JP6301441B2 - Method for producing silica glass crucible for pulling silicon single crystal and method for producing silicon single crystal - Google Patents

Description

The present invention relates to the production how preparation and a silicon single crystal of the silica glass crucible for pulling up a silicon single crystal.

  The diameter of the silicon single crystal has shifted from the current mainstream of 300 mmφ to 400 to 450 mmφ. Accordingly, the diameter of the silica glass crucible is also shifted from about 600 mm to 700 mm or more. As the diameter of the silica glass crucible becomes larger, the distance from the heater arranged outside the silica glass crucible to the center of the silicon single crystal becomes longer than before. For example, when the diameter is shifted from about 600 mm to 700 mm, the distance from the heater to the center of the single crystal is 50 mm or more. In addition, the amount of silicon melt increases as the diameter of the quartz glass crucible increases.

  The increase in the distance from the heater to the center of the silicon single crystal and the amount of polysilicon to be melted leads to an increase in the temperature applied to the silica glass crucible and a longer pulling time. Brown cristobalite is generated on the inner surface of the silica glass crucible when it is in contact with a high-temperature silicon melt for a long time. As the single crystal pulling progresses, cristobalite grows on the inner surface of the silica glass crucible or in a direction perpendicular to the inner surface to form ring-like spots (Brown ring). The formed brown ring is easy to peel off. When the peeled brown ring falls and mixes in the silicon melt, it is carried to the silicon single crystal. As a result, the pulled silicon ingot is polycrystallized and the single crystallization rate is lowered.

  Bubbles contained in the inner layer of the silica glass crucible also cause a reduction in the single crystallization rate. As the erosion of the inner layer of the silica glass crucible proceeds, bubbles in the inner layer of the silica glass crucible enter the silicon melt. Since the bubbles in the silicon melt are contained in the silicon ingot, the single crystallization rate is lowered. In addition, the bubbles contained in the inner layer of the silica glass crucible significantly expand under high temperature conditions for a long time. The expanded bubbles deform the silica glass crucible or make the inner surface non-uniform. As a result, molten metal surface vibration occurs in the silicon melt, and the single crystallization rate is reduced.

  Furthermore, as the diameter of the silica glass crucible increases, stress concentrates at the interface between the synthetic layer formed from the synthetic silica powder and the natural layer formed from the natural silica powder, and breaks during transportation of the silica glass crucible. There was a fear.

JP 2002-080230 A JP-A-11-116388 Patent 2923720

  In the said patent document 1, the processing method which sinters synthetic quartz powder at the temperature more than decarburization temperature and less than powder sintering temperature under the vacuum of 100 Pa or less of vacuum is disclosed. However, even the synthetic quartz powder produced by such a method has a problem that it is difficult to suppress the occurrence of brown rings and open bubbles on the inner surface of a quartz glass crucible used under a high temperature condition for a long time. It was.

  Patent Document 2 discloses a method of removing the inner surface by heating to a temperature equal to or higher than the evaporation temperature of quartz glass in the latter half of arc melting. However, when producing a silica glass crucible by such a method, there has been a problem that the inner surface to be removed is wasted.

  In the said patent document 3, the quartz crucible which reduced the bubble content rate of the transparent glass layer is disclosed. However, it is unclear whether a single crystallization rate of 90% or more can be achieved even with a large-diameter crucible.

  The present invention has been made in view of such circumstances, and a silica glass crucible that suppresses the occurrence of brown rings and open bubbles on the inner surface of the silica glass crucible even when used under long-term high temperature conditions. Is used to provide a method for producing a silicon single crystal having a good single crystallization rate and a silicon single crystal having a good single crystallization rate.

According to one aspect of the present invention, a silica glass crucible for pulling a silicon single crystal comprising an inner surface layer formed from synthetic silica powder and an outer surface layer formed from natural silica powder, the inner surface layer comprising: the peak intensities D1 wavelength 492cm ^-1 measured by Raman intensity ratio D1 / D2 of the peak intensity D2 of wavelength 606 cm ^-1 is formed by melting the 1.8 to 2.0 of the synthetic silica powder A silica glass crucible is provided. Then, polycrystalline silicon is charged into the silica glass crucible, the polycrystalline silicon is melted, and a silicon single crystal is manufactured by immersing the end of the silicon seed crystal in the polycrystalline silicon melt and pulling it up while rotating. .

The inventors investigated the relationship between the degree of strain of the synthetic silica powder and the occurrence of brown rings and open bubbles. When analyzing the Raman spectra of the silica glass, a peak is detected in the 492cm ^-1 and 606 cm ^-1. Each peak corresponds to a planar four-membered ring (D1) and a planar three-membered ring (D2), which constitute a strain in silica glass. A silica glass crucible manufactured using synthetic silica powder having an intensity ratio D1 / D2 between the peak intensity D1 and the peak intensity D2 of 1.8 or more and 2.0 or less is that of a silicon single crystal under a high temperature condition for a long time. The present invention has been completed by finding that generation of brown rings and open bubbles is suppressed on the inner surface of the silica glass crucible during pulling.

Moreover, according to one aspect of the present invention, there is provided a method for producing a silica glass crucible for pulling a silicon single crystal comprising an inner surface layer formed from synthetic silica powder and an outer surface layer formed from natural silica powder. , the method of synthetic silica powder, a peak intensity D1 wavelength 492cm ^-1 measured by Raman intensity ratio D1 / D2 of the peak intensity D2 of wavelength 606 cm ^-1 is 1.8 to 2.0 A step of providing a rotating mold for producing a silica glass crucible, a step of supplying natural silica powder to the mold, and a step of supplying the treated synthetic silica powder onto the natural silica powder. And producing a silica glass crucible by generating an arc discharge inside the mold on which the natural silica powder and the treated synthetic silica powder are deposited. Method for producing a silica glass crucible for con single crystal pulling is provided.

  Also in this case, a silica glass crucible having the same effect can be obtained efficiently.

FIG. 1 is a cross-sectional view of a silica glass crucible and a drawing depicting a strain observation method. FIG. 2 is an electron micrograph of the appearance and cross section of the plasma-treated synthetic silica powder. FIG. 3 is an electron micrograph of the appearance and cross section of a synthetic silica powder not subjected to plasma treatment. FIG. 4 is a Raman spectrum of natural silica powder, plasma-treated synthetic silica powder and non-plasma-treated synthetic silica powder. FIG. 5 is a graph plotting the bubble content in each part of the silica glass crucible using the plasma-treated synthetic silica powder and the silica glass crucible using the non-plasma-treated synthetic silica powder. FIG. 6 is a laser confocal micrograph of the inner surface of a silica glass crucible using a plasma-treated synthetic silica powder. FIG. 7 is a laser confocal micrograph of the inner surface of a silica glass crucible using synthetic silica powder not subjected to plasma treatment. FIG. 8 is a polarization photograph of a slice piece obtained by slicing a silica glass crucible using synthetic silica powder not subjected to plasma treatment in the thickness direction. FIG. 9 is a polarized photograph of a slice piece obtained by slicing a silica glass crucible using synthetic silica powder subjected to plasma treatment in the thickness direction.

The silica glass crucible of this embodiment is a silica glass crucible for pulling up a silicon single crystal comprising an inner surface layer formed from synthetic silica powder and an outer surface layer formed from natural silica powder, and the inner surface layer is the peak intensities D1 wavelength 492cm ^-1 measured by Raman intensity ratio D1 / D2 of the peak intensity D2 of wavelength 606 cm ^-1 is formed by melting the 1.8 to 2.0 of the synthetic silica powder This is a silica glass crucible. Hereinafter, each component will be described in detail.

1. Synthetic Silica Powder A silica glass crucible is composed of two layers: an inner surface layer (hereinafter referred to as “synthetic layer”) formed from synthetic silica powder and an outer surface layer (hereinafter referred to as “natural layer”) formed from natural silica powder. The synthetic layer also includes a seal layer formed on the outermost surface of the silica glass crucible during arc melting. The natural silica powder is a silica powder produced by pulverizing a natural mineral mainly composed of α-quartz. Synthetic silica powder is produced by chemical synthesis methods such as gas phase oxidation (dry synthesis method) of silicon tetrachloride (SiCl ₄ ) and hydrolysis (sol-gel method) of silicon alkoxide (Si (OR) ₄ )). can do.

  A silica glass crucible manufactured using a synthetic silica powder having an intensity ratio D1 / D2 between the peak intensity D1 and the peak intensity D2 of 1.8 or more and 2.0 or less is a silicon single crystal under a long-time high temperature condition. In the pulling up, the occurrence of brown rings and open bubbles on the inner surface of the silica glass crucible is suppressed. The peak intensity ratio D1 / D2 is, for example, 1.80, 1.84, 1.87, 1.89, 1.93, 1.97, 2.00. It may be within the range.

  The synthetic silica powder in this embodiment may be subjected to plasma treatment using a plasma reactor.

It is known that the peak intensity of D1 and D2 of synthetic silica changes depending on the fictive temperature. Therefore, the desired peak intensity ratio D1 / D2 can be obtained by heating the synthetic silica powder and then rapidly cooling it to reflect the heat history. The plasma treatment by the plasma reactor can easily perform rapid heating and rapid cooling. The plasma treatment is not particularly limited, but thermal plasma treatment capable of continuously performing the heating-cooling process is preferable. Further, in order to obtain a desired peak intensity ratio D1 / D2, it is necessary to rapidly cool the synthetic silica powder after heating, preferably a cooling rate of 10 ⁵ K / s or more, more preferably 10 ⁶ K / s or more. Quench quickly at the cooling rate. The cooling method is not particularly limited, but is preferably cooling by air cooling, more preferably cooling by water cooling. In particular, in the case of thermal plasma processing, it is possible to efficiently cool by providing a water-cooled cooling device at the plasma torch outlet. Thereby, the synthetic silica powder in which the thermal history is reflected can be obtained stably.

  Moreover, it is preferable that the synthetic silica powder in this embodiment has a circularity of 0.8 or more and 1.0 or less.

The synthetic silica powder is preferably spherical. Since the spherical synthetic silica powder has a small gap between the particles, the gap is easily closed during melting, and the residual gas component in the silica glass crucible can be prevented. The circularity of the synthetic silica powder is preferably 0.8 or more. When the circularity is less than 0.8, the gap between the particles is large, so that the gap is not closed during melting, the gas component remains in the silica glass crucible, and the bubble content increases. The expression of circularity is circularity = 4πS / L ² (S: area in the particle projection diagram of the recorded image taken; L: perimeter of the particle projection diagram). Synthetic silica powder is dispersed in a liquid and the liquid is allowed to flow into a planar extension flow cell. 200 powder particles that move into the planar elongated flow cell are recorded as an image on the objective lens, and the circularity is calculated from the recorded image. The measurement is performed twice, and the average value is taken as the circularity of the powder. When the particle is a perfect circle, the circularity is 1.

In this embodiment, the synthetic silica powder has an average particle size of 80 μm or more and 150 μm or less, a tap bulk density of 1.38 g / cm ³ or more and 1.44 g / cm ³ or less, and a specific surface area of 0.026 m ² / g or more and 0.0. It is preferably 040 m ² / g or less.

  By the way, the term “particle size” generally refers to the definition of the term of JIS Z 8901 “Test Powder and Test Particles”. In this specification, the particle size distribution is measured by using a laser beam as a light source. The laser diffraction / scattering measurement method described above is used. The “average particle size” means a particle size (D50) at an integrated value of 50% in the obtained particle size distribution, and in this specification means a volume average particle size.

  The average particle diameter is, for example, 80, 85, 90, 95, 100, 110, 120, 130, 140, 150 μm, and may be in the range of any two numerical values exemplified here. When the average particle size is 150 μm or less, the size of one void between the particles is small, so that the bubble size in the silica glass crucible formed due to the entrainment of atmospheric gas can be reduced. As a result, bubbles can be contracted and eliminated during arc melting. However, when the average particle size is more than 150 μm, the size of one void between the particles is large, so that the bubble size in the silica glass crucible formed due to the entrainment of atmospheric gas is large, and the bubble is not melted during arc melting. Shrinkage progresses, but bubbles do not disappear.

When the tap bulk density is less than 1.38 g / cm ³ or the specific surface area is more than 0.040 m ² / g, the bubble content increases. This is probably because there are many scratches and cracks inside and outside the particles, or the particles contain gas.

Accordingly, the tap bulk density is, for example, 1.38, 1.39, 1.40, 1.41, 1.42, 1.43, 1.44 g / cm ³ , and any two of the examples illustrated here It may be within a numerical range. The tap bulk density refers to a density when the sample is dispersed and placed in the container, and then the container is impacted by the tap and the volume of the sample is not changed. The measurement container containing the sample is placed on the tap device with the auxiliary cylinder attached, and tapping is performed 600 times. After grinding the sample, the mass is measured. The sample was replenished again, the measurement container with the auxiliary cylinder attached was placed in the tap device, and the tap was performed 100 times. After grinding the sample, the mass is measured, and the operation is repeated until the mass difference from the previous mass is within 0.3%. The mass of the sample is divided by the volume of the measurement container to obtain the tap bulk density. The measurement is performed 3 times and the average value is adopted.

The specific surface area is, for example, 0.026, 0.028, 0.030, 0.032, 0.034, 0.036, 0.038, 0.040 m ² / g, and any of the two exemplified here It may be within a range of two numerical values. The specific surface area can be determined by a nitrogen adsorption method. In the nitrogen adsorption method, (1) nitrogen gas is adsorbed on the synthetic silica powder while gradually increasing the pressure from a high vacuum. (2) An adsorption isotherm is created by plotting relative pressure on the X-axis and nitrogen adsorption on the Y-axis. (3) This is a method for obtaining the specific surface area by applying such adsorption isotherm data to various adsorption isotherms. Examples of the adsorption isotherm include the Henry adsorption isotherm, the Langmuir adsorption isotherm, and the BET adsorption isotherm. In this specification, a BET adsorption isotherm using a multi-layer molecule as a model formula is used.

  It is preferable that all of the characteristic values of the average particle diameter, tap bulk density, and specific surface area satisfy the above conditions.

2. Silica glass crucible Silica glass crucible is a synthetic silica powder produced by supplying natural silica powder to a rotating mold for producing silica glass crucible, supplying synthetic silica powder onto natural silica powder, and melting the silica powder by arc discharge. A silica glass crucible comprising two layers of an inner surface layer (synthetic layer) formed from powder and an outer surface layer (natural layer) formed from natural silica powder is produced.

  The silica glass crucible manufactured using the synthetic silica powder described above substantially does not contain bubbles near the inner surface of the crucible. Here, “substantially free of bubbles” means that the bubble content and bubble diameter are such that the single crystallization rate does not decrease due to bubbles.

  If even a small amount of bubbles exist in the vicinity of the inner surface of the silica glass crucible, bubble expansion occurs in the synthetic layer when the silicon single crystal is pulled up. The generated bubbles enter the silicon melt together with the dissolution of the inner surface side of the synthetic layer, and the bubbles are taken into the pulled silicon single crystal. The incorporated bubbles cause dislocations (crystal defects) due to crystal transition and lower the single crystallization rate. For example, optical detection means can be used to detect bubbles present in the vicinity of the inner surface of the silica glass crucible.

  The optical detection means includes a light receiving device that receives transmitted light or reflected light of light irradiated on the silica glass crucible. The light emitting means for irradiating light may be built-in or may use an external light emitting means. As the optical detection means, one that can be rotated along the inner surface of the silica glass crucible is used. As the irradiation light, in addition to visible light, ultraviolet light and infrared light, X-rays or laser light can be used, and any light can be applied as long as it can reflect and detect bubbles. The light receiving device is selected according to the type of irradiation light. For example, a digital camera including an optical lens and an image sensor can be used. In order to detect bubbles existing at a certain depth from the surface, the focal point of the objective lens may be scanned in the depth direction from the surface.

  The measurement result by the optical detection means is taken into the image processing apparatus, and the bubble content P (%) is calculated. An image of the inner surface of the crucible is picked up using an optical camera, the inner surface of the crucible is divided into a predetermined volume to obtain a reference volume W1, and an occupied volume W2 of bubbles with respect to the reference volume W1 is obtained, and P (%) = (W2 / W1) × 100. The average bubble content from the inner surface of the crucible to the outer surface with a thickness of 0.3 mm is preferably 0.05 vol% or less. At this time, bubbles having a bubble diameter of 10 μm or more are measured. When it is higher than 0.05 vol%, the single crystallization rate is remarkably reduced. Moreover, it is preferable that the maximum bubble content is 0.10 vol% or less. When it is larger than 0.10 vol%, the single crystallization rate is remarkably reduced. Furthermore, the average cell diameter is preferably 50 μm or less. If it exceeds 50 μm, the silica glass crucible is deformed due to the expansion of bubbles. Moreover, when the maximum bubble diameter is larger than 100 μm, since the reduction of the single crystallization rate is remarkable, it is preferably 100 μm or less.

  In addition, if there are irregularities on the inner surface of the silica glass crucible when pulling up the silicon single crystal, non-uniform nucleation is likely to occur. When the nuclei are peeled off due to melting damage and float in the silicon solution, if they adhere to the growth interface of the silicon single crystal being pulled, quality defects such as polycrystallization or dislocation are caused. The arithmetic average roughness (Ra) of the inner surface of the silica glass crucible is preferably 0.02 μm or less. When it is larger than 0.02 μm, the probability that heterogeneous nucleation occurs is increased, and the single crystallization rate is remarkably lowered. Ra is extracted from the roughness curve by the reference length in the direction of the average line, the X-axis is taken in the direction of the average line of the extracted portion, the Y-axis is taken in the direction of the vertical magnification, and the roughness curve is taken from the center line. The integrated value is expressed in micrometers.

  In this embodiment, a silica glass crucible provided with an inclined birefringent layer in which the birefringence at the boundary between the inner surface layer and the outer surface layer continuously changes is provided.

  Since the natural layer and the synthetic layer are formed by melting silica powder having different characteristics, the thermal expansion characteristics of each layer are different. The amount of shrinkage of each layer during curing after arc melting is also different. As a result, different internal stresses (strains) exist in each layer of the silica glass crucible.

  The strain of each layer in the fusion product of layers having different thermal expansion characteristics can be observed using two polarizing plates. As shown in FIG. 1, a crucible is sliced in the thickness direction, and a sliced crucible piece (fused body) is placed between two polarizing plates combined in a crossed Nicol state, and observed through white light. At this time, the sliced crucible piece is about 2 mm thick by polishing. When there is no distortion in the fused body, the fused body does not give an optical path difference with respect to the white polarized light, so the white polarized light that has passed through the fused body must pass through an orthogonal polarizing plate (analyzer). I can't. When there is distortion in the fused material, the fused material gives an optical path difference with respect to the white polarized light, so that the polarization plane of the white polarized light rotates and the components that can pass through the orthogonal polarizing plate (analyzer) are observed. Is done. In addition, the birefringence is different for each wavelength. For this reason, when white polarized light is passed through a fused body having strain, an optical path difference corresponding to the strain is generated for each wavelength, and thus the amount of light passing through the polarizing plate is different for each wavelength. As a result, the color of the fused body observed through the polarizing plate (analyzer) is observed. It is also possible to evaluate the distortion of the fused body from this color. For example, the distortion of the fused body can be evaluated by using an interference color diagram or a polarization color diagram representing the relationship between chromaticity and birefringence.

  In the conventional silica glass crucible, the change in strain is clear from the natural layer to the synthetic layer, and the continuity of strain is not substantially observed. Therefore, the interface between the natural layer and the synthetic layer is clear, and when the external stress applied to the silica glass crucible is concentrated on the interface, there is a problem that the silica glass crucible is lost. In the silica glass crucible according to the present invention, the strain continuously changes (inclines) from the natural layer to the synthetic layer (seal layer). Therefore, since there is no clear interface between the natural layer and the synthetic layer, external stress applied to the silica glass crucible is dispersed throughout, and the silica glass crucible can be prevented from being lost. In particular, the large-diameter crucible has a higher degree of deformation and a higher risk of breakage than the small-diameter crucible with respect to external stress applied to the crucible. For this reason, the large-diameter crucible including the inclined birefringent layer can reduce the risk of breakage.

  Therefore, the silica glass crucible of this embodiment is effective when the diameter is 32 inches (about 800 mm) or more. Furthermore, a remarkable effect is exhibited also in the case of a large crucible having a diameter of 40 inches (about 1000 mm) or more.

3. Silica glass crucible after pulling up the silicon single crystal In the silicon single crystal, polycrystalline silicon (polysilicon) is charged into the silica glass crucible and heated by a heater to melt the polycrystalline silicon. It is manufactured by pulling up the seed crystal while rotating the end portion of the silicon seed crystal immersed in a silicon melt. The shape of the silicon single crystal is a cylindrical silicon seed crystal from the top, a conical silicon single crystal below (top portion), a cylindrical silicon single crystal having the same diameter as the bottom of the upper cone (straight barrel portion), It consists of a conical silicon single crystal (tail part) whose apex is downward.

  The silica glass crucible in the present embodiment can remarkably reduce the maximum bubble content, open cell number density, brown ring number density, and brown ring diameter of the silica glass crucible after pulling up the silicon single crystal. Can be improved. The single crystallization rate is defined as the weight ratio of the single crystal to the silicon raw material. However, not all the silicon melt in the crucible is used, and only the straight body portion excluding the top portion and the tail portion of the silicon single crystal ingot is subject to calculation of the single crystallization rate. Therefore, even if a sufficient silicon single crystal is pulled, the single crystallization rate is 100% or less, and it is good if it is 80% or more.

  The maximum bubble content of the silica glass crucible after pulling up the silicon single crystal is preferably 0.1 (Vol%) or less. Under these conditions, mixing of bubbles into the silicon single crystal can be reduced, and a good single crystallization rate can be realized.

The average open cell number density is preferably 7 (number / cm ² ) or less. Under these conditions, mixing of bubbles into the silicon single crystal can be reduced, and a good single crystallization rate can be realized. An open bubble is a recess derived from bubbles that appears on the inner surface of the silica glass crucible due to melting of the inner surface of the silica glass crucible when the silicon single crystal is pulled up. The open cell number density can be calculated by counting the number of open cells per unit area formed on the inner surface of the silica glass crucible after pulling up the silicon single crystal by microscopic observation. The average open cell number density can be calculated from the average value of the open cell number density in the straight body part, bottom part, and corner part from the straight body part to the bottom part of the silica glass crucible.

The average brown ring number density is preferably 7 (number / cm ² ) or less. Under these conditions, mixing of bubbles into the silicon single crystal can be reduced, and a good single crystallization rate can be realized. The brown ring number density can be calculated by counting the number of brown rings per unit area formed on the inner surface of the silica glass crucible after pulling up the silicon single crystal by microscopic observation. The average brown ring number density can be calculated from the average value of the brown ring number density of the straight barrel portion, bottom portion, and corner portion from the straight barrel portion to the bottom portion of the silica glass crucible.

  The average brown ring diameter is preferably 4 (mm) or less. Under these conditions, mixing of bubbles into the silicon single crystal can be reduced, and a good single crystallization rate can be realized. The brown ring diameter is the diameter of the brown ring formed on the inner surface of the silica glass crucible after the silicon single crystal is pulled. The average brown ring diameter can be calculated from an average value obtained by measuring the diameters of 100 brown rings from the straight body part, bottom part, and corner part from the straight body part to the bottom part of the silica glass crucible.

  In Examples and Comparative Examples, silica glass crucibles were produced based on the rotational mold method. The diameter of the mold was 32 inches (81.3 cm), the average thickness of the silica powder layer deposited on the inner surface of the mold was 15 mm, and arc discharge was performed with three electrodes of three-phase alternating current. In the arc melting process, the energization time was 90 minutes, the output was 2500 kVA, and the silica powder layer was evacuated for 10 minutes from the start of energization.

Examples 1-6
The synthetic silica powder obtained by hydrolysis of alkoxysilane was fired under low vacuum conditions to obtain a fired synthetic silica powder. The obtained synthetic silica powder was put into a high frequency induction thermal plasma generator. After the thermal plasma treatment, the synthetic silica powder was recovered, and the synthetic silica powder was washed by ultrasonic cleaning using ultrapure water. The washing was performed until there were no fine particles adhering to the synthetic silica powder. The synthetic silica powder after washing was classified. The Raman spectrum, average particle diameter, specific surface area, tap bulk density and circularity of the synthetic silica powder were measured. Thereafter, silica glass crucibles were produced using the respective synthetic silica powders, the average cell diameter and the maximum cell content were measured, and the birefringence layer was observed. For Example 1, the arithmetic surface roughness was also calculated.

Comparative Examples 1-12
A synthetic silica powder is produced based on the methods of Examples 1 to 6 excluding the thermal plasma treatment. The Raman spectrum, average particle diameter, specific surface area, tap bulk density and circularity of the synthetic silica powder were measured. Thereafter, silica glass crucibles were produced using the respective synthetic silica powders, the average cell diameter and the maximum cell content were measured, and the birefringent layer was observed. For Comparative Example 1, the arithmetic surface roughness was also calculated.

  2 is an electron micrograph of the appearance and cross section of the synthetic silica powder according to Example 1, and FIG. 3 is an electron micrograph of the appearance and cross section of the synthetic silica powder according to Comparative Example 1. The synthetic silica powder according to Example 1 was spherical, whereas the synthetic silica powder according to Comparative Example 1 was found to be amorphous.

Raman Spectrum FIG. 4 is a Raman spectrum of the synthetic silica powder of Example 1 and the synthetic silica powder of Comparative Example 1. For the Raman spectrum, a dispersive microscopic Raman apparatus SENTERRA manufactured by BRUKER was used. Measurement conditions are laser wavelength: 532 nm (5 mw), exposure time: 20 seconds, and integration count: once. When the peak intensity ratio D1 / D2 was determined with the peak intensity of D2 being 1, the synthetic silica powder of Example 1 was 1.89, and the synthetic silica powder of Comparative Example 1 was 1.74. Therefore, it was found that the synthetic silica powder of Example 1 had a higher peak intensity ratio D1 / D2 than the synthetic silica powder of Comparative Example 1.

Bubble Content In the silica glass crucibles of Example 1 and Comparative Example 1, the bubble content from the bottom of the crucible through the corner to the straight body (wall) was measured. FIG. 5 is a graph plotting the X-axis as each part of the silica glass crucible and the Y-axis as the bubble content (Vol%). In Comparative Example 1, the bubble content increased from the corner portion to the wall portion, whereas Example 1 showed a bubble content rate of 0.01 Vol% or less only in the wall portion. Therefore, it was found that the silica glass crucible of Example 1 was a glass crucible substantially free of bubbles.

Arithmetic surface roughness (Ra)
The inner surfaces of the silica glass crucibles of Example 1 and Comparative Example 1 were cut out, and the roughness of the inner surface was measured in a non-contact manner using a laser confocal microscope (Lasertec H1200, manufactured by Lasertec Corporation). 6 is a photograph of the surface of the silica glass crucible of Example 1, and FIG. 7 is a photograph of Comparative Example 1. When Ra was calculated, Example 1 was 0.012 μm, and Comparative Example 1 was 0.025 μm. Therefore, the silica glass crucible of Example 1 showed Ra superior to that of Comparative Example 1.

  The results of Examples 1 to 6 and Comparative Examples 1 to 12 are shown in Table 1.

Strain observation At a position 600 mm from the intersection of the central axis of the silica glass crucible and its bottom to the rim along the inner surface, the silica glass crucible was cut in the thickness direction and the sliced pieces were polished to a thickness of about 2 mm. . The polished cross-sectional sample (slice piece) was placed between two polarizing plates arranged so as to be in a crossed Nicol state, and observed by applying white light.

  FIG. 8 is a polarization photograph of a cross-sectional sample in the silica glass crucible of Comparative Example 1. In the silica glass crucible of Comparative Example 1, the boundary between the synthetic layer and the natural layer is clear. FIG. 9 is a polarized photograph of a cross-sectional sample in the silica glass crucible of Example 1. In the silica glass crucible of Example 1, the boundary between the synthetic layer and the natural layer is not clear. That is, the strain continuously changes from the synthetic layer toward the natural layer, and the synthetic layer and the natural layer are fused without a clear boundary. Therefore, in Example 1, a silica glass crucible having a birefringent layer and resistant to external stress was obtained. Similarly, birefringent layers were also observed from the silica glass crucibles of Examples 2-6.

Pulling effect Using the silica glass crucibles produced in Examples 1 to 6 and Comparative Examples 1 to 12, the silicon single crystal was pulled, the maximum bubble content (Vol%), the average open cell number density (number / cm ^2). ), Average brown ring number density (number / cm ² ), average brown ring diameter (mm), and single crystallization rate (%). These results are listed in Table 2.

When Table 1 and Table 2 are seen, Examples 1-6 whose intensity ratio D1 / D2 is 1.8 or more and 2.0 or less had a favorable single crystallization rate. In particular, the average particle size is 87 to 150 μm, the specific surface area is 0.026 to 0.034 (m ² / g), the tap bulk density is 1.39 to 1.44 (g / cm ³ ), and the circularity is 0.85. The silica glass crucible manufactured using the synthetic silica powders of Examples 1 to 3 having a ratio of 0.91 exceeded 0.9%. In addition, the silica glass crucibles of Examples 1 to 6 after pulling up the silicon single crystal have lower average open cell number density, average Brown ring number density and average Brown ring diameter, compared with Comparative Examples 1 to 12, and in particular, Examples 1-3, the average open cell number density is 4.0 to 4.5 (number / cm ² ), the average brown ring number density is 2.6 to 3.1 (number / cm ² ), and the average brown ring diameter is It was 2.3 to 2.5 (mm) and was extremely low. Therefore, the silica glass crucibles of Examples 1 to 6 have low generation of bubbles and brown rings during pulling of the silicon single crystal and good single crystallization rate. In particular, Examples 1 to 3 have a single crystallization rate. Remarkably good.

Claims (8)

A method for producing a silica glass crucible for pulling a silicon single crystal,
The method
Processing synthetic silica powder, a peak intensity D1 wavelength 492cm ^-1 measured by Raman, so that the intensity ratio D1 / D2 of the peak intensity D2 of wavelength 606 cm ^-1 becomes 1.8 to 2.0 And forming a synthetic silica powder after treatment,
A method for producing a silica glass crucible for pulling a silicon single crystal, comprising a step of producing a silica glass crucible comprising an inner surface layer formed from the treated synthetic silica powder and an outer surface layer formed from natural silica powder. . The step of forming the synthetic silica powder after the treatment,
The silica glass crucible for pulling up a silicon single crystal according to claim 1, wherein the synthetic silica powder is heated and then cooled so that the strength ratio D1 / D2 is 1.8 to 2.0. Manufacturing method.   The method for producing a silica glass crucible for pulling a silicon single crystal according to claim 2, wherein the cooling after heating is a thermal plasma treatment in which a heating-cooling process can be continuously performed. The method for producing a silica glass crucible for pulling up a silicon single crystal according to claim 2, wherein the cooling method for cooling after heating is rapid cooling at a cooling rate of 10 ⁵ K / s or more. The method for producing a silica glass crucible for pulling up a silicon single crystal according to claim 2, wherein the cooling method for cooling after heating is rapid cooling at a cooling rate of 10 ⁶ K / s or more.   The method for producing a silica glass crucible for pulling up a silicon single crystal according to claim 2, wherein the cooling method for cooling after heating is air cooling.   The method for producing a silica glass crucible for pulling up a silicon single crystal according to claim 2, wherein the cooling method for cooling after heating is water cooling. An inner surface layer formed from synthetic silica powder;
An outer surface layer formed from natural silica powder, and a method for producing a silicon single crystal using a silica glass crucible,
It said inner surface layer has a peak intensity D1 wavelength 492cm ^-1 measured by Raman intensity ratio D1 / D2 of the peak intensity D2 of wavelength 606 cm ^-1 is treated to be 1.8 to 2.0 Formed by synthetic silica powder after treatment
Polycrystalline silicon is charged into the silica glass crucible, the polycrystalline silicon is melted, and the end of the silicon seed crystal is immersed in the melt of the polycrystalline silicon and pulled up while rotating.
A method for producing a silicon single crystal.