Precise seeding method for bubble-solution sapphire single crystal based on lotus leaf vein cold heart positioning

Abstract

The present application relates to the technical field of sapphire single crystal growth, in particular to a precise seeding method for sapphire single crystal by the Kyropoulos method based on lotus vein cold heart positioning, which provides a crucible with lotus vein radiating heat conduction channels on the bottom surface, the channels diverge from the center convergence zone to the edge, forming a path for strengthening downward heat conduction in the center area of the crucible bottom, and a cold heart is formed at the center of the melt surface; under the condition that the crucible is not loaded and filled with inert gas, the temperature distribution of the melt bearing surface is obtained by thermal imaging, the lowest temperature area is calibrated as the cold heart position coordinates and the mapping relationship with the fine adjustment platform coordinates of the seed crystal rod is established; according to the mapping relationship, the seed crystal is positioned above the cold heart through the X-Y fine adjustment platform; the seed crystal is lowered to contact the melt, and after contact, it remains stationary until single crystal nucleation and pulling, so that the cold heart position is stable and can be calibrated, the seed crystal and the cold heart are precisely aligned, and the seeding success rate and single crystal rate are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of sapphire single crystal growth technology, specifically to a precise crystal pulling method for sapphire single crystals based on the cold-core positioning of lotus leaf veins. Background Technology

[0002] Sapphire's main component is high-purity α-Al₂O₃, with a Mohs hardness second only to diamond. It also possesses excellent optical transmittance, thermal conductivity, chemical stability, and mechanical properties, making it widely used in light-emitting diode substrates, optical windows, wear-resistant devices, and microelectronics. Among the industrial methods for producing large-size sapphire single crystals, the Czochralski method is one of the most widely used methods due to its small temperature gradient, low stress caused by the crystal being mostly immersed in the melt during growth, and its suitability for growing large-size, low-defect single crystals. Approximately 70% of sapphire single crystals worldwide are produced using the Czochralski method.

[0003] The basic process of the bubble growth method is as follows: High-purity alumina raw material is placed in a high-temperature resistant crucible and heated to above 2050°C under vacuum or inert atmosphere to completely melt the raw material and form a melt. Then, a seed crystal is lowered to the surface of the melt and made slight contact. When the temperature of the interface between the seed crystal and the melt is lower than the solidification point, the surface of the seed crystal slightly melts and recrystallizes to form the initial growth interface. Then, by gradually lowering the melt temperature and slowly pulling the crystal, the growth stages such as necking, shoulder formation, equal diameter formation, and tailing are completed. Throughout the growth process, except for the crystal introduction and shoulder formation stages, the growth morphology of the crystal is basically determined by the temperature field. The temperature field is formed by the combined action of the heating element, crucible, and reflector screen in the growth furnace. It usually exhibits a distribution of low temperature at the center, high temperature around the edges, low temperature at the top, and high temperature at the bottom. The part with the lowest temperature in the central area of ​​the melt surface is commonly referred to as the cold core.

[0004] Seed crystal growth is the core process in the Czochralski method for growing sapphire single crystals. The quality of seed crystal growth directly affects the overall crystal quality, especially for growing large-sized, optically high-quality sapphire single crystals. Ideally, the seed crystal should be positioned precisely at the cold core, where the melt surface temperature is lowest, allowing for moderate melting of the seed crystal's end face and recrystallization starting from a single nucleus, thus avoiding polycrystalline nucleation. However, existing Czochralski seed crystal growth techniques have significant shortcomings in both cold core formation and seed crystal alignment.

[0005] Firstly, in the traditional Czochralski method, the temperature field on the melt surface is passively formed by the superposition of multiple components such as the heating element, crucible, and reflector. The position and shape of the cold core not only change with the adjustment of the heating power but also drift with the growth time, resulting in significant uncertainty in its position. It is difficult for the seed crystal to align with a target with an uncertain position. If the contact position of the seed crystal deviates from the cold core during crystal pulling, nucleation can easily occur simultaneously in multiple locations in the colder region, leading to polycrystalline formation or crystal pulling failure. Chinese patent CN102978694A discloses an improved Czochralski method that reduces mechanical disturbance by stopping the rotation and movement of the seed crystal rod after single crystal formation. However, its cold core is still formed by a passive temperature field and its position is not fixed. The yield of crystal growth using this method is typically only about 65%, and the crystal neck defect is relatively large.

[0006] Secondly, while existing technologies have recognized the influence of the cooling core on the thermal field structure, their control targets differ from those of this invention. Chinese Patent CN105019023B discloses a seed crystal growth method for sapphire crystals using the Czochralski method. This method utilizes a non-rotationally symmetric structure of the seed crystal's radiative heat dissipation opening, allowing the seed crystal to preferentially grow a crystal disk to one side by waiting for heat dissipation from the opening. However, the heat dissipation structure in this method is located on one side of the seed crystal, and its control over heat dissipation at the melt center mainly relies on the top heat-shielding radiation structure. The position of the cooling core is still indirectly determined by the longitudinal heat dissipation gradient of the entire thermal field component, and its position is not fixed, failing to provide a definite spatial alignment coordinate for the seed crystal. Furthermore, the scheme of controlling the cooling core by adjusting the heater power distribution suffers from slow response and low control precision due to the large thermal inertia of the melt, making it difficult to stably fix the cooling core in the predetermined position.

[0007] Third, the positioning of the cold core in existing crystal pulling operations largely relies on the operator's visual observation and experience in judging the halo on the melt surface. This is highly subjective and inconsistent, with significant differences in crystal pulling results between different operators, and even between different batches by the same operator. The crystal pulling process has always been a trade secret of crystal manufacturers and is difficult to standardize. Chinese patents with publication numbers CN102877120A and CN104109904A disclose automatic crystal pulling technology that uses a weighing system to monitor the weight change of the seed crystal to determine the moment when the seed crystal contacts the melt. However, this type of technology only solves the problem of when the seed crystal contacts the melt, but does not solve the problem of the spatial positioning of the seed crystal in the melt. The seed crystal may still contact the melt at a position deviating from the cold core, thus failing to fundamentally guarantee single crystal nucleation.

[0008] Furthermore, the uncertainty of the cold core position not only affects whether a single crystal can nucleate during the seeding stage, but also has a chain reaction on subsequent shoulder formation and constant-diameter growth. If the seed crystal fails to align with the cold core during the seeding stage and forms multiple crystal nuclei in the colder region, the grains grown from different nuclei will compete and squeeze each other during the shoulder formation process, easily forming defects such as grain boundaries, small-angle grain boundaries, and dislocations inside the crystal, which can lead to the scrapping of the entire batch of crystals in severe cases. Even if growth is barely completed, the single crystal ratio and optical quality of the resulting crystal will be difficult to meet the requirements of sapphire for LED substrates. Due to the inherent limitations of the Czochralski method in its necking ability, the temperature gradient at the crucible mouth is too small, making it impossible to effectively reduce the inheritance of seed crystal defects to the crystal through the necking process like the Czochralski method. This further highlights the importance of accurately aligning the cold core and ensuring the initiation of a single crystal nucleus during the seeding stage.

[0009] It should be noted that although there are improved methods in this field, such as the cold-core shoulder-forming micro-pulling method, which are developed based on the pulling method and the bubble growth method, the cold core in these methods is still a low-temperature zone located near the center of the melt surface, passively formed by the overall temperature field of the growth furnace. Its position and shape depend on the combined effect of components such as the heating element, crucible, and reflector. The position of the cold core is not actively and geometrically fixed by the structure of the crucible itself. Therefore, the problem of the cold core position drifting with power and time and relying on experience judgment still exists.

[0010] In summary, the core problems faced by existing Czochralski crystal-leading techniques are: the cold core position on the melt surface is passively formed and uncertain with power and time drift; there is a lack of reliable spatial alignment methods between the seed crystal and the cold core; and the cold core positioning relies on human experience. How to form a stable, predictable, and calibrable cold core on the melt surface, and on this basis achieve precise spatial alignment between the seed crystal and the cold core, thereby improving the success rate of crystal-leading and the single crystal yield, and eliminating reliance on human experience, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0011] To address the shortcomings of the existing technologies, the present invention aims to provide a precise crystal pulling method for sapphire single crystals using the Czochralski method, based on the cold-core positioning of lotus leaf veins. Inspired by the radial distribution of lotus leaf veins, this invention actively modifies the heat conduction path at the bottom of the crucible by machining radial heat-conducting channels resembling lotus leaf veins on the crucible bottom plate. This forms a stable and calibrable cold core at a fixed point on the center of the melt surface. Precise alignment between the seed crystal and the cold core is achieved through infrared calibration and coordinate positioning, thereby improving the success rate of crystal pulling and the single crystal yield while eliminating reliance on operator experience.

[0012] To achieve the above objectives, the present invention adopts the following technical solution. A method for crystallizing sapphire single crystals using the Czochralski method based on cold core positioning includes the following steps: S1, providing a crucible, wherein the upper surface of the bottom plate of the crucible is provided with radial heat-conducting channels resembling lotus leaf veins, the radial heat-conducting channels being distributed from the convergence area in the center of the crucible bottom plate to the edge of the crucible, so that the central area of ​​the crucible bottom plate forms a path for enhanced downward heat conduction, forming a cold core in the central area of ​​the melt surface; S2, under the condition that the crucible is not loaded with material and is filled with inert gas, the temperature distribution of the molten bearing surface of the crucible is obtained by thermal imaging, the lowest temperature area is marked as the position coordinates of the cold core, and a mapping relationship is established between the position coordinates and the coordinates of the seed crystal rod fine-tuning platform; S3, according to the mapping relationship, the seed crystal is positioned directly above the cold core by the XY fine-tuning platform of the seed crystal rod; S4, the seed crystal is lowered to contact the melt, and after contact, it is kept stationary until single crystal nucleation, and then the seed crystal is pulled up.

[0013] Furthermore, the radial heat-conducting channels include multiple main vein channels radiating from the converging region towards the edge of the crucible. Each main vein channel gathers heat from the edge area of ​​the crucible bottom plate towards the converging region, and then conducts it downwards from the converging region towards the bottom of the crucible bottom plate, thereby forming a cold core with a lower temperature than its surroundings in the central area of ​​the melt surface. Furthermore, the converging region is constructed such that the heat-conducting cross-section at the center of the crucible bottom plate is larger than the sum of the heat-conducting cross-sections of the main vein channels, thus reducing the sensitivity of the cold core's location to fluctuations in heating power.

[0014] Further, in step S2, an infrared thermal imager is used to acquire the temperature field image of the crucible melt bearing surface, and the centroid of the lowest temperature region is extracted to obtain the position coordinates of the cold core; in step S3, the seed crystal is positioned directly above the cold core once by the XY fine-tuning platform according to the mapping relationship, and the position coordinates are reused in multiple single crystal growths in the same crucible; in step S4, after the seed crystal contacts the melt, an image of the contact interface between the seed crystal and the melt is acquired. When the image of the contact interface shows a closed and symmetrical halo, it is determined that the single crystal has nucleated, and the seed crystal is pulled up after the single crystal nucleation is determined. Furthermore, the radial heat-conducting channels have 8-12 main veins, a channel width of 2-3 mm, a channel depth of 1-2 mm, and a convergence zone diameter of 15-20 mm; the cold core has a diameter of 10-15 mm, and its temperature is 3-8°C lower than the surrounding melt surface; the seed crystal descends at a speed of 0.5 mm / min to contact the melt, and remains stationary for 10-15 minutes after contact; the inert gas is argon; the XY fine-tuning platform has a stroke of ±5 mm and a positioning accuracy of ±0.05 mm.

[0015] Compared with existing technologies, this invention has the following advantages. First, by processing radial heat-conducting channels resembling lotus leaf veins on the crucible bottom plate, this invention actively modifies the heat conduction path at the bottom of the crucible, making the net downward heat dissipation rate in the central area of ​​the crucible bottom plate greater than that around it, thereby forming a cold core at a fixed point in the center of the melt surface. The position of this cold core is directly determined by the geometry of the crucible bottom plate, rather than being passively formed by the superposition of the entire set of thermal field components. Therefore, the position is geometrically definite, stable, and calibrable, overcoming the defect of traditional cold cores drifting with power and time. Second, by calibrating the position coordinates of the cold core using an infrared thermal imager under empty crucible conditions and establishing a coordinate mapping relationship, this invention replaces the operator's visual experience judgment of the halo with objective coordinate data, eliminating reliance on human experience and significantly improving the consistency and repeatability of crystal pulling. Third, by using the XY fine-tuning platform to position the seed crystal directly above the cold core based on the calibration coordinates, the precise spatial alignment between the seed crystal and the cold core is achieved without the need for repeated trial probes. Furthermore, since the position of the cold core is fixed, the coordinates obtained from the calibration of the same crucible can be reused in multiple growth furnaces, further improving production efficiency and batch consistency.

[0016] Fourth, by acquiring images of the interface between the seed crystal and the melt during the crystal pulling process and using a closed, symmetrical halo as the criterion for single-crystal nucleation, this invention achieves online objective determination of whether crystal nucleation is successful. This determination is then linked to subsequent pulling actions, forming a closed-loop control system that connects structural positioning, coordinate calibration, and online verification. Fifth, and most importantly, there is a significant nonlinear synergistic relationship between the structural positioning of the cold core and the infrared calibration of the cold core coordinates in this invention: on the one hand, the radial heat-conducting channels fix the cold core position, ensuring the coordinates obtained from infrared calibration remain valid throughout the entire crystal pulling cycle; on the other hand, the infrared calibration provides objective coordinates, allowing the cold core formed by the structural positioning to be accurately hit by the seed crystal. These two aspects are mutually enabling and indispensable, resulting in a crystal pulling success rate that exceeds the linear summation of any single measure. Experimental verification shows that using the crystal pulling method of this invention, the crystal pulling success rate can be increased from approximately 70% in existing technologies to over 95%, the single-crystal rate can be increased from approximately 80% to over 98%, and the problem of frequent polycrystalline defects is effectively suppressed. Sixth, the present invention has a simple structure, which can be achieved by simply processing radial channels on the bottom plate of the crucible. The increase in crucible manufacturing cost is less than 5%, and it does not change the rest of the structure of the sapphire growth furnace. It is suitable for various specifications of sapphire growth furnaces, such as 50-200kg class, and is easy to promote in industrialization. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments are briefly introduced below.

[0018] Figure 1This is a top view of the radial heat-conducting channels resembling lotus leaf veins on the bottom plate of the crucible of the present invention.

[0019] Figure 2 This is a schematic diagram of the longitudinal cross-sectional structure of the crucible of the present invention;

[0020] Figure 3 This is a schematic diagram of the process flow for the crystal-driving method of the present invention;

[0021] Figure 4 This is a schematic diagram illustrating the principle of infrared calibration of the cold core position coordinates and seed crystal XY positioning in this invention;

[0022] Figure 5 This is a schematic diagram of the temperature field distribution on the melt surface and the location of the cold core in this invention;

[0023] Figure 6 This is a schematic diagram of the symmetrical halo that appears at the interface between the seed crystal and the melt during the crystal pulling process of this invention.

[0024] In the diagram: 1. Crucible; 11. Crucible base plate; 12. Radial heat-conducting channels; 121. Main vein channel; 13. Convergence zone; 2. Melt; 21. Cold core; 3. Seed crystal; 31. Seed crystal rod; 32. XY fine-tuning platform; 4. Infrared thermal imager; 5. Image acquisition device; 6. Heating element; 7. Reflecting screen; 22. Halo. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0026] like Figure 1 and Figure 2 As shown, the crucible 1 used in this invention is typically a tungsten crucible, and its bottom plate 11 has radially arranged heat-conducting channels 12, mimicking the veins of a lotus leaf, processed on its upper surface. The radially arranged heat-conducting channels 12 include a converging area 13 located at the center of the crucible bottom plate 11 and multiple main vein channels 121 radiating from the converging area 13 towards the edge of the crucible 1. The multiple main vein channels 121 are arranged approximately evenly along the circumference with the converging area 13 as the center, their shape mimicking the main veins radiating from the petiole to the leaf edge on a lotus leaf, with one end converging at the converging area 13 and the other end extending to the edge region of the crucible bottom plate 11. In some embodiments of this invention, the number of main vein channels 121 is 8 to 12, the width of each main vein channel 121 is 2 to 3 mm, the channel depth is 1 to 2 mm, and the diameter of the converging area 13 is 15 to 20 mm.

[0027] like Figure 5As shown, the radial heat-conducting channels 12 on the crucible bottom plate 11 change the heat conduction path at the bottom of the crucible, thereby forming a cold core 21 at a fixed point in the central area of ​​the surface of the melt 2. The formation mechanism is as follows. During the bubble growth process, the crucible 1 is supported by the support component below it and conducts some heat downwards. Compared with the traditional crucible with a flat bottom plate, the multiple main vein channels 121 that radiate from the edge area of ​​the crucible bottom plate 11 to the converging area 13 in this invention increase the effective area of ​​the central area of ​​the crucible bottom plate 11 participating in downward heat conduction. This allows the heat in the edge area of ​​the crucible bottom plate 11 to be concentrated in the central converging area 13 along each main vein channel 121, and efficiently discharged downwards from the crucible bottom plate 11 through the converging area 13. As a result, the net downward heat dissipation rate of the central area of ​​the crucible bottom plate 11 is greater than that of its surroundings. Since the temperature of a certain part of the surface of melt 2 depends on the net downward heat dissipation rate of that part, the enhanced net heat dissipation in the central region of the crucible bottom plate 11 makes the central region of the surface of melt 2 the lowest temperature part, that is, a cold core 21 is formed at a fixed point in the central region of the surface of melt 2. It is worth noting that the radial structure imitating the veins of a lotus leaf is usually used in the field of heat dissipation of electronic devices to evenly spread heat from the center to the periphery to achieve uniform heating. However, in this invention, the radial converging structure is used in reverse to concentrate and enhance the heat dissipation capacity of the center of the crucible bottom plate, thereby creating and locking a lowest temperature point on the surface of the melt. The two have exactly opposite orientations.

[0028] When using the crucible 1 with radially conductive heat-conducting channels 12 of the present invention, the temperature field on the surface of the melt 2 can be considered as being composed of two superimposed parts: one is the basic temperature field formed by components such as the heating element 6 and the reflector 7, which is distributed with a low center and high periphery; the other is a fixed-point low-temperature distribution formed by the radially conductive heat-conducting channels 12 on the crucible bottom plate 11, which enhances downward heat dissipation in the central region of the crucible bottom plate 11, with the center of the crucible bottom plate 11 as the lowest point. The latter superimposed on the former ensures that the lowest point of the surface temperature of the melt 2 is stably locked directly above the center of the crucible bottom plate 11, that is, the position of the cold core 21 is directly determined by the geometry of the crucible bottom plate 11, and is no longer determined solely by the distribution of the basic temperature field. Therefore, even if the basic temperature field changes to some extent due to the adjustment of the heating power, the position of the cold core 21 can remain stable, thereby overcoming the defect of the cold core drifting with power and time in the traditional bubble-forming method. This fixed-point low-temperature distribution is the key difference between the present invention and the prior art that relies solely on adjusting the heating power or the reflector to indirectly affect the cold core.

[0029] Furthermore, the convergence zone 13 is configured such that the thermally conductive cross-section at the center of the crucible bottom plate 11 is larger than the sum of the thermally conductive cross-sections of the main vein channels 121. The stability of the position of the cold core 21 depends on the thermal conductivity of the central region of the crucible bottom plate 11: when relying solely on the geometric convergence of the main vein channels 121, the position of the cold core 21 is prone to shift with fluctuations in heating power; however, the convergence zone 13 further increases the thermally conductive cross-section at the center of the crucible bottom plate 11, giving the central region of the crucible bottom plate 11 a sufficiently large and concentrated downward thermal conductivity. Thus, when the heating power fluctuates to a certain extent, the lowest point of the surface temperature of the melt 2 can still be stably maintained directly above the center of the crucible bottom plate 11, thereby reducing the sensitivity of the position of the cold core 21 to fluctuations in heating power and converging the positional drift of the cold core 21 within a small range.

[0030] In some embodiments of the present invention, the cross-section of each main vein channel 121 can be rectangular, V-shaped, or arc-shaped groove, extending radially from the convergence area 13 to the edge along the crucible bottom plate 11. The width and depth of the main vein channel 121 together determine the degree of modification of the heat conduction path at the bottom of the crucible: the greater the width and depth of the channel, the larger the effective channel cross-section participating in the heat conduction towards the center, the more significant the net downward heat dissipation of the central area of ​​the crucible bottom plate 11 relative to its periphery, and the greater the temperature difference between the cold core 21 and the surrounding melt surface; however, the width and depth of the channel should not be too large, so as not to excessively weaken the structural strength of the crucible bottom plate 11 or make the central low temperature area too concentrated and its size too small. Correspondingly, the number of main vein channels 121 determines the number of paths for heat conduction from the edge area of ​​the crucible bottom plate 11 towards the center. If the number is too small, the heat conduction will be insufficient, the temperature difference of the cold core 21 will be too small, and its position will be easy to drift; if the number is too large, the spacing between adjacent main vein channels 121 will be too small, the processing will be difficult, and the modification of the heat conduction path will tend to be saturated. Therefore, setting the number of main vein channels 121 to 8 to 12, the channel width to 2 to 3 mm, and the channel depth to 1 to 2 mm can achieve a good balance between forming a cold core 21 with a sufficiently low temperature and appropriate size and ensuring the structural strength and machinability of the crucible bottom plate 11.

[0031] Regarding the calibration of the cold core position coordinates and the establishment of the coordinate mapping relationship in step S2, the following method can be used: Place the infrared thermal imager 4 above the crucible 1, ensuring its field of view covers the molten surface of the crucible. Heat the crucible 1 according to the actual growth power regime, with the crucible 1 unloaded and filled with inert gas. After the temperature field of the crucible 1 stabilizes, acquire an image of the temperature field on the molten surface of the crucible. Because the radial heat-conducting channels 12 on the crucible bottom plate 11 make the central region of the crucible bottom plate 11 the lowest temperature area, the region corresponding to the center of the crucible bottom plate 11 appears as the lowest temperature area in the acquired temperature field image. By thresholding the lowest temperature region and calculating its centroid, the position of the cold core 21 in the image coordinate system can be obtained. Then, based on the pre-calibrated geometric correspondence between the field of view of the infrared thermal imager 4 and the XY fine-tuning platform 32 of the seed crystal rod 31, the image coordinates of the cold core 21 are converted into the platform coordinates of the XY fine-tuning platform 32, thereby establishing a mapping relationship between the position coordinates of the cold core 21 and the fine-tuning platform coordinates of the seed crystal rod 31. Since the position of the cold core 21 in this invention is directly determined by the geometric structure of the crucible bottom plate 11 and remains stable, the mapping relationship obtained by the above calibration remains effective in multiple growth furnaces of the same crucible 1, without the need for repeated calibration for each furnace.

[0032] In some embodiments of the present invention, to further improve the accuracy of the cold core 21 position calibration, the position coordinates of the cold core 21 obtained by calibrating the crucible 1 with inert gas before it is filled with material in step S2 can be checked and fine-tuned by combining the observation of the surface temperature field of the melt 2 after the crucible 1 is filled and melted. Since the position of the cold core 21 in the present invention is mainly determined by the radial heat-conducting channels 12 of the crucible bottom plate 11, the coordinates obtained by the empty crucible calibration are highly consistent with the cold core position during actual material growth. Therefore, the above-mentioned verification usually only requires a small correction to the calibration coordinates, which does not affect the repeated use of the calibration mapping relationship in multiple growth furnaces. By combining the empty crucible pre-calibration with the post-loading verification, the alignment of the seed crystal 3 and the cold core 21 can be made more accurate, thereby further improving the crystal pulling success rate and single crystal rate.

[0033] Furthermore, in the crystal-leading method of the present invention, the inert gas introduced in step S2 is preferably argon. On the one hand, argon has stable chemical properties, which can prevent the tungsten crucible 1 and seed crystal 3 from being oxidized during calibration and subsequent growth. On the other hand, introducing argon helps to obtain heat transfer conditions similar to the actual growth atmosphere during the calibration stage, so that the coordinates of the cold core 21 obtained by calibration are closer to the cold core position during actual growth. The seed crystal 3 used is a high-quality alumina single crystal with a defined crystal orientation. Its crystal orientation can be selected as a-axis, c-axis, m-axis, or r-axis according to the crystal orientation of the crystal to be grown. The end face of the seed crystal 3 is cut and ground to ensure that the seed crystal 3 can recrystallize from a single crystal nucleus after contacting the melt 2. In step S4, the seed crystal 3 is slowly lowered to contact the melt 2 at a speed of 0.5 mm / min. The low descent speed can avoid thermal shock cracking caused by the sudden contact of the seed crystal 3 with the high temperature melt 2. After contact, it is kept still for 10 to 15 minutes to allow a stable solid-liquid interface to be fully established between the end face of the seed crystal 3 and the melt 2 and to complete the single crystal nucleation. After that, the pulling can be started, which can effectively reduce the probability of polycrystalline nucleation and dislocation generation.

[0034] The principle behind using the image acquisition device 5 to observe the contact interface between the seed crystal 3 and the melt 2 in step S4, and using a symmetrical halo as the nucleation criterion, is as follows: When the seed crystal 3 is accurately positioned directly above the cold core 21 and in proper contact with the melt 2, the melt 2 around the end face of the seed crystal 3 is cooled by the seed crystal 3, forming a ring-shaped meniscus around the contact interface. This meniscus reflects light from the surface of the melt 2, appearing as a bright halo 22 in the image acquired by the image acquisition device 5. If the seed crystal 3 is precisely aligned with the cold core 21, the temperature field at the contact interface is symmetrical about the central axis of the seed crystal 3, and the resulting halo is closed, symmetrical in shape, and uniform in brightness. Conversely, if the seed crystal 3 deviates from the cold core 21, the temperature on one side of the contact interface is higher than that on the other side, and the meniscus and halo will show openings, skewness, or uneven brightness. Therefore, the closed and symmetrical halo is not only a sign of stable single-crystal nucleation, but also an online verification of the accurate alignment of the seed crystal 3 with the cold core 21. By binding this criterion with the subsequent pulling action, it can be ensured that the pulling stage is only entered after stable single-crystal nucleation, thus avoiding continued growth in the case of polycrystalline or eccentric nucleation.

[0035] In this invention, the positional drift of the cold core 21 refers to the maximum distance that the centroid of the lowest temperature region on the melt surface shifts relative to its calibrated position during the entire crystal-seeding cycle, influenced by factors such as heating power fluctuations and melt convection. A smaller positional drift indicates a more stable position for the cold core 21, more reliable calibrated coordinates, and higher alignment accuracy between the seed crystal 3 and the cold core 21. When using the crucible 1 with radially conductive heat channels 12 of this invention, the positional drift of the cold core 21 can typically be controlled within 0.5 mm, while when using a traditional flat-bottomed crucible, the positional drift often exceeds 2 mm—a difference of several times. This difference is the basis for the precise positioning of the seed crystal in this invention.

[0036] The radial heat-conducting channels 12 on the crucible bottom plate 11 can be directly formed on the upper surface of the crucible bottom plate 11 by means of machining, electrical discharge machining or laser machining. The processing technology is simple. It only adds channel structure to the crucible bottom plate 11 without changing the overall shape and volume of the crucible 1, nor does it change the structure of other components such as the heating element 6, the reflector 7, and the seed crystal rod 31 of the Czochralski growth furnace. Therefore, the increase in manufacturing cost of crucible 1 is very small, usually less than 5%, and it is easy to promote and apply in existing Czochralski growth furnaces.

[0037] Compared to traditional crystal pulling methods that rely on operator experience, this invention transforms the most critical and difficult-to-standardize step in the crystal pulling process—cold core positioning—from the operator's subjective visual judgment of the molten surface halo into objective coordinate calibration based on infrared temperature field images and mechanical positioning based on an XY fine-tuning platform. This ensures consistency in crystal pulling results across different operators and batches, thereby achieving standardization and reproducibility of the crystal pulling process. This is of great significance for the large-scale and stable industrial production of sapphire single crystals: on the one hand, it significantly reduces reliance on the operator's personal experience and skill level, shortening the training cycle for new operators; on the other hand, by improving the success rate and single crystal yield, it reduces material waste and furnace losses caused by crystal pulling failures or polycrystalline defects, thus lowering the production cost of large-size sapphire single crystals.

[0038] Example 1.

[0039] This embodiment uses a 50kg-class sapphire growth furnace via the Czochralski method as an example to illustrate the crystal pulling method of the present invention. The process flow is as follows: Figure 3 As shown, the principle of infrared calibration and seed crystal positioning is as follows: Figure 4 As shown. First, step S1 is performed, providing a tungsten crucible 1. The upper surface of its bottom plate 11 is machined with radial heat-conducting channels 12 that resemble the veins of a lotus leaf. In this embodiment, there are 10 main vein channels 121, each with a channel width of 2.5 mm and a channel depth of 1.5 mm. The diameter of the converging area 13 is 17.5 mm. High-purity alumina raw material is then loaded into the crucible 1. Subsequently, step S2 is performed, with the crucible 1 empty and without material, the furnace body is evacuated to 1×10⁻⁶. -4After Pa, argon gas is introduced, and an infrared thermal imager 4 is used to acquire temperature field images of the melt-bearing surface of crucible 1, i.e., the upper surface of the crucible bottom plate 11 and its adjacent area. The centroid of the lowest temperature region in the temperature field image is extracted, and the centroid coordinates of the lowest temperature region are obtained and calibrated as the position coordinates of the cold core 21. Then, a mapping relationship is established between the position coordinates and the coordinates of the XY fine-tuning platform 32 of the seed crystal rod 31. Next, step S3 is executed. According to the mapping relationship, the seed crystal 3 is positioned directly above the cold core 21 by the XY fine-tuning platform 32 of the seed crystal rod 31. In this embodiment, the stroke of the XY fine-tuning platform 32 is ±5mm and the positioning accuracy is ±0.05mm. Finally, step S4 is executed. After the raw material melts and the surface of the melt 2 is stable, the seed crystal 3 is slowly lowered at a speed of 0.5mm / min to contact the melt 2. After contact, it is kept still for 12min to ensure stable nucleation of the single crystal. During this period, an image acquisition device 5 is used to acquire images of the contact interface between the seed crystal 3 and the melt 2. Figure 6 As shown, when the image of the contact interface presents a closed and symmetrical halo, it is determined that the single crystal has nucleated, and then the seed crystal 3 is pulled to complete the crystal pulling. Upon inspection, the diameter of the cold core 21 in this embodiment is approximately 12 mm, the temperature of the cold core 21 is about 5°C lower than the surrounding melt surface, the positional drift of the cold core 21 is approximately 0.3 mm, the crystal pulling success rate reaches 96%, the single crystal rate reaches 98.5%, and no polycrystalline defects are observed.

[0040] Example 2.

[0041] The crystal-leading method in this embodiment is basically the same as that in Embodiment 1, except that the parameter values ​​of the radial heat-conducting channels 12 of the crucible 1 and some crystal-leading process parameters are taken near the lower limit of the range of this invention. In this embodiment, there are 8 main vein channels 121, each with a channel width of 2 mm and a channel depth of 1 mm, and the diameter of the convergence area 13 is 15 mm; in step S4, the seed crystal 3 remains stationary after contacting the melt 2 for 10 minutes. The remaining steps are the same as in Embodiment 1. Upon testing, the diameter of the cold core 21 in this embodiment is approximately 10 mm, the temperature of the cold core 21 is approximately 3°C lower than the surrounding melt surface, the positional drift of the cold core 21 is approximately 0.4 mm, the crystal-leading success rate reaches 95%, the single crystal rate reaches 98%, and no polycrystalline defects are observed.

[0042] Example 3.

[0043] The crystal-leading method in this embodiment is basically the same as that in Embodiment 1, except that the parameter values ​​of the radial heat-conducting channels 12 of the crucible 1 and some crystal-leading process parameters are taken near the upper limit of the range of this invention. In this embodiment, there are 12 main vein channels 121, each with a channel width of 3 mm and a channel depth of 2 mm, and the diameter of the convergence area 13 is 20 mm; in step S4, the seed crystal 3 remains stationary after contacting the melt 2 for 15 minutes. The remaining steps are the same as in Embodiment 1. Upon testing, the diameter of the cold core 21 in this embodiment is approximately 15 mm, the temperature of the cold core 21 is approximately 8°C lower than the surrounding melt surface, the positional drift of the cold core 21 is approximately 0.3 mm, the crystal-leading success rate reaches 96%, the single crystal rate reaches 98.8%, and no polycrystalline defects are observed. Examples 1 to 3 above show that within the range of 8 to 12 main vein channels 121, channel width of 2 to 3 mm, channel depth of 1 to 2 mm, and diameter of convergence zone 13 of 15 to 20 mm, and within the range of cold core 21 diameter of 10 to 15 mm and cold core 21 temperature of 3 to 8°C lower than the surrounding area, a cold core with small positional drift can be stably formed and a high crystal growth success rate and single crystal rate can be obtained.

[0044] Example 4.

[0045] The crystal pulling method in this embodiment is basically the same as that in Embodiment 1, except that the number of main vein channels 121 is 11, the channel width of each main vein channel 121 is 2.8 mm, the channel depth is 1.8 mm, and the diameter of the convergence area 13 is 18 mm; in step S4, the seed crystal 3 remains stationary after contacting the melt 2 for 13 minutes. The remaining steps are the same as in Embodiment 1. Upon inspection, the diameter of the cold core 21 in this embodiment is approximately 13 mm, the temperature of the cold core 21 is approximately 6°C lower than the surrounding melt surface, the positional drift of the cold core 21 is approximately 0.3 mm, the crystal pulling success rate reaches 96%, the single crystal rate reaches 98.6%, and no polycrystalline defects are observed.

[0046] Example 5.

[0047] The crystal-guiding method in this embodiment is basically the same as that in Embodiment 1, except that this embodiment uses a 200kg-class Czochralski sapphire growth furnace, and the number of main vein channels 121 is 9, with each main vein channel 121 having a channel width of 2.2mm, a channel depth of 1.2mm, and a convergence zone 13 diameter of 16mm; in step S4, the seed crystal 3 remains stationary after contacting the melt 2 for 11 minutes. The remaining steps are the same as in Embodiment 1. The results show that for a 200kg-class large growth furnace, the present invention can also stably and precisely position the cold core at the center of the melt surface and achieve precise seed crystal positioning. Testing showed that the diameter of the cold core 21 in this embodiment is approximately 11mm, the temperature of the cold core 21 is about 4°C lower than the surrounding melt surface, the positional drift of the cold core 21 is approximately 0.4mm, the crystal-guiding success rate reaches 95%, the single crystal rate reaches 98.2%, and no polycrystalline defects are observed. This embodiment demonstrates that the crystal-guiding method of the present invention is applicable to Czochralski sapphire growth furnaces of different specifications, ranging from 50kg to 200kg.

[0048] Comparative Example 1.

[0049] This comparative example uses a traditional tungsten crucible with a flat bottom and no radial heat-conducting channels. The remaining crystal-seeding operations are kept as consistent as possible with Example 1, i.e., the temperature field is also acquired using an infrared thermal imager for seed crystal positioning. Because the cold core of a traditional flat-bottomed crucible is passively formed by the stacking of the entire thermal field components and its position is not fixed, the cold core position coordinates obtained from infrared calibration quickly become invalid due to power adjustment and melt convection during the raw material melting and subsequent holding process. The actual seed crystal positioning still relies on the operator's visual judgment of the halo. Testing revealed that the cold core position drift in this comparative example reached approximately 2.5 mm, the crystal-seeding success rate was only 72%, the single crystal rate was only 80%, and polycrystalline defects were frequent.

[0050] Comparative Example 2.

[0051] This comparative example uses the same crucible 1 as Example 1, with radially shaped heat-conducting channels 12 machined on the base plate. However, the step S2, which involves calibrating the cold core position coordinates using an infrared thermal imager and establishing a coordinate mapping, is omitted. Instead, the seed crystal is positioned based on the operator's visual observation of the halo on the melt surface, following existing technology. Because the crucible 1 in this comparative example can form a cold core at a fixed point on the center of the melt surface, and the cold core position is relatively stable, the crystal pulling effect is improved compared to Comparative Example 1. However, due to the lack of objective cold core coordinates, the seed crystal positioning still suffers from deviations caused by visual experience. Testing showed that the seed crystal's positioning deviation relative to the cold core in this comparative example was approximately 1.2 mm, the crystal pulling success rate was 85%, the single crystal rate was 90%, and polycrystalline defects occurred occasionally.

[0052] Comparative Example 3.

[0053] The crystal pulling method in this comparative example is basically the same as that in Example 1, except that the parameters of the radial heat-conducting channels 12 of the crucible 1 exceed the scope of this invention. Specifically, the number of main vein channels 121 is reduced to 4, and the diameter of the converging region 13 is increased to 25 mm. Because the number of main vein channels 121 is too small, there is insufficient pathway for heat to converge from the edge of the crucible bottom plate 11 to the center, and the enhancement of net downward heat dissipation in the central area of ​​the crucible bottom plate 11 is limited. At the same time, the diameter of the converging region 13 is too large, causing the low-temperature region in the center of the crucible bottom plate 11 to be too diffuse and not concentrated, making it difficult to form a cold core of suitable size and fixed position on the surface of the melt 2. Testing showed that the temperature of the lowest temperature region on the melt surface in this comparative example was only about 1.8°C lower than the surrounding area, and the region diameter was greater than 20 mm, with an unstable position. The positional drift of the cold core was about 1.5 mm, the crystal pulling success rate was only 78%, and the single crystal rate was 85%.

[0054] Comparative Example 4.

[0055] The crystal pulling method in this comparative example is basically the same as that in Example 1, except that the depth of the radial heat-conducting channels 12 in the crucible 1 is too shallow, only 0.3 mm, which is lower than the 1-2 mm range of this invention. Due to the shallow channel depth, the main vein channel 121 does not adequately modify the heat conduction path at the bottom of the crucible, and the enhancement of net downward heat dissipation in the central area of ​​the crucible bottom plate 11 is not significant, failing to form a sufficiently low-temperature and stable cold core at the center of the melt 2 surface. Testing showed that the temperature of the cold core in this comparative example was only about 2.2°C lower than the surrounding area, the positional drift of the cold core was about 1.4 mm, the crystal pulling success rate was only 75%, and the single crystal rate was 83%.

[0056] To more intuitively compare the crystal-driving effects of the various embodiments of the present invention with those of the comparative examples, the key parameters and test results of the above embodiments 1 to 3 and comparative examples 1 to 4 are summarized in Table 1, as shown in Table 1.

[0057] Group Should radial heat conduction channels be provided? Number of main veins / line Convergence area diameter / mm Infrared calibration? Cold core temperature difference / °C Cold core position drift / mm Crystallization success rate / % Single crystal ratio / % Polycrystalline defect situation Example 1 yes 10 17.5 yes 5 0.3 96 98.5 Not seen Example 2 yes 8 15 yes 3 0.4 95 98 Not seen Example 3 yes 12 20 yes 8 0.3 96 98.8 Not seen Comparative Example 1 no — — Attempted but failed — 2.5 72 80 Frequent Comparative Example 2 yes 10 17.5 No (visual inspection) 5 1.2 85 90 Occasional Comparative Example 3 yes 4 25 yes 1.8 1.5 78 85 Occasional Comparative Example 4 Yes (0.3mm deep) 10 17.5 yes 2.2 1.4 75 83 Occasional

[0058] The data shown in Table 1 allows for an in-depth analysis of the technical effects of this invention. Using Comparative Example 1, which employs a traditional flat-bottomed crucible and relies on visual positioning, as a benchmark, its crystal pulling success rate is 72% and its single crystal rate is 80%. When only the cold core structure positioning method implemented by radial heat-conducting channels in this invention is introduced (e.g., Comparative Example 2, which uses a crucible with radial heat-conducting channels but still relies on visual positioning), the crystal pulling success rate increases to 85%, meaning that the cold core structure positioning alone contributes approximately 13 percentage points to the crystal pulling success rate. However, when only infrared calibration is introduced without addressing the cold core stability issue, the coordinates obtained from infrared calibration quickly become invalid due to the drift of the cold core position with power and time (e.g., in Comparative Example 1, even with attempts at infrared acquisition, valid coordinates could not be maintained). Therefore, the contribution of infrared calibration alone to the crystal pulling success rate is very limited, approximately 3 percentage points. Accordingly, if the effects of the cold core structure positioning and infrared calibration are simply linearly superimposed, the expected crystal pulling success rate is at most 72% + 13% + 3%, or approximately 88%. However, when the present invention combines cold core structure positioning (step S1) with infrared coordinate calibration (step S2) and seed crystal coordinate positioning based thereon (step S3), as shown in Examples 1 to 3, the actual crystal pulling success rate reaches more than 95%, exceeding the above linear superposition expectation by more than 7 percentage points, showing a significant nonlinear synergistic effect.

[0059] The fundamental reason for the aforementioned nonlinear synergistic effect lies in the bidirectional enabling relationship between cold core structure positioning and cold core coordinate calibration: on the one hand, it is precisely because the radial heat-conducting channels on the crucible bottom plate geometrically determine and stabilize the position of the cold core that the cold core position coordinates obtained by infrared calibration remain effective throughout the entire cycle from raw material melting and heat preservation to crystal seeding, thus truly guiding the positioning of the seed crystal; on the other hand, it is precisely because infrared calibration provides objective cold core position coordinates that the cold core formed by the structure positioning can be accurately hit by the seed crystal via the XY fine-tuning platform, thereby fully realizing the value of the cold core structure positioning. Both are prerequisites and indispensable; the effect of either measure used alone is greatly reduced. Only by combining the two can a stable and precise docking between the seed crystal and the cold core be achieved, which is the fundamental reason why this invention achieves an improved superlinear crystal seeding success rate. Furthermore, Comparative Examples 3 and 4 further demonstrate that when the parameters of the radial heat-conducting channels, such as the number of main vein channels, the diameter of the convergence zone, and the channel depth, exceed the range defined by this invention, the structural positioning capability of the cold core will be significantly weakened or even fail. At this time, even if infrared calibration and coordinate positioning are retained, it is impossible to obtain a cold core of appropriate size and stable position, and the crystal pulling success rate will drop to a low level of 75% to 78%. This proves the necessity of the radial heat-conducting channel parameter range defined by this invention for forming a stable cold core and achieving accurate crystal pulling.

[0060] Furthermore, as shown in Table 1, the single crystal ratio of each embodiment of the present invention reaches over 98%, while the single crystal ratio of each comparative example is between 80% and 90%. This is consistent with the trend of crystal pulling success rate, and also confirms the synergistic effect of combining cold core structure positioning and infrared coordinate calibration in ensuring single crystal quality. This is because the seed crystal 3 is accurately aligned with the stable cold core 21 and recrystallizes from a single crystal nucleus, avoiding polycrystalline and grain boundary defects caused by multi-point nucleation in the cold region from the source. Thus, while improving the crystal pulling success rate, the single crystal ratio and optical quality of the obtained crystal are correspondingly improved.

[0061] It should be noted that the above-described embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for crystallizing sapphire single crystals based on cold-center positioning using the Czochralski method, characterized in that, Includes the following steps: S1. A crucible is provided, the upper surface of which is provided with radial heat-conducting channels resembling lotus leaf veins. The radial heat-conducting channels are distributed from the converging area in the center of the crucible bottom plate to the edge of the crucible, so that the central area of ​​the crucible bottom plate forms a channel for enhanced downward heat conduction, forming a cold core in the central area of ​​the melt surface; S2. Under the condition that the crucible is not loaded with material and is filled with inert gas, the temperature distribution of the crucible melt bearing surface is obtained by thermal imaging, the lowest temperature area is marked as the position coordinates of the cold core, and a mapping relationship is established between the position coordinates and the coordinates of the seed crystal rod fine-tuning platform; S3. According to the mapping relationship, the seed crystal is positioned directly above the cold core by the XY fine-tuning platform of the seed crystal rod; S4. Lower the seed crystal to contact the melt, keep it stationary until a single crystal nucleation occurs, and then pull up the seed crystal.

2. The method according to claim 1, characterized in that, The radial heat-conducting channels include multiple main vein channels radiating from the convergence area to the edge of the crucible. Each main vein channel gathers heat from the edge area of ​​the crucible bottom plate to the convergence area and then conducts it downwards from the convergence area to form a cold core in the central area of ​​the melt surface with a temperature lower than that around it.

3. The method according to claim 2, characterized in that, The convergence zone is constructed such that the thermally conductive cross-section at the center of the crucible bottom plate is larger than the sum of the thermally conductive cross-sections of the main channels, thereby reducing the sensitivity of the location of the cold core to fluctuations in heating power.

4. The method according to claim 3, characterized in that, In step S2, an infrared thermal imager is used to acquire a temperature field image of the crucible melt bearing surface, and the centroid of the lowest temperature region in the temperature field image is extracted to obtain the position coordinates of the cold center.

5. The method according to claim 4, characterized in that, In step S3, the seed crystal is positioned directly above the cold core using the XY fine-tuning platform according to the mapping relationship, and the position coordinates are reused in multiple single crystal growths in the same crucible.

6. The method according to claim 5, characterized in that, In step S4, after the seed crystal contacts the melt, an image of the contact interface between the seed crystal and the melt is acquired. When the image of the contact interface shows a closed and symmetrical halo, it is determined that the single crystal has nucleated, and the seed crystal is pulled up after the single crystal has nucleated.

7. The method according to claim 1, characterized in that, The radial heat-conducting channels have 8 to 12 main veins, a channel width of 2 to 3 mm, a channel depth of 1 to 2 mm, and a convergence area diameter of 15 to 20 mm.

8. The method according to claim 1, characterized in that, The diameter of the cold core is 10-15 mm, and the temperature of the cold core is 3-8°C lower than that of the surrounding melt surface.

9. The method according to claim 1, characterized in that, In step S4, the seed crystal descends at a speed of 0.5 mm / min to contact the melt, and remains stationary for 10 to 15 minutes after contact.

10. The method according to claim 1, characterized in that, The inert gas mentioned in step S2 is argon; the XY fine-tuning platform of the seed crystal rod has a stroke of ±5mm and a positioning accuracy of ±0.05mm.

Images