, but is typically 18 inches (approximately 450 mm) or larger, with 22 inches (approximately 560 mm) being preferred, and 32 inches (approximately 800 mm) or larger being particularly preferred. This is because such large crucibles are used to pull large silicon single crystal ingots with a diameter of 300 mm or larger, and are required to be able to withstand long-term use without affecting the quality of the single crystal.

The thickness of the crucible varies slightly depending on the location, but is preferably 8 to 20 mm. In particular, it is preferable that the thickness of the side wall 10a of an 18-inch crucible be 8 to 9 mm, the thickness of the side wall 10a of a 24-inch crucible be 10 to 12 mm, the thickness of the side wall 10a of a 28-inch or larger crucible be 13 to 15 mm, and the thickness of the side wall 10a of a 32-inch or larger crucible be 17 to 18 mm. This allows a large amount of silicon melt to be stably held at high temperatures. It is preferable that the thickness of the crucible be thickest at the corners 10c, and that the side wall 10a and bottom 10b be thinner than the corners 10c.

As shown in Figure 2, the quartz glass crucible 1 mainly has a two-layer structure, comprising a transparent layer 11 (bubble-free layer) made of silica glass that is substantially bubble-free, and a bubble layer 12 (opaque layer) made of silica glass that contains numerous tiny bubbles.

The transparent layer 11 is a glass layer that constitutes the inner surface 10i of the crucible that comes into contact with the silicon melt. It is provided to prevent a decrease in silicon single crystal yield due to bubbles in the silica glass. Because the inner surface 10i of the quartz glass crucible 1 reacts with the silicon melt and is dissolved, bubbles near the inner surface cannot be contained within the silica glass. Thermal expansion may cause the bubbles to burst, resulting in crucible fragments (silica fragments) peeling off. If crucible fragments released into the silicon melt are carried by melt convection to the growth interface of the silicon single crystal and are incorporated into the silicon single crystal, they can cause dislocations in the silicon single crystal. Furthermore, if bubbles released into the silicon melt float up to the solid-liquid interface and are incorporated into the single crystal, they can cause pinholes to form in the silicon single crystal.

The phrase "substantially bubble-free" means that the transparent layer 11 may have a bubble content and bubble size that do not reduce the single crystal yield due to bubbles. For example, the bubble content is 0.1 vol% or less, and the bubble diameter is 100 μm or less.

The bubble content per unit volume can be calculated by integrating the bubble content per unit area in the depth direction. To detect bubbles present at a certain depth from the surface, the focus of an optical lens can be scanned from the surface in the depth direction. The bubble content per unit area can be calculated by dividing an image of the inner surface of the crucible taken with a digital camera into fixed areas and using these as reference areas, and then calculating the bubble content per unit area as the ratio of the area occupied by bubbles to this reference area.

The transparent layer 11 is preferably 0.5 to 10 mm thick, and is set to an appropriate thickness for each part of the crucible so that it does not completely disappear due to melting during the crystal pulling process, exposing the bubble layer 12. It is preferable that the transparent layer 11 be provided over the entire crucible, from the side wall 10a to the bottom 10b, but it is also possible to omit the transparent layer 11 from the upper end (rim) of the crucible, which does not come into contact with the silicon melt.

The bubble content and bubble diameter of the transparent layer 11 can be measured non-destructively using optical detection means. The optical detection means includes a light receiving device that receives transmitted or reflected light irradiated onto the crucible. The light receiving device can be a digital camera equipped with an optical lens and an image sensor. Irradiated light can be visible light, ultraviolet light, infrared light, X-rays, laser light, or other light. The measurement results obtained by the optical detection means are input into an image processing device, and the bubble diameter and bubble content per unit volume are calculated.

The bubble layer 12 is the main glass layer of the quartz glass crucible 1, located outside the transparent layer 11. It is provided to improve the heat retention of the silicon melt in the crucible and to distribute radiant heat from the heater of the single crystal pulling device to heat the silicon melt in the crucible as uniformly as possible. For this reason, the bubble layer 12 is provided throughout the entire crucible, from the side wall 10a to the bottom 10b. The thickness of the bubble layer 12 is approximately equal to the thickness of the quartz glass crucible 1 minus the thickness of the transparent layer 11, and varies depending on the part of the crucible. The temperature of the silicon melt is determined by the extent to which the bubble layer 12 transmits radiant heat from the heater, so the state of the bubbles inside the crucible (number, size, and density of bubbles) is important.

The bubble content of the bubble layer 12 is preferably higher than that of the transparent layer 11, greater than 0.1 vol% and less than 5 vol%. If the bubble content of the bubble layer 12 is less than 0.1 vol%, the bubble layer 12 will not exhibit the required heat retention function. Furthermore, if the bubble content of the bubble layer 12 exceeds 5 vol%, the thermal expansion of the bubbles may cause deformation of the crucible, resulting in a decrease in single crystal yield and insufficient heat transfer. From the perspective of balancing heat retention and heat transfer, a bubble content of 1 to 4 vol% is particularly preferred. The above bubble content is a value measured on an unused crucible at room temperature. The bubble content of the bubble layer 12 can be determined, for example, by measuring the specific gravity (Archimedes' method) of an opaque silica glass piece cut from the crucible.

In order to increase the strength of the quartz glass crucible 1 at room temperature, this invention focuses on the fictive temperature of the silica glass that makes up the crucible. The fictive temperature refers to the freezing temperature of a supercooled liquid. When cooled to a temperature below the melting point at a rate fast enough to prevent crystallization, structural relaxation keeps up with the cooling rate at temperatures near the melting point, and the glass structure always reaches the most stable state of the supercooled liquid. However, as the temperature drops further, structural relaxation becomes unable to keep up with the cooling, and the glass structure ultimately fails to reach the thermal equilibrium state of the supercooled liquid and remains in a non-equilibrium state. The fictive temperature indicates the temperature at which the glass structure corresponds to the stable structure ((quasi) thermal equilibrium state) of the supercooled liquid.

The volume of a substance decreases continuously as the temperature decreases from a high temperature, but the volume changes discontinuously when the substance changes phase from liquid to crystalline solid. However, in some substances, the change in volume from liquid to solid occurs continuously, and the substance may become solid while remaining in an amorphous state. This continuous change in volume from liquid to amorphous solid is called the glass transition, and the state below this temperature range is called the glassy state. In other words, glass is defined as an amorphous solid that exhibits the glass transition phenomenon. Unlike crystals, glass is not a thermodynamically stable state, but rather has a disordered structure without long-range regularity. For this reason, glass easily changes structure when external energy is applied, transitioning to a more stable state and constantly changing structure.

As mentioned above, even if the glass has the exact same composition, its structure will differ depending on the freezing temperature. If the cooling rate is slow, there is enough time for the structure to relax, and the fictive temperature will decrease in line with the actual temperature. On the other hand, if the cooling rate is fast, the fictive temperature will quickly deviate from the actual temperature of the glass. Compared to slowly cooled glass, rapidly cooled glass freezes its high-temperature structure, making it more unstable and more susceptible to structural changes. It is also more likely to exhibit properties such as transparency, brittleness, and continuity of physical properties.

Figure 3 is a schematic diagram illustrating the fictive temperature distribution on the crucible wall of a quartz glass crucible according to this embodiment.

As shown in Figure 3, the quartz glass crucible 1 according to this embodiment is characterized in that, when comparing the fictive temperature of the crucible surface with the fictive temperature of the interior of the crucible at least 3 mm deeper, the fictive temperature of the crucible surface is lower than the fictive temperature of the interior. Specifically, the fictive temperature _TVi1 of the silica glass constituting the inner surface 10i of the crucible is at least 30°C lower, and more preferably at least 50°C lower, than the fictive temperature _TVi2 of the inner interior at least 3 mm deeper. Furthermore, the fictive temperature TVo1 of the silica glass constituting the outer surface 10o of the _crucible is at least 30°C lower, and more preferably at least 50°C lower than the fictive temperature _TVo2 of the outer interior at least 3 mm deeper.

When measuring the fictive temperature T _Vi1 of the inner surface of the crucible, the depth region from the inner surface 10i to 1/10 of the crucible's wall thickness is defined as the measurement region di of the inner surface 10i of the crucible. Also, when measuring the fictive temperature inside the crucible relative to the inner surface 10i of the crucible, the depth region from the inner surface 10i of the crucible to a depth position of 1/3 to 2/3 of the crucible's wall thickness is defined as the measurement region dd inside the crucible. When measuring the inside of the crucible that is at least 3 mm deep from the inner surface 10i, it is sufficient to measure any of the measurement regions dd.

Furthermore, when measuring the fictive temperature T _Vo1 of the outer surface 10o of the crucible, a depth region from the outer surface 10o to 1/10 of the crucible's wall thickness is defined as the measurement region do of the outer surface 10o of the crucible. Furthermore, when measuring the fictive temperature inside the crucible relative to the outer surface 10o of the crucible, a depth region from the outer surface 10o to a depth position of 1/3 to 2/3 of the crucible's wall thickness is defined as the measurement region dd of the crucible's interior. This is the same region as the inside of the crucible relative to the inner surface 10i of the crucible. When measuring the outside interior at least 3 mm deep from the outer surface 10o, it is sufficient to measure any of the measurement regions dd.

As mentioned above, the thickness of a crucible varies depending on the size and location of the crucible, but the sidewall thickness of a crucible of 18 inches or larger is 8 mm or more, and the sidewall thickness of a 32-inch crucible is 17 mm or more. Therefore, for example, for a small-diameter crucible, the fictive temperature inside the crucible can be determined by measuring the fictive temperature at a position at least 3 mm deep from the inner surface 10i of the crucible. Also, the fictive temperature inside the crucible's exterior can be measured by measuring the fictive temperature at a position at least 3 mm deep from the outer surface 10o of the crucible.

The fictive temperature of silica glass can be measured using Raman spectroscopy or FT-IR. Raman spectroscopy involves irradiating the surface of the sample to be measured with laser light, and determining the area intensity ratio of the peaks resulting from the three-membered rings and four-membered rings, which are ring structures of Si, from the Raman (scattering) spectrum. The fictive temperature of the sample can be calculated by plotting the results on a calibration curve obtained from measurements of samples with known fictive temperatures.

When measuring fictive temperature using the FT-IR method, a thin sample of the glass to be measured is irradiated with laser light, and the peak wavelength derived from the quartz glass structure is detected from the transmission spectrum. The fictive temperature of the sample can be calculated by plotting the results on a calibration curve obtained from measurements of samples with known fictive temperatures.

At least in the sidewall portion, the OH group concentration of the silica glass that constitutes the inner surface 10i and outer surface 10o of the crucible is preferably 15 ppm to 70 ppm. By increasing the OH group concentration in the silica glass, the effect of reducing the fictive temperature of the silica glass can be enhanced. Therefore, it is possible to realize a silica glass crucible with a relatively low fictive temperature on the inner and outer surfaces of the crucible.

Next, we will explain the manufacturing method of the quartz glass crucible 1. The quartz glass crucible 1 according to this embodiment can be manufactured by the so-called rotational mold method.

Figure 4 is a schematic diagram showing the manufacturing method of a quartz glass crucible 1 using the rotating mold method.

As shown in Figure 4, in the rotating mold method, a carbon mold 14 is prepared with a cavity that matches the outer shape of the crucible, and natural silica powder 16a and synthetic silica powder 16b are filled in sequence along the inner surface 14i of the rotating carbon mold 14 to form a deposition layer 16 of raw silica powder. The raw silica powder remains in a fixed position, adhering to the inner surface 14i of the carbon mold 14 due to centrifugal force, and the crucible shape is maintained.

Next, the tip of the arc electrode 15 is placed inside the carbon mold 14, and the deposited layer 16 of raw silica powder is arc-melted from inside the carbon mold 14. Specific conditions such as heating time and heating temperature are determined appropriately taking into account the characteristics of the raw silica powder, the size of the crucible, etc.

During arc melting, the amount of bubbles in the molten quartz glass is controlled by evacuating the raw silica powder deposition layer 16 through numerous vent holes 14a provided on the inner surface 14i of the carbon mold 14. Specifically, at the start of arc melting, the raw silica powder deposition layer 16 is evacuated to form a transparent layer 11, and after the transparent layer is formed, the vacuum on the raw silica powder is stopped or the suction force is weakened to form a bubble layer 12.

Since the arc heat propagates from the inside to the outside of the deposited layer 16 of raw silica powder, melting the raw silica powder, it is possible to create either the transparent layer 11 or the bubble layer 12 by changing the vacuum conditions when the raw silica powder begins to melt. That is, if vacuum melting is performed, in which the vacuum is increased when the raw silica powder that forms the inner surface of the crucible melts, atmospheric gases are not trapped within the glass, and the fused silica becomes silica glass that does not contain bubbles. On the other hand, if normal melting (atmospheric pressure melting) is performed, in which the vacuum is decreased when the raw silica powder that forms the outer surface of the crucible melts, atmospheric gases are trapped within the glass, and the fused silica becomes silica glass that contains numerous bubbles.

Then, arc melting is terminated and the crucible is cooled. This completes the quartz glass crucible 1, with the transparent layer 11 and bubble layer 12 formed in that order from the inside to the outside of the crucible wall. The crucible is then shaped by cutting the rim, etc., and then washed with cleaning fluid and rinsed with pure water. The cleaning fluid is preferably prepared by diluting semiconductor-grade or higher hydrofluoric acid with pure water with a TOC of 2 ppb or less to a concentration of 10 to 40 wt %.

The fictive temperature of silica glass changes depending on the cooling temperature. Rapidly cooled silica glass has a higher fictive temperature, while slowly cooled silica glass has a lower fictive temperature. Because the arc electrode 15, which serves as the heat source during arc melting, is located on the inner surface of the crucible, cooling of the inner surface of the crucible begins immediately after arc heating ends. Meanwhile, the carbon mold 14 is located on the outer surface of the crucible, and remains at a high temperature even after arc heating ends. Here, by temporarily stopping the air or water cooling of the carbon mold 14, the cooling rate of the outer surface 10o of the quartz glass crucible 1 becomes slower than that of the inner surface 10i. Furthermore, because the outer surface 10o of the quartz glass crucible 1 is cooled more slowly than the interior, the fictive temperature of the outer surface 10o becomes lower than that of the interior. In this embodiment, the heat retention of the carbon mold 14 is improved by heating, etc., and the cooling rate on the outer surface 10o side of the quartz glass crucible 1 is slowed, thereby increasing the fictive temperature difference between the outer surface 10o of the quartz glass crucible 1 and the inside of the base by 30°C or more.

Figure 5 is a sequence diagram showing the details from the arc melting process to the cooling process.

As shown in Fig. 5, at the start of the arc melting process, the arc electrode 15 is placed at the discharge position shown in Fig. 4, and arc discharge (arc heating) starts at time _t0 . At this time, the carbon mold 14 is also cooled, thereby preventing the carbon mold 14 from overheating.

When the arc discharge ends at time _t1 , a certain time after the start of the arc discharge (the time until the raw silica powder is completely melted), in this embodiment, the arc electrode 15 is not moved from the discharge position to the retracted position, but is maintained at the discharge position. Conventionally, the arc electrode 15 is retracted above the carbon mold 14 immediately after the end of arc melting in order to rapidly cool the molten crucible. However, by maintaining the arc electrode at the discharge position, the crucible can be slowly cooled from the inner surface 10i side. This makes it possible to manufacture a crucible in which the fictive temperature of the inner surface 10i is relatively lower than that of the inside of the crucible.

Furthermore, when arc heating is completed, the forced cooling of the carbon mold 14 is temporarily stopped and the mold is allowed to cool naturally. Conventionally, the carbon mold 14 has always been forcedly cooled, but by temporarily stopping the forced cooling, the crucible can be slowly cooled from the outer surface 10o side. In other words, it is possible to manufacture a crucible in which the fictive temperature of the outer surface 10o is relatively lower than that of the inside of the outside.

Furthermore, at the end of the arc heating, humidification of the atmosphere in the processing chamber is initiated. The humidification conditions are preferably 8 g/ ^cm³ or more, and the humidification time is preferably 60 seconds or more. This increases the OH group concentration in the silica glass near the inner and outer surfaces of the crucible, thereby enhancing the effect of reducing the fictive temperature of the inner surface 10i and outer surface 10o of the crucible. This also has the effect of relaxing the glass structure, thereby further reducing the fictive temperature of the inner surface 10i and outer surface 10o of the crucible.

At time _t2 , a certain time (e.g., 60 seconds) after the start of the slow cooling of the crucible, the slow cooling of the crucible is completed and the arc electrode is moved to the waiting position. At this time, forced cooling of the carbon mold 14 is resumed to accelerate the cooling of the crucible. This allows the temperature of the crucible to be quickly lowered to room temperature, thereby shortening the time required to complete the arc melting process.

As explained above, the quartz glass crucible 1 according to this embodiment has a fictive temperature of the inner surface 10i at least in the side wall portion 10a that is 30°C or more lower than the fictive temperature inside the crucible at a depth of at least 5 mm from the inner surface 10i, thereby increasing the strength of the crucible's inner surface 10i at room temperature. Furthermore, the quartz glass crucible 1 according to this embodiment has a fictive temperature of the inner surface 10i at least in the side wall portion 10a that is 30°C or more lower than the fictive temperature inside the crucible at a depth of at least 5 mm from the outer surface 10o, thereby increasing the strength of the crucible's outer surface 10o at room temperature.

The above describes a preferred embodiment of the present invention, but the present invention is not limited to the above embodiment and various modifications are possible without departing from the spirit of the present invention, and it goes without saying that these modifications are also included within the scope of the present invention.

For example, in the above embodiment, both the fictive temperature of the inner surface 10i and the fictive temperature of the outer surface 10o of the crucible are lower than the fictive temperature inside the crucible, but it is also acceptable for only one of them to be lower. If only the inner surface of the crucible is like this, it is possible to improve the strength of only the inner surface of the crucible. Also, if only the outer surface of the crucible is like this, it is possible to improve the strength of only the outer surface of the crucible.

<Evaluation of fictive temperature distribution in a 32-inch crucible>
A 32-inch diameter quartz glass crucible was manufactured using the rotating mold method described above. During the arc melting process, the tip of the arc electrode was held in the mold immediately after the end of arc heating to slowly cool the inner surface of the crucible. Cooling of the carbon mold was also temporarily suspended to slowly cool the outer surface of the crucible. Furthermore, humidification was started immediately after the end of arc heating in the chamber.

The wall thickness, transparent layer thickness, and bubble layer thickness at the half-height position (side wall portion) of the quartz glass crucible of Example A1 obtained in this way were measured. The crucible wall thickness was 17.6 mm, the transparent layer thickness was 5.6 mm, and the bubble layer thickness was 12.0 mm.

Next, the fictive temperature distribution of the obtained quartz glass crucible was measured using Raman spectroscopy. The measurement position for the fictive temperature distribution was set at half the height of the quartz glass crucible. At this position, 1/3 of the crucible's wall thickness was 5.9 mm, and this value was used as the measurement position for the fictive temperature inside the crucible. That is, to measure the fictive temperature inside the crucible, the inner surface was polished by 5.9 mm, and then the polished surface was irradiated with laser light to measure the fictive temperature. To measure the fictive temperature inside the crucible, the outer surface was polished by 5.9 mm, and then the polished surface was irradiated with laser light to measure the fictive temperature.

Figure 6 is a graph showing the measurement results of the fictive temperature distribution of the quartz glass crucible according to Example A1, with the horizontal axis representing the distance from the inner surface (mm) and the vertical axis representing the fictive temperature (°C).

As shown in Figure 6, the fictive temperature of the inner surface was lower than the fictive temperature of the inner interior, and the fictive temperature of the outer surface was lower than the fictive temperature of the outer interior.

<Evaluation of fictive temperature distribution in a 24-inch crucible>
A 24-inch diameter quartz glass crucible was manufactured by the above-mentioned rotating mold method. During the arc melting process, the arc electrode was kept in the mold immediately after the end of arc heating to slowly cool the inner surface of the crucible. Cooling of the carbon mold was also temporarily stopped to slowly cool the outer surface of the crucible.

The wall thickness, transparent layer thickness, and bubble layer thickness at the half-height position (side wall portion) of the quartz glass crucible of Example A2 obtained in this way were measured. The crucible wall thickness was 11.5 mm, the transparent layer thickness was 7.5 mm, and the bubble layer thickness was 4.0 mm.

Next, the fictive temperature distribution of the obtained quartz glass crucible was measured using Raman spectroscopy. The measurement position for the fictive temperature distribution was set at half the height of the quartz glass crucible. At this position, 1/3 of the crucible's wall thickness was 3.8 mm, and this value was used as the measurement position for the fictive temperature inside the crucible. That is, to measure the fictive temperature inside the crucible, the inner surface was polished by 3.8 mm, and then the polished surface was irradiated with laser light to measure the fictive temperature. To measure the fictive temperature inside the crucible, the outer surface was polished by 3.8 mm, and then the polished surface was irradiated with laser light to measure the fictive temperature.

Figure 7 is a graph showing the measurement results of the fictive temperature distribution of the quartz glass crucible according to Example A2, with the horizontal axis representing the distance from the inner surface (mm) and the vertical axis representing the fictive temperature (°C).

As shown in Figure 7, even in the 24-inch crucible, the fictive temperature of the inner surface was lower than the fictive temperature of the inner interior, and the fictive temperature of the outer surface was lower than the fictive temperature of the outer interior.

<Evaluation of the inner surface of the crucible after pulling the crystal>
Five quartz glass crucible samples were prepared, and the fictive temperature of the inner surface of each crucible was measured. The quartz glass crucibles of Examples B1, B2, and B3 were measured using the rotating mold method described above, and in the arc melting process, the arc electrode was kept in the mold even immediately after the end of arc heating to slowly cool the inner surface of the crucible. The quartz glass crucibles of Comparative Examples B1 and B2 were manufactured using the conventional method of rapid cooling after the end of arc heating.

Next, a silicon single crystal was pulled using a crucible identical to this one, and the condition of the inner surface of the crucible after pulling was visually observed, and the number of silicon pieces that had become embedded in the inner surface of the crucible was evaluated. This embedding is thought to be caused by impacts that occurred when the inner surface of the crucible came into contact with the raw silicon material when the silicon material was being filled into the crucible or while it was being melted. The results are shown in Table 1.

As shown in Table 1, in Example B1, the fictive temperature difference was 30°C and the number of silicon bites was 1. In Example B2, the fictive temperature difference was 30°C and the number of silicon bites was 2. In Example B3, the fictive temperature difference was 50°C and the number of silicon bites was 0.

On the other hand, in Comparative Example B1, the fictive temperature difference was -10°C, and the number of silicon bites was 13. Furthermore, in Comparative Example B2, the fictive temperature difference was -30°C, and the number of silicon bites was 10. From these results, it was confirmed that the inner surfaces of the quartz glass crucibles of Examples B1, B2, and B3, in which the fictive temperature of the crucible inner surface was 30°C or more lower than that of the crucible interior, suffered less damage and had improved strength.

<Evaluation of the outer surface of the crucible after pulling the crystal>
Three quartz glass crucible samples were prepared, and the fictive temperature of the outer surface of each crucible was measured. The quartz glass crucibles of Examples C1 and C2 were measured using the rotating mold method described above. In the arc melting process, the arc electrode was kept in the mold even immediately after the end of arc heating, and the mold cooling was stopped immediately after the end of arc heating, and the inner and outer surfaces of the crucible were slowly cooled. The quartz glass crucible of Comparative Example C1 was manufactured using the conventional method of rapid cooling after the end of arc heating.

Next, a silicon single crystal was pulled using a crucible equivalent to this one, and the condition of the outer surface of the crucible after pulling was visually inspected to determine the number of dents present on the outer surface of the crucible. These dents are thought to be the result of external pressure or impact received before pulling, such as during transportation, which manifested as dents during the pulling process. The results are shown in Table 2.

As shown in Table 2, in Example C1, the fictive temperature difference was 40°C and the number of dents on the outer surface was 0. In Example C2, the fictive temperature difference was 70°C and the number of dents on the outer surface was 0.

On the other hand, in Comparative Example C1, the fictive temperature difference was 10°C, and the number of dents on the outer surface was 5. From these results, it was confirmed that the outer surfaces of the quartz glass crucibles of Examples C1 and C2, in which the fictive temperature of the crucible outer surface was 30°C or more lower than that of the crucible interior, had fewer dents and improved strength.

1 Quartz glass crucible 10a Side wall portion 10b Bottom portion 10c Corner portion 10i Inner surface 10o Outer surface 11 Transparent layer 12 Bubble layer 14 Carbon mold 14a Ventilation hole 14i Inner surface 15 Arc electrode 16 Deposition layer 16a Natural silica powder 16b Synthetic silica powder T Fictitious temperature of inner surface _Vi1 T Fictitious temperature inside _Vi2 Fictitious temperature inside _Vo1 outer surface T Fictitious temperature inside _Vo2 dd Measurement area inside crucible di Measurement area inside crucible do Measurement area outside crucible

Claims (8)

a cylindrical side wall portion, a bottom portion provided below the side wall portion, and a corner portion provided between the side wall portion and the bottom portion and having a larger curvature than the bottom portion,
A quartz glass crucible characterized in that the fictive temperature of the inner surface of the side wall portion is 30°C or more lower than the fictive temperature of the inner interior at a depth of at least 3 mm from the inner surface. The quartz glass crucible of claim 1, wherein the fictive temperature of the inner surface of the side wall portion is at least 50°C lower than the fictive temperature of the inner portion. The quartz glass crucible according to claim 1, wherein the OH group concentration on the inner surface of the side wall portion is 15 ppm to 70 ppm. A quartz glass crucible according to any one of claims 1 to 3, wherein the fictive temperature of the outer surface of the side wall portion is at least 30°C lower than the fictive temperature of the outer interior at a depth of at least 3 mm from the outer surface. A quartz glass crucible as described in claim 4, wherein the fictive temperature of the outer surface of the side wall portion is at least 50°C lower than the fictive temperature of the outer interior. A method for manufacturing a quartz glass crucible having a cylindrical side wall portion, a bottom portion provided below the side wall portion, and a corner portion provided between the side wall portion and the bottom portion and having a larger curvature than the bottom portion,
forming a deposition layer of raw silica powder on the inner surface of the carbon mold;
arc-melting the deposited layer of the raw silica powder from inside the carbon mold;
and a step of cooling the fused silica crucible formed in the carbon mold,
The step of cooling the quartz glass crucible includes:
Immediately after the arc melting of the raw silica powder is completed, a step of holding the arc electrode in a high temperature state in the carbon mold for a certain period of time to slowly cool the fused silica crucible in the carbon mold;
and after the slow cooling, the arc electrode is retracted above the carbon mold to increase the cooling rate of the quartz glass crucible. the step of arc-melting the deposited layer of raw silica powder includes a step of cooling the carbon mold;
7. The method for manufacturing a silica glass crucible according to claim 6, wherein the step of cooling the silica glass crucible includes a step of stopping cooling of the carbon mold for a certain period of time and a step of restarting cooling of the carbon mold. The method for manufacturing a quartz glass crucible according to claim 6, wherein the step of cooling the quartz glass crucible includes a step of humidifying the chamber in synchronization with the start of the slow cooling.