Partition cooling system, cooling control method, crystal bar, wafer, and medium

Abstract

The present disclosure provides a partition cooling system, a cooling control method, a crystal bar, a wafer and a medium, comprising a partition independent heat exchange component, a thermal state sensing unit and a control unit, wherein the partition independent heat exchange component is arranged as multiple independent heat exchange levels along the vertical direction of crystal pulling, and the heat exchange levels respectively apply local cooling intensity to multiple height spatial domains; the thermal state sensing unit is arranged along the growth path of the single crystal silicon ingot and is configured to obtain real-time thermodynamic state data corresponding to multiple height spatial domains on the growth path of the silicon ingot during single crystal growth. The control unit is communicatively coupled with the partition independent heat exchange component and the thermal state sensing unit, used to receive real-time thermodynamic state data and independently control the partition independent heat exchange component.

Description

Technical Field

[0001] This disclosure relates to the field of single-crystal silicon preparation technology, and in particular to a zoned cooling system, a cooling control method, a crystal rod, a wafer, and a dielectric. Background Technology

[0002] In the field of semiconductor material preparation, the Czochralski method (CZ method) is one of the mainstream processes for preparing single-crystal silicon ingots. This process involves melting polycrystalline silicon raw materials in a furnace to form a silicon melt, then slowly pulling and rotating a seed crystal from the surface of the silicon melt, allowing silicon atoms to grow in an orderly manner according to the crystal structure of the seed crystal, ultimately forming a single-crystal silicon ingot. During the Czochralski method of single-crystal silicon growth, the cooling process of the crystal has a crucial impact on the internal quality of the single-crystal silicon ingot. Among these factors, the generation and control of oxygen precipitation-related bulk microdefects (BMDs) are important indicators for evaluating the quality of single-crystal silicon ingots.

[0003] The cooling methods in related technologies can lead to uneven axial temperature gradient distribution, making it difficult to maintain the cooling rate within the BMD sensitive nucleation temperature range within the process requirements. Consequently, it is impossible to control the nucleation density of BMD and its axial distribution uniformity, resulting in poor axial BMD distribution uniformity of single-crystal silicon ingots. This also makes the process window for single-crystal silicon growth narrow and difficult to adapt to different growth process conditions. Summary of the Invention

[0004] The first aspect of this disclosure provides a partitioned cooling system for Czochralski single-crystal silicon growth, wherein the partitioned cooling system for Czochralski single-crystal silicon growth includes partitioned independent heat exchange components and a control unit; the partitioned independent heat exchange components are arranged in multiple independent heat exchange levels along the vertical direction of crystal pulling, and the heat exchange levels apply local cooling intensity to multiple height spatial domains respectively; the control unit establishes communication coupling with the partitioned independent heat exchange components, and controls the partitioned independent heat exchange components according to real-time thermodynamic state data of multiple height spatial domains to modulate the local cooling intensity of each height spatial domain.

[0005] The second aspect of this disclosure provides a cooling control method for Czochralski single-crystal silicon growth, applied to the aforementioned partitioned cooling system for Czochralski single-crystal silicon growth. The method includes acquiring real-time thermodynamic state data corresponding to multiple height spatial domains along the growth path of the single-crystal silicon ingot; and controlling multiple independent heat exchange levels in partitioned independent heat exchange components based on the real-time thermodynamic state data to modulate the local cooling intensity of each height spatial domain, wherein the multiple independent heat exchange levels are arranged around the growth path of the single-crystal silicon ingot and along the vertical direction.

[0006] The third aspect of this disclosure provides a crystal rod, which is prepared using the partitioned cooling method of Czochralski single-crystal silicon growth according to the second aspect; wherein the average density of bulk microdefects along the entire pulling axis of the crystal rod is 1.40 × 10⁻⁶. 8 pcs / cm 3 ~2.0×10 8 pcs / cm 3 Between; and the axial standard deviation of the bulk microdefect density of the crystal rod is less than or equal to 0.8 × 10⁻⁶. 8 pcs / cm 3 .

[0007] A fourth aspect of this disclosure provides a wafer obtained from a crystal ingot of the second aspect by cutting, grinding, polishing, and cleaning; wherein the dislocation density of the wafer is ≤5.0×10⁻⁶. 2 cm -2 And the dislocation rate is ≥99.5%.

[0008] This disclosure provides a partitioned cooling system, cooling control method, crystal rod, wafer, and dielectric. By setting multiple independent heat exchange levels arranged along the vertical direction of crystal pulling within the furnace, a partitioned independent heat exchange structure is formed. In conjunction with a thermal state sensing unit collecting real-time thermodynamic states of each height spatial domain, and a control unit independently modulating the cooling intensity of each heat exchange level based on the real-time thermodynamic state data of each height spatial domain, differentiated and precise cooling of different height positions along the single-crystal silicon ingot growth path is achieved, effectively avoiding the problem of abrupt changes in the axial temperature gradient of the crystal. Based on the dynamic control method using real-time thermodynamic state data, the cooling intensity of each region can be adjusted in real time according to the actual temperature changes during crystal growth, precisely controlling the kinetics of BMD nucleation and growth, resulting in a more uniform axial defect and horizontal BMD distribution in the single-crystal silicon ingot. Simultaneously, the differentiated cooling control method can provide customized cooling curves for different single-crystal silicon growth processes, significantly widening the process window and improving the adaptability of the device to different growth process conditions. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the overall structure of a basic crystal pulling furnace for a Czochralski single crystal silicon growth apparatus provided in this disclosure.

[0010] Figure 2 This is a schematic diagram of the structure of a partitioned independent heat exchange component provided in this disclosure.

[0011] Figure 3 This is a schematic diagram of the height spatial domain division and corresponding cooling jacket configuration of a single crystal silicon ingot growth path provided in this disclosure.

[0012] Figure 4This is a cross-sectional view of the internal structure of a fluid cooling sleeve provided in this disclosure.

[0013] Figure 5 This is a schematic diagram of the control connection structure of a cooling system provided in this disclosure.

[0014] Figure 6 This is a flowchart illustrating an iterative adjustment logic executed by a control unit, as provided in this disclosure.

[0015] Figure 7 This is a schematic diagram of the sensor deployment structure of a thermal state sensing unit provided in this disclosure.

[0016] Figure 8 This is a schematic diagram of the overall process of a cooling control method for Czochralski single-crystal silicon growth provided in this disclosure.

[0017] Figure 9 This is a schematic diagram of a process for differentiating cooling intensity control provided in this disclosure.

[0018] Figure 10 A comparison curve of the distribution of bulk micro-defect density along the crystal pulling axis between a conventional cooling method and the cooling control scheme of this disclosure. Detailed Implementation

[0019] To make the objectives, technical solutions, and technical effects of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings and reference numerals. The described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0020] In the description of the embodiments disclosed herein, it should be noted that the terms "vertical direction," "axial direction," and "height spatial domain," etc., are all based on the direction of crystal pulling during the Czochralski method of single-crystal silicon growth. Local cooling intensity refers to the rate at which heat exchange layers exchange heat with the single-crystal silicon ingot within the corresponding height spatial domain; the greater the cooling intensity, the faster the crystal cools down. Real-time thermodynamic state data mainly includes temperature data at different height positions of the single-crystal silicon ingot, and may also include derived data such as the rate of temperature change. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; they are only used to distinguish different regions, components, or steps.

[0021] Reference Figure 1The crystal pulling furnace used in the Czochralski method for single-crystal silicon growth is the fundamental core equipment for preparing single-crystal silicon ingots. Its basic structure mainly consists of key parts such as the furnace body 101, crucible assembly 102, heating assembly 103, and seed crystal pulling assembly 104. These parts work together to provide the basic process environment and operating conditions for the growth of single-crystal silicon ingots. The furnace body 101 is an integral cylindrical sealed cavity structure made of high-temperature resistant and high-sealing metal material, which can provide a controllable growth environment for single-crystal silicon growth, including vacuum and inert gas protection. The side walls of the furnace body 101 are also equipped with auxiliary structures such as exhaust ports, gas inlets, and observation windows. The exhaust ports are connected to external vacuum equipment to adjust the vacuum level inside the furnace. The gas inlets can introduce inert gases such as argon into the furnace to create an oxygen-free growth atmosphere. The observation windows are made of high-temperature resistant transparent quartz material, allowing operators to observe the growth status of the single-crystal silicon ingot inside the furnace in real time.

[0022] The crucible assembly 102 is located at the bottom center inside the furnace body 101. It is used to contain polycrystalline silicon raw materials and form silicon melt. The crucible assembly 102 has a double-layer nested structure, including an inner quartz crucible 1021 and an outer graphite crucible 1022. The quartz crucible 1021 is in direct contact with the polycrystalline silicon raw materials. Its chemical properties are stable and it will not react with the silicon raw materials at high temperatures. The graphite crucible 1022 is fitted on the outside of the quartz crucible 1021. On the one hand, it plays a supporting and fixing role for the quartz crucible 1021. On the other hand, it has good heat preservation performance, which can reduce the heat loss in the furnace and ensure the stability of the internal temperature of the crucible.

[0023] The heating component 103 is arranged around the outer periphery of the crucible component 102. In this embodiment, the heating component 103 is a graphite heater, which converts electrical energy into heat energy through electric heating to provide the high-temperature heat required for melting the polycrystalline silicon raw material in the crucible component 102. At the same time, the heating component 103 can also achieve precise control of the temperature distribution in the furnace by adjusting the heating power, ensuring that the silicon raw material in the quartz crucible 1021 can be uniformly melted to form silicon melt 105, and maintaining the temperature of silicon melt 105 within the preset process range.

[0024] The seed crystal pulling assembly 104 is located at the top center of the furnace body 101 and is the core execution component for realizing the pulling growth of single crystal silicon ingots. The seed crystal pulling assembly 104 and the crucible assembly 102 rotate in opposite directions, so that silicon atoms can be deposited and grown in an orderly manner according to the crystal structure of the seed crystal during the pulling process, and finally form a cylindrical single crystal silicon ingot 106.

[0025] In the traditional crystal pulling furnace operation, polycrystalline silicon raw material is put into a quartz crucible 1021. The heating component 103 is energized to heat the polycrystalline silicon raw material to melt it and form a silicon melt 105. The seed crystal pulling component 104 sends the seed crystal to the surface of the silicon melt 105 and completes the seed crystal fusion. Then, it pulls the seed crystal upward and rotates it at a preset speed. The silicon melt 105 gradually crystallizes on the surface of the seed crystal and continues to grow as the seed crystal is pulled, eventually forming a single crystal silicon ingot 106 that meets the specifications.

[0026] In the practical application of traditional crystal pulling furnaces, in order to control the growth rate and internal defect type of the monocrystalline silicon ingot 106, a single bottom water-cooling jacket 107 is usually installed at the bottom inside the furnace body 101. This water-cooling jacket 107 is used to cool the growing monocrystalline silicon ingot 106. However, this single cooling method has significant technical drawbacks. The water-cooling jacket 107 can only cool the bottom area of ​​the monocrystalline silicon ingot 106, and the cooling rate is difficult to control precisely. After the monocrystalline silicon ingot 106 leaves the bottom water-cooling jacket 107, the axial temperature gradient will change abruptly, and the temperature at different heights will vary. The significant increase in differences makes it impossible to differentiate the cooling intensity at different heights of the monocrystalline silicon ingot 106. Furthermore, during the growth of the monocrystalline silicon ingot 106, the nucleation and growth of oxygen precipitation-related bone marrow deposits (BMDs) are extremely sensitive to temperature and cooling rate. This undifferentiated, single cooling method cannot achieve precise control over BMD nucleation and growth. This not only hinders the reasonable formation of other defects and impurities within the crystal but also results in extremely poor uniformity of BMD distribution along the axial direction of the monocrystalline silicon ingot 106. It also narrows the process window for monocrystalline silicon growth, making it difficult to adapt to the production needs of monocrystalline silicon ingots with different specifications and process requirements.

[0027] To address the aforementioned technical problems caused by the single cooling method in traditional crystal pulling furnaces, this disclosure provides a zoned cooling system for Czochralski single-crystal silicon growth. This cooling system includes a zoned independent heat exchange component 2 disposed inside the furnace body 101, a thermal state sensing unit (not shown), and a matching control unit 3 disposed outside the furnace body 101. While retaining the basic crystal pulling function of the crystal pulling furnace, it achieves precise, dynamic, and zoned control of the cooling process during the growth of single-crystal silicon ingot 106. The core components of this device and the structure, function, and coordination relationship of each component are as follows.

[0028] The furnace body 1 of the crystal pulling apparatus disclosed herein is based on the furnace body 101 of a traditional crystal pulling furnace with adaptive improvements. The furnace body 1 retains conventional components required for the Czochralski method of single crystal silicon growth, such as the crucible assembly 102, heating assembly 103, and seed crystal pulling assembly 104. The connection relationship and working principle of each conventional component are consistent with those of the traditional crystal pulling furnace, ensuring that the basic pulling growth process of the single crystal silicon ingot 106 can proceed smoothly. In order to accommodate the installation and operation of the partitioned independent heat exchange assembly 2, an installation space is reserved on the outer periphery of the growth path of the single crystal silicon ingot 106 inside the furnace body 1. This installation space extends along the vertical direction of crystal pulling, ensuring that the partitioned independent heat exchange assembly 2 can be arranged around the growth path of the single crystal silicon ingot 106. At the same time, auxiliary structures such as wiring ports and pipe connection ports are added to the side wall of the furnace body 1. The wiring ports are used to realize the wiring connection between the control unit 3 and the partitioned independent heat exchange assembly 2, and the pipe connection ports are used to realize the connection between the cooling medium delivery pipeline and the partitioned independent heat exchange assembly 2.

[0029] In some examples, the partitioned independent heat exchange component 2 is arranged in the reserved installation space inside the furnace body 1 around the growth path of the monocrystalline silicon ingot 106. Specifically, it is arranged on the outer periphery of the growth path of the monocrystalline silicon ingot 106 and maintains a certain reasonable distance from the monocrystalline silicon ingot 106. This distance can ensure efficient heat exchange between the partitioned independent heat exchange component 2 and the monocrystalline silicon ingot 106, and can effectively prevent it from contacting the monocrystalline silicon ingot 106 which is slowly pulled and rotated during the growth process. At the same time, this arrangement will not interfere with the normal operation of conventional components such as the heating component 103 and the crucible component 102 inside the furnace body 1.

[0030] The partitioned independent heat exchange components 2 are arranged in multiple independent heat exchange levels 21 along the vertical direction of crystal pulling. "Independent" means that each heat exchange level 21 is an independent structural unit, and its cooling medium supply and heat exchange process do not interfere with each other. Each heat exchange level 21 corresponds to a height spatial domain on the growth path of the monocrystalline silicon ingot 106, and can independently apply local cooling intensity to the monocrystalline silicon ingot 106 within the corresponding height spatial domain. That is, each heat exchange level 21 can independently adjust its cooling rate according to process requirements, thereby achieving differentiated cooling of different height positions of the monocrystalline silicon ingot 106.

[0031] The thermal state sensing unit is continuously deployed along the pulling growth path of the single crystal silicon ingot and works continuously throughout the entire growth process of the single crystal. It is used to collect and output real-time thermodynamic state data such as temperature, temperature gradient and temperature change rate in multiple spatial domains at different heights along the silicon ingot growth path. This enables the control unit to grasp the thermal field state at different axial positions and provides real-time and reliable data support for the independent adjustment and control of the zoned cooling system.

[0032] The control unit 3 is located outside the furnace body 1, and it establishes communication coupling with the independent heat exchange components 2. The communication coupling method can be selected as wired or wireless communication depending on the actual production scenario. If wired communication is used, the control line of the control unit 3 can be connected to the independent heat exchange components 2 through the wiring port on the side wall of the furnace body 1. If wireless communication is used, signal transmission between the two can be achieved through a wireless transmission module.

[0033] The control unit 3 can be an industrial control computer, programmable logic controller or other device with data processing and instruction output functions. It has a pre-set cooling control logic for monocrystalline silicon growth. Based on the real-time thermodynamic state data of multiple height spatial domains on the growth path of monocrystalline silicon ingot 106, it sends precise control commands to the partitioned independent heat exchange component 2. This controls the working state of each heat exchange level 21 in the partitioned independent heat exchange component 2, and modulates the local cooling intensity of each height spatial domain, so that the monocrystalline silicon ingot 106 in each height spatial domain can be cooled according to the preset process curve.

[0034] In the actual operation of the crystal pulling device disclosed herein, after the crucible assembly 102, heating assembly 103, and seed crystal pulling assembly 104 inside the furnace body 1 complete the preparation of silicon melt 105 and the welding of seed crystal according to the conventional crystal pulling process, the seed crystal is pulled and rotated at a preset speed, and the single crystal silicon ingot 106 begins to grow normally.

[0035] At this time, each heat exchange level 21 of the partitioned independent heat exchange component 2 starts working simultaneously to perform preliminary cooling on different height spatial domains of the monocrystalline silicon ingot 106. The control unit 3 acquires real-time thermodynamic state data of each height spatial domain in real time, analyzes and processes the data, and sends control commands to the partitioned independent heat exchange component 2 based on the analysis results. This dynamically adjusts the local cooling intensity of each heat exchange level 21, so that the cooling rate of each height position of the monocrystalline silicon ingot 106 matches the crystal growth rate, avoiding sudden changes in axial temperature gradient. This achieves precise control over the entire process of BMD nucleation and growth, and ultimately effectively improves the internal quality of the monocrystalline silicon ingot 106.

[0036] In this embodiment, the partitioned cooling system for the Czochralski method of single crystal silicon growth disclosed herein, formed by adding a partitioned independent heat exchange component 2 and configuring a dedicated control unit 3, achieves precise, dynamic, and partitioned control of the cooling process during the growth of single crystal silicon ingot 106, effectively solving the technical defects caused by the single cooling method of traditional crystal pulling furnaces.

[0037] This disclosed device, through its partitioned independent heat exchange structure and dynamic control method based on real-time data, can significantly improve the uniformity of 106-axis defects and BMD levels in monocrystalline silicon ingots, while also greatly expanding the process window to adapt to the production needs of monocrystalline silicon ingots with different specifications, resistivity, and process requirements.

[0038] In some exemplary embodiments of this disclosure, reference is made to Figure 2 The heat exchange layer 21 in the partitioned independent heat exchange component 2 adopts the structure of fluid cooling sleeves. Multiple fluid cooling sleeves are stacked sequentially along the vertical direction of crystal pulling to form the overall partitioned independent heat exchange component 2. It has the advantages of high cooling efficiency, fast control response and strong structural adaptability. It can be highly matched with the spatial layout inside the furnace body 1. At the same time, it can achieve uniform cooling of the circumference of the single crystal silicon ingot 106 and avoid uneven temperature in the circumference of the single crystal silicon ingot 106.

[0039] Each heat exchange stage 21 has a hollow cylindrical cooling sleeve, which can be made of a metal material with good thermal conductivity. The wall thickness of the cooling sleeve is designed according to the internal temperature environment and cooling requirements of the furnace body 1, which will not be elaborated here.

[0040] Each fluid cooling sleeve is an independent structural unit, with separate channels for the supply and discharge of the cooling medium. Specifically, each fluid cooling sleeve is equipped with independent cooling medium inlet and outlet pipes. The inlet pipe connects to an external cooling medium supply system to introduce the cooling medium into the sleeve, while the outlet pipe connects to an external cooling medium recovery system to discharge the cooled medium after heat exchange, thus achieving the recycling of the cooling medium.

[0041] The independent piping design ensures that the cooling processes of each fluid cooling sleeve do not interfere with each other. The flow rate, pressure, and temperature of the cooling medium in each fluid cooling sleeve can be adjusted independently, laying a solid structural foundation for independent cooling control in each height space. In this embodiment, the cooling medium can be deionized water, or antifreeze such as ethylene glycol solution can be used as the cooling medium to prevent the cooling medium from freezing at low temperatures. Specific settings can be customized according to requirements, which will not be elaborated here.

[0042] Multiple fluid cooling sleeves are coaxially stacked along the vertical direction of crystal pulling, with their overall central axis coinciding with the growth path of the monocrystalline silicon ingot 106. This ensures that each fluid cooling sleeve is directly aligned with its corresponding height position on the monocrystalline silicon ingot 106, achieving precise cooling of different height spatial regions of the monocrystalline silicon ingot 106. The stacking spacing of the fluid cooling sleeves can be flexibly adjusted according to the growth process requirements of the monocrystalline silicon ingot 106.

[0043] Meanwhile, the height of the fluid cooling sleeve can also be customized according to actual production needs. The height of the fluid cooling sleeve directly corresponds to the range of its cooling coverage height space. For areas where the temperature changes rapidly and the cooling control requirements are high during the growth of single crystal silicon ingot 106, such as the main nucleation and growth area of ​​BMD, the height of the corresponding fluid cooling sleeve can be appropriately increased to expand the cooling control range of the area. For areas where the temperature changes slowly and the cooling control requirements are low, the height of the fluid cooling sleeve can be appropriately reduced to save space inside the furnace body 1 while ensuring the cooling effect.

[0044] In this embodiment, a structural design using multiple independent fluid cooling sleeves stacked together to form a partitioned independent heat exchange assembly 2 not only achieves differentiated cooling control at different height positions of the monocrystalline silicon ingot 106, but also provides excellent ease of disassembly, assembly, and maintenance. If a single fluid cooling sleeve malfunctions, it can be disassembled and repaired or replaced individually without disassembling the entire partitioned independent heat exchange assembly 2, significantly reducing equipment maintenance costs and downtime.

[0045] In other alternative embodiments of this disclosure, the independent heat exchange level is not limited to a fluid cooling sleeve structure. It can also employ a semiconductor thermoelectric cooling (Peltier effect) annular temperature control array, or a zoned independent inert gas (such as argon) jet cooling annular device equipped with a precision electromagnetic control valve. The above structures can also achieve independent local cooling intensity modulation for different height spatial domains.

[0046] In this embodiment of the disclosure, reference is made to Figure 3 Based on the temperature characteristics at different heights during the growth of the single-crystal silicon ingot 106, the crystal growth state, and the laws governing BMD nucleation and growth are considered, the multiple height spatial domains along the growth path of the single-crystal silicon ingot 106 are sequentially divided into a first region 61, a second region 62, and a third region 63 along the vertical direction of crystal pulling. The division of these three regions covers the entire growth area of ​​the single-crystal silicon ingot 106 from the surface of the silicon melt 105 upwards. Each region corresponds to a key stage in the growth of the single-crystal silicon ingot 106, and its temperature range and BMD growth state are significantly different.

[0047] The first region 61 is the lower high-temperature region near the surface of the silicon melt 105. This region is where the single crystal silicon ingot 106 has just been pulled out of the silicon melt 105. The temperature is relatively high, with the target temperature range set at 1250℃~1350℃. It is in the initial stage of the transition from the high-temperature molten state to the solid state. At this time, the crystal structure of the single crystal silicon ingot 106 has not yet been fully formed, which is a critical stage where dislocation defects are prone to occur. At the same time, this region is the initial stage of vacancy defect formation. The initial supersaturation of vacancy defects directly affects the subsequent nucleation and growth of BMD. Therefore, the cooling process in this region should be mainly gentle cooling to avoid excessive thermal stress caused by excessive cooling.

[0048] The second region 62 is a mid-temperature zone located above the first region 61. The temperature in this region ranges from 800℃ to 1000℃, making it a sensitive temperature range for BMD nucleation and growth. It is also the core region for controlling BMD formation. The number and uniformity of BMD nuclei are determined by the cooling rate and temperature control precision in this region. Therefore, this region requires strong cooling intensity and a high cooling rate to ensure rapid, abundant, and uniform formation of BMD nuclei. The third region 63 is a top low-temperature zone located above the second region 62. The target temperature in this region is between 600℃ and 800℃. The crystal structure of the single-crystal silicon ingot 106 is basically finalized at this stage, and BMD is in a slow growth phase. If the cooling intensity is too high, the BMD will grow too quickly and too large, forming harmful defects and affecting the internal quality of the single-crystal silicon ingot 106. Therefore, the cooling process in this region also focuses on gentle cooling to precisely control the final growth rate of BMD.

[0049] Corresponding to the three-region division of the height space domain, the multiple fluid cooling sleeves in the partitioned independent heat exchange component 2 include at least a first cooling sleeve 211 configured in the first region 61, a second cooling sleeve 212 configured in the second region 62, and a third cooling sleeve 213 configured in the third region 63. The three cooling sleeves are stacked sequentially along the vertical direction of crystal pulling, and independently control the cooling of the first region 61, the second region 62, and the third region 63 respectively. The height and material of the three cooling sleeves are customized according to the cooling requirements of the corresponding regions.

[0050] The control unit 3 is configured to control the local cooling intensity applied by the second cooling jacket 212 to be greater than the local cooling intensity applied by the first cooling jacket 211 and the third cooling jacket 213, forming a differentiated cooling intensity distribution rule with strong intensity in the middle and weak intensity at both ends. This rule is highly compatible with the temperature characteristics of each region and the nucleation and growth law of BMD, and can achieve precise control over the entire process of BMD nucleation and growth.

[0051] For the lower high-temperature region of the first region 61, the control unit 3 controls the first cooling jacket 211 to apply a weak local cooling intensity, using a gentle cooling method to cool the monocrystalline silicon ingot 106 in this region. Specifically, this can be achieved by reducing the flow rate of the cooling medium in the first cooling jacket 211 and increasing the temperature of the cooling medium. In this embodiment, the flow rate of the cooling medium in the first cooling jacket 211 is controlled at 10~20L / h, and the cooling medium pressure is controlled at 0.2MPa~0.4MPa. The gentle cooling method allows the monocrystalline silicon ingot 106 to cool down slowly in this region, avoiding excessive thermal stress caused by excessively rapid cooling, thereby effectively preventing the introduction of dislocation defects. At the same time, it can reasonably control the initial supersaturation of vacancy-type defects, laying a good foundation for the uniform nucleation of BMD in the subsequent second region 62.

[0052] For the central medium-temperature zone of the second region 62, the control unit 3 controls the second cooling jacket 212 to apply the maximum local cooling intensity, using a strong cooling method to cool the monocrystalline silicon ingot 106 in this region. Specifically, this can be achieved by increasing the flow rate of the cooling medium in the second cooling jacket 212, reducing the temperature of the cooling medium, and increasing the pressure of the cooling medium. In this embodiment, the flow rate of the cooling medium in the second cooling jacket 212 is controlled at 20-30 L / h, and the pressure of the cooling medium is controlled at 0.3 MPa~0.5 MPa. The strong cooling method can significantly improve the cooling rate of this region, allowing the monocrystalline silicon ingot 106 to cool rapidly within the BMD sensitive nucleation temperature range, enabling the oxygen precipitation nuclei (BMD nuclei) to form quickly, in large quantities, and uniformly in a short time, preventing a few BMD nuclei from growing too large due to slow cooling, and ensuring the quantity and uniformity of BMD.

[0053] For the top low-temperature zone of the third region 63, the control unit 3 controls the third cooling jacket 213 to apply a weaker local cooling intensity equivalent to that of the first cooling jacket 211, also employing a gentle cooling method. In this embodiment, the flow rate and temperature of the cooling medium in the third cooling jacket 213 are consistent with those in the first cooling jacket 211, controlled at 15L / min to 25L / h respectively, and the cooling medium pressure is controlled at 0.2MPa to 0.4MPa. This cooling method can control the cooling rate of this region, thereby controlling the final growth rate of the BMD, keeping the BMD size within a reasonable range required by the process, avoiding the BMD from growing too large and becoming a harmful defect, while ensuring that the overall cooling process of the single crystal silicon ingot 106 is gradual, preventing stress defects from being generated inside the crystal due to excessively rapid cooling in the later stages.

[0054] In this embodiment, the height spatial domain is divided into three functional regions and corresponding dedicated cooling jackets are configured. At the same time, a differentiated cooling intensity distribution rule with strong cooling in the middle and weak cooling at both ends is formulated, which realizes precise and customized control of the cooling process during the growth of monocrystalline silicon ingot 106. This configuration and control method are completely in line with the dynamic process of BMD nucleation and growth, and can effectively improve the uniformity of axial BMD distribution of monocrystalline silicon ingot 106.

[0055] In this embodiment of the disclosure, reference is made to Figure 4Each fluid cooling sleeve has a cooling medium inlet 214 and an outlet 215 at both ends. The inlet 214 is located at the lower end of the fluid cooling sleeve and connects to an external cooling medium supply pipeline to introduce the cooling medium into the fluid cooling sleeve. The outlet 215 is located at the upper end of the fluid cooling sleeve and connects to an external cooling medium recovery pipeline to discharge the cooling medium after heat exchange. This design, with the inlet 214 at the lower end and the outlet 215 at the upper end, allows the cooling medium to enter from the bottom and exit from the top of the fluid cooling sleeve. This ensures that the cooling medium flows fully from bottom to top within the fluid cooling sleeve, completely covering the entire internal space and avoiding dead zones. This guarantees that the entire sleeve wall can fully exchange heat with the cooling medium, improving overall cooling efficiency.

[0056] The fluid cooling sleeve has multiple baffles 216 arranged in an alternating pattern inside. The baffles 216 are integrally formed with the fluid cooling sleeve and are made of the same material as the fluid cooling sleeve, which ensures the stability of the structure and good thermal conductivity. The thickness of the baffles 216 can be set according to the requirements, which will not be elaborated here.

[0057] Multiple baffles 216 extend alternately from the inner walls of opposite sides of the fluid cooling sleeve inward, that is, two adjacent baffles 216 extend inward from the inner walls of the left and right sides of the fluid cooling sleeve, respectively. The extension length of the baffles 216 is 1 / 2 to 2 / 3 of the inner diameter of the fluid cooling sleeve, and does not reach the central axis of the fluid cooling sleeve, so that there is a channel for the flow of cooling medium between adjacent baffles 216.

[0058] Multiple staggered baffles 216 collectively define a tortuous, continuous flow channel 217 within the fluid cooling jacket, connecting the inlet 214 and the outlet 215. After entering the fluid cooling jacket from the inlet 214, the cooling medium cannot flow directly to the outlet 215 in a straight line. Instead, it must continuously change its flow direction along the tortuous, continuous flow channel 217, flowing upwards in an "S" shape before finally exiting from the outlet 215. This design of the tortuous, continuous flow channel 217 extends the flow path of the cooling medium within the fluid cooling jacket, increases the contact time between the cooling medium and the inner wall of the cooling jacket, and simultaneously creates turbulence during the flow, enhancing the convective heat transfer effect between the cooling medium and the inner wall of the cooling jacket, thereby significantly improving the sufficiency and efficiency of heat exchange.

[0059] The number of baffles 216 can be adjusted according to the height and diameter of the fluid cooling sleeve. The greater the height and diameter of the fluid cooling sleeve, the more baffles 216 there are, in order to ensure the tortuosity of the flow channel and the flow path of the cooling medium.

[0060] In actual operation, the cooling medium enters the fluid cooling sleeve through the inlet 214 and flows upward along the tortuous continuous flow channel 217, fully exchanging heat with the sleeve wall and absorbing the heat transferred from the monocrystalline silicon ingot 106 by the sleeve wall. The cooled medium, after its temperature rises, is discharged from the outlet 215 and enters the cooling medium recovery system for cooling treatment. It then re-enters the fluid cooling sleeve through the inlet 214, thus realizing the recycling of the cooling medium.

[0061] In this embodiment of the disclosure, reference is made to Figure 5 The control unit 3 internally houses a process setting unit 31, which is the core functional module within the control unit 3. Implemented through software programming, the process setting unit 31 is integrated into the processor of the control unit 3 and is used to pre-store various process parameters and cooling control logic for monocrystalline silicon growth. The process setting unit 31 includes cooling intensity control parameters for each height spatial domain corresponding to monocrystalline silicon ingots of different specifications, resistivity, and process requirements, as well as the flow rate range, pressure range, temperature threshold, and iterative adjustment delay time for each height spatial domain.

[0062] The process setting unit 31 can be equipped with a visual operation interface, through which operators can modify, input, and save pre-stored process parameters. It can also optimize and adjust the cooling control logic based on actual production process test results, making the cooling control logic more aligned with actual production conditions. Simultaneously, the process setting unit 31 can store all process parameters and adjustment records, facilitating subsequent process traceability and optimization.

[0063] Control unit 3 independently modulates the local cooling intensity of each height space region by independently adjusting the flow rate and pressure of the cooling medium flowing into each fluid cooling sleeve. The flow rate and pressure of the cooling medium directly determine its flow velocity inside the fluid cooling sleeve and its heat exchange efficiency with the sleeve wall. The higher the flow rate and pressure, the faster the flow velocity of the cooling medium, the stronger the convective heat exchange effect with the sleeve wall, the higher the heat exchange efficiency of the fluid cooling sleeve, and the greater the applied local cooling intensity; conversely, the lower the flow rate and pressure, the smaller the local cooling intensity.

[0064] To achieve independent regulation of the cooling medium flow rate and pressure, the control unit 3 is electrically connected to the mass flow controller 7 on the cooling medium supply pipeline corresponding to each fluid cooling sleeve. The mass flow controller 7 uses an electric regulating valve, which can steplessly adjust the valve opening according to the electrical signal sent by the control unit 3, with high regulation accuracy and fast response speed. Specifically, each cooling sleeve is connected to the cooling water supply unit 8 through the flow mass controller.

[0065] Each fluid cooling sleeve has an independent set of mass flow controllers 7 on its cooling medium supply pipeline. Specifically, the supply pipelines of the first cooling sleeve 211, the second cooling sleeve 212, and the third cooling sleeve 213 are equipped with corresponding mass flow controllers 7, namely the first mass flow controller 71, the second mass flow controller 72, and the third mass flow controller 73. The control unit 3 can independently control each mass flow controller 7 with electrical signals, thereby realizing independent adjustment of the flow rate and pressure of the cooling medium in each fluid cooling sleeve, and thus realizing independent modulation of the local cooling intensity of each height space domain.

[0066] In this embodiment of the disclosure, reference is made to Figure 6 The control unit 3 is configured to execute iterative adjustment logic, which continuously adjusts the flow rate and pressure of the cooling medium in each fluid cooling sleeve according to the real-time temperature changes during the growth of the single crystal silicon ingot 106. This ensures that the temperature in each height space domain remains within the preset process range, avoiding temperature deviations caused by excessive or insufficient adjustment in a single instance. Simultaneously, it guarantees that the adjustment effect of the cooling intensity is fully realized, improving the accuracy and stability of cooling control. This iterative adjustment can specifically execute steps S610 to S640.

[0067] In step S610, data detection and initial adjustment are performed.

[0068] The control unit 3 receives real-time thermodynamic state data from each spatial domain at different heights. After analyzing and processing the data, it compares the real-time temperature of each region with the temperature range stored in the process setting unit 31. If the real-time temperature of one or more regions deviates from the preset process range, the control unit 3 calculates the corresponding cooling medium flow rate and pressure adjustment parameters according to the pre-stored cooling control logic, sends a control command to the corresponding mass flow controller 7, and completes the adjustment of the cooling medium flow rate and pressure of the fluid cooling sleeve, thereby achieving the first adjustment of the local cooling intensity of the corresponding region.

[0069] In step S620, the adjustment state is maintained and delayed.

[0070] After adjusting the flow rate and pressure of the cooling medium, the control unit 3 controls each mass flow controller 7 to maintain the adjusted valve opening, keeping the flow rate and pressure of the cooling medium stable. Data monitoring continues, but no new adjustments are performed immediately, allowing the device to undergo a preset delay time. This delay time is set to allow the effect of the cooling intensity adjustment to be fully realized. Because the temperature change of the monocrystalline silicon ingot 106 has a certain lag, after the cooling intensity is adjusted, the temperature of the monocrystalline silicon ingot 106 needs a certain amount of time to change accordingly. If the data is detected and adjusted again immediately after adjustment, the temperature lag will cause control errors, resulting in over-adjustment.

[0071] The preset delay time can be flexibly set in the process setting unit 31 according to the process requirements of monocrystalline silicon growth. In this embodiment, according to the kinetic law of BMD nucleation and growth, the delay time is set to 10~20 minutes. This time range can ensure that the effect of cooling intensity adjustment is fully reflected, so that the temperature of monocrystalline silicon ingot 106 changes significantly, and will not cause the temperature to deviate too much from the preset range due to excessive time, thus ensuring the timeliness of control.

[0072] In step S630, a second detection and judgment are performed.

[0073] After a preset delay time, the control unit 3 receives real-time thermodynamic state data of each height space domain again, re-analyzes and processes the data, and compares the real-time temperature of each region with the preset temperature range in the process setting unit 31 for the second time to determine whether the temperature of each region has returned to the preset process range.

[0074] In step S640, iteratively adjust or maintain.

[0075] If, after the second test, the real-time temperature of each area has returned to the preset process range, the control unit 3 controls each mass flow controller 7 to maintain the current valve opening and continues to monitor real-time data. If the temperature deviates again in subsequent tests, the iterative adjustment logic is restarted. If, after the second test, the real-time temperature of any area still deviates from the preset process range, the control unit 3 recalculates and adjusts the parameters based on the new real-time data, performs the next adjustment of the cooling medium flow rate and pressure, and then maintains the adjustment state again and goes through a preset delay time, repeating the above detection and judgment steps until the temperature of that area returns to the preset process range.

[0076] During the iterative adjustment logic loop, the control unit 3 synchronously monitors the pulling growth status of the monocrystalline silicon ingot and the preset process termination conditions. Among them, the criteria for determining the end of the monocrystalline silicon ingot pulling growth include: the pulling height of the seed crystal pulling component 104 reaches the preset target length of the monocrystalline silicon ingot, and the pulling load reaches the preset threshold (indicating that the weight of the monocrystalline silicon ingot meets the standard).

[0077] The preset process termination conditions include abnormal defects during the growth of monocrystalline silicon ingots (such as excessive dislocation density or BMD size exceeding the allowable range) and cooling system failures (such as cooling medium leakage or temperature sensor signal interruption). When the control unit 3 detects any of the above situations, it immediately terminates the iterative adjustment logic, stops the cooling intensity control of the partitioned independent heat exchange component 2, and simultaneously sends a signal to the main control system of the crystal pulling furnace to trigger the monocrystalline silicon ingot pulling termination or emergency handling process, ensuring the safety of the monocrystalline silicon ingot growth process and the controllability of product quality.

[0078] In this embodiment, the execution process of the iterative adjustment logic is completed automatically by the control unit 3 without manual intervention, realizing the automation of the cooling control process, greatly reducing the workload of operators. At the same time, the logic has high control precision, which can control the temperature fluctuation of each height spatial domain within ±5℃, which is far higher than the temperature control precision of traditional crystal pulling furnaces. This effectively ensures the stability of BMD nucleation and growth, and further improves the uniformity of BMD distribution along the 106 axis of the single crystal silicon ingot.

[0079] In this embodiment of the disclosure, reference is made to Figure 7 The thermal state sensing unit includes multiple sets of temperature measurement sensors 9 deployed along the crystal pulling axis to acquire the aforementioned real-time thermodynamic state data.

[0080] Each set of temperature measurement sensors 9 corresponds one-to-one with multiple height spatial domains on the growth path of the single crystal silicon ingot 106. That is, the first set of temperature measurement sensors 91 is deployed in the first region 61, the second set of temperature measurement sensors 92 is deployed in the second region 62, and the third set of temperature measurement sensors 93 is deployed in the third region 63.

[0081] Three sets of sensors are arranged sequentially along the axial direction of the crystal pulling process, with their deployment height consistent with the center height of the corresponding area. This ensures accurate detection of the temperature data of the monocrystalline silicon ingot 106 in the corresponding area. Each set of temperature measurement sensors 9 includes multiple sensor probes, which are evenly arranged circumferentially along the monocrystalline silicon ingot 106. This allows for comprehensive detection of the circumferential temperature distribution of the monocrystalline silicon ingot 106, avoiding detection errors caused by uneven circumferential temperature of the monocrystalline silicon ingot 106 and improving the representativeness and accuracy of the detection data.

[0082] The axial deployment spacing of the temperature measurement sensors 9 matches the stacking spacing of the fluid cooling sleeves in the partitioned independent heat exchange assembly 2. In this embodiment, the deployment height of the first group of temperature measurement sensors 91 is consistent with the center height of the first cooling sleeve 211, the deployment height of the second group of temperature measurement sensors 92 is consistent with the center height of the second cooling sleeve 212, and the deployment height of the third group of temperature measurement sensors 93 is consistent with the center height of the third cooling sleeve 213. This ensures that the detection position of the temperature measurement sensors 9 precisely corresponds to the cooling action position of the corresponding cooling sleeve, and the detection data can directly reflect the cooling effect of the corresponding cooling sleeve, providing direct data support for the precise control of the control unit 3.

[0083] In this embodiment, the axial deployment of the temperature measurement sensor 9 is precisely matched with the division of each height spatial domain. The circumferential multi-probe arrangement can comprehensively reflect the temperature distribution of the single crystal silicon ingot 106. Combined with the data analysis and processing function of the thermal state sensing unit, it can provide accurate and comprehensive temperature data for the control unit 3, ensuring the accuracy and effectiveness of cooling regulation.

[0084] This disclosed partitioned independent cooling scheme achieves a significant improvement in both overall BMD density and axial distribution uniformity by precisely modulating the cooling intensity of different height spatial domains along the growth path of a single-crystal silicon ingot. Compared to traditional single-region cooling, this disclosure increases the BMD density at the crystal head from 1.2 × 10⁻⁶ to 1.2 × 10⁻⁶. 8 pcs / cm 3 Increased to 1.5×10 8 pcs / cm 3 The increase was 25%; the BMD density in the middle of the crystal rod increased from 2.8 × 10⁻⁶. 8 The number of units / cm³ increased to 2.9×10 8 pcs / cm 3 The increase was 3.6%; the BMD density at the tail of the crystal rod increased from 1×10 8 pcs / cm 3 Increased to 1.4×10 8 pcs / cm 3 The increase was 40%; the average BMD density increased from 1.67 × 10⁻⁶. 8 pcs / cm 3 Increased to 1.93×10 8 pcs / cm 3 The increase was 15.6%, providing more effective sites for the removal of metal impurities.

[0085] Meanwhile, the standard deviation of BMD density decreased from 0.92 × 10⁻⁶. 8 pcs / cm 3 Reduced to 0.76×10 8 pcs / cm 3 The reduction of 17.4% effectively reduced the BMD density difference at various positions along the crystal ingot axis, precisely controlled the kinetics of BMD nucleation and growth, avoided the uneven BMD distribution problem caused by traditional single-region cooling, significantly improved the internal quality and process stability of single-crystal silicon ingots, and provided a higher quality raw material basis for the subsequent preparation of semiconductor devices.

[0086] Furthermore, this disclosure also provides a cooling control method for Czochralski single-crystal silicon growth. This method is applied to the partitioned cooling system for Czochralski single-crystal silicon growth in any of the above embodiments, which can realize precise, dynamic, and partitioned control of the cooling process during the growth of single-crystal silicon ingots, effectively improving the internal quality of single-crystal silicon ingots. The overall process of this method is highly compatible with the structural design of the device. The core steps, execution logic, and specific implementation methods of each step are described in detail below.

[0087] The overall process of cooling control for Czochralski single-crystal silicon growth mainly includes two core steps, which are executed cyclically to form a closed-loop control. (Refer to...) Figure 8 Specifically, this includes steps S810 to S820.

[0088] In step S810, real-time thermodynamic state data corresponding to multiple height spatial domains along the growth path of a single-crystal silicon ingot are acquired.

[0089] This step is the basic data acquisition stage of the cooling control method. The main execution unit is the thermal state sensing unit of the device, which is specifically composed of multiple sets of temperature measurement sensors 9 deployed along the crystal pulling axis to complete data detection. During the growth of the single crystal silicon ingot 106, the thermal state sensing unit continuously drives each set of temperature measurement sensors 9 to detect the temperature in the corresponding height spatial domain. The detection frequency is consistent with the data transmission frequency, which can reflect the temperature changes at different height positions of the single crystal silicon ingot 106 in real time.

[0090] In step S820, based on real-time thermodynamic state data, multiple independent heat exchange levels in the partitioned independent heat exchange components are controlled to modulate the local cooling intensity of each height spatial domain.

[0091] After receiving the real-time thermodynamic state data transmitted by the thermal state sensing unit, the control unit 3 analyzes and processes the data according to the preset cooling control logic and generates corresponding control commands. These commands control each heat exchange level 21 of the independent heat exchange components 2 to perform cooling operations, achieving precise modulation of the local cooling intensity in each height spatial domain 6. The specific execution process can be referred to in the description of the partitioned cooling system for Czochralski single-crystal silicon growth, and will not be elaborated upon here.

[0092] The cooling control method of this embodiment relies on the partitioned independent heat exchange components 2 and the thermal state sensing unit of the Czochralski single-crystal silicon growth apparatus to achieve precise and dynamic control of the cooling process of the single-crystal silicon ingot 106. The method has a high degree of automation and requires minimal manual intervention, making it suitable for continuous production in industrial manufacturing. Furthermore, the control logic of this method can be flexibly adjusted according to actual production process requirements. Operators can modify pre-stored parameters through the process setting unit 31 of the control unit 3, adapting the method to the growth of single-crystal silicon ingots of different specifications and with different process requirements, significantly broadening the application scope of the method.

[0093] In this embodiment, multiple height spatial domains 6 along the growth path of the monocrystalline silicon ingot 106 are vertically divided into a first region 61 (lower high-temperature region), a second region 62 (middle medium-temperature region), and a third region 63 (top low-temperature region). The corresponding heat exchange layers 21 of the partitioned independent heat exchange components 2 include a first cooling jacket 211, a second cooling jacket 212, and a third cooling jacket 213. The differentiated cooling intensity control process revolves around these three regions and the three cooling jackets, as described above. Figure 9 The specific steps may include steps S910 to S940.

[0094] In step S910, it is determined whether the temperature of the second region is within the preset micro-defect sensitive nucleation temperature range.

[0095] After receiving the real-time thermodynamic state data transmitted by the thermal state sensing unit, the control unit 3 first extracts the real-time average temperature of the second region 62 and compares it with the preset sensitive nucleation temperature range for bulk microdefects (BMD) (1000℃-800℃) to determine whether the temperature of this region is within a reasonable range. The preset sensitive nucleation temperature range is determined based on the kinetic laws of BMD nucleation and growth. Within this range, BMD nuclei can form rapidly and uniformly, and it is the core temperature range for controlling BMD generation.

[0096] If the real-time temperature of the second region 62 is within the sensitive nucleation temperature range of 1000℃-800℃, it indicates that the current temperature conditions are suitable for the formation of BMD nuclei, and step S920 is executed; if the real-time temperature of the second region 62 is higher than 1000℃ or lower than 800℃, it indicates that the current temperature conditions deviate from the optimal range for BMD nucleation, and step S940 is executed.

[0097] Step S920: Control the heat exchange layer configured in the first region to apply a first local cooling intensity, the heat exchange layer configured in the second region to apply a second local cooling intensity, and the heat exchange layer configured in the third region to apply a third local cooling intensity.

[0098] When the temperature of the second region 62 is within the preset sensitive nucleation temperature range, the control unit 3 controls the second cooling jacket 212 to apply a second local cooling intensity, which is the highest cooling intensity among the three regions. The purpose is to increase the cooling rate of the second region 62 so that BMD nuclei can be formed quickly, in large quantities, and uniformly.

[0099] While applying a first local cooling intensity to the second region 62, the control unit 3 controls the first cooling jacket 211 to apply a first local cooling intensity and the third cooling jacket 213 to apply a third local cooling intensity. The first local cooling intensity and the third local cooling intensity are less than the first local cooling intensity. The first region 61 and the third region 63 are cooled in a gentle cooling manner to avoid thermal stress caused by excessive cooling or affecting the normal growth of BMD.

[0100] In step S930, iterative adjustment logic is executed to maintain the temperature of each region within a preset range.

[0101] After applying differentiated cooling intensities, the control unit 3 executes iterative adjustment logic. After completing one cooling intensity adjustment, it maintains the adjusted state for 10 to 20 minutes until the cooling effect is fully realized. Then, it adjusts the cooling medium flow rate and pressure again based on the real-time thermodynamic state data transmitted by the thermal state sensing unit. This cycle continues, ensuring that the temperature of each area remains within the preset process range.

[0102] In step S940, the cooling intensity of the heat exchange level in the second region is adjusted. The control unit 3 needs to first adjust the cooling intensity of the second cooling jacket 212 to quickly bring the temperature of the second region 62 back to the sensitive nucleation temperature range before proceeding with subsequent steps. For example, if the real-time temperature of the second region 62 is 1050°C, which is 50°C higher than the upper limit of the sensitive nucleation temperature range, the control unit 3 needs to immediately increase the flow rate and pressure of the cooling medium in the second cooling jacket 212 to accelerate the cooling rate of the region. After the temperature drops below 1000°C, the normal differentiated cooling control process will then begin.

[0103] The differentiated cooling intensity control process in this embodiment perfectly matches the temperature characteristics and BMD nucleation growth patterns of each region during the growth of the single-crystal silicon ingot 106. By employing strong cooling for the core region (second region 62) and gentle cooling for the non-core regions (first and third regions), the cooling process is made precise and customized. In actual production applications, this control process can improve the uniformity of axial BMD distribution in the single-crystal silicon ingot 106 by more than 60%, while effectively reducing the generation of defects such as dislocations and harmful large-size BMDs, significantly improving the internal quality of the single-crystal silicon ingot 106, and providing a high-quality raw material foundation for the subsequent fabrication of semiconductor devices.

[0104] Furthermore, this disclosure also provides a crystal rod, which is prepared using a cooling control method for Czochralski single-crystal silicon growth; wherein the average density of bulk microdefects along the entire pulling axis of the crystal rod is between 1.40 × 10⁸ defects / cm³ and 2.0 × 10⁸ defects / cm³; and the axial standard deviation of the bulk microdefect density of the crystal rod is less than or equal to 0.8 × 10⁸ defects / cm³. Simultaneously, the crystal rod can be cut, ground, polished, and cleaned to obtain the wafer of this disclosure, wherein the dislocation density of the wafer is ≤ 5.0 × 10² cm⁻², and the dislocation-free rate is ≥ 99.5%.

[0105] The following description uses specific examples and comparative models for illustration.

[0106] Specifically, the basic experimental parameters are as follows: Growth method: Czochralski method for single-crystal silicon growth.

[0107] Uniform parameters: raw material specifications, thermal field structure, lifting rate, argon pressure in the furnace, and crucible rotation speed are consistent.

[0108] Detection components: a first set of temperature measurement sensors 91, a second set of temperature measurement sensors 92, and a third set of temperature measurement sensors 93.

[0109] Cooling execution components: first cooling sleeve 211, second cooling sleeve 212 and third cooling sleeve 213.

[0110] The cooling parameters of the embodiment are shown in Table 1 (cooling structure: three-zone independent cooling jacket, matching the technical solution of this disclosure): Table 1

[0111] As shown in Table 1, the core of this scheme limits the average density of bulk microdefects (BMD) in the crystal rod to 1.40 × 10⁻⁶. 8 pcs / cm³ ~ 2.0×10 8 BMD density axial standard deviation ≤ 0.8 × 10⁻⁶ units / cm³ 8 Dislocation density per cm³, wafer dislocation density ≤ 5.0 × 10⁻⁶ 2 cm -2 The four core numerical ranges—wafer dislocation-free rate ≥99.5%—are, and combined with experimental data from examples and comparative studies, demonstrate the strong rationality and necessity of these ranges. Regarding the average BMD density range, the lower limit is 1.40 × 10⁻⁶. 8 The number of impurities per cm³ is fundamental to ensuring the ability to control impurities in monocrystalline silicon ingots. The average BMD density of Comparative Example 4 is 1.45 × 10⁻⁶. 8 The number of nuclei per cm³ is approaching the lower limit, indicating insufficient BMD nucleation due to an incorrect cooling strategy. This fails to provide enough effective sites for the removal of metal impurities. Falling below this lower limit will directly reduce the suitability of single-crystal silicon ingots for semiconductor device fabrication; the upper limit is 2.0 × 10⁻⁶. 8 The number per cm³ is to avoid excessive cooling that could lead to abnormal growth of BMD nuclei and the formation of harmful large-size defects. Example 6 uses 2.01 × 10⁻⁶. 8 pcs / cm 3 Slightly exceeding the upper limit, while Examples 1-5 are all between 1.78 and 1.99 × 10⁻⁶. 8 pcs / cm 3 The optimal range was found, and all embodiments fell within this numerical range, achieving a significant improvement in BMD density compared to the comparative example. Within the BMD axial standard deviation range, ≤0.8×10⁻⁶. 8 pcs / cm 3 The axial quality uniformity of monocrystalline silicon ingots is a core indicator. Although Examples 2 and 5 slightly exceed this value (0.83 and 0.81 respectively), these are minor fluctuations acceptable for the process and the dislocation rate still meets the standard. Meanwhile, the standard deviation of all comparative examples is greater than 0.8 × 10⁻⁶. 8 pcs / cm 3 The minimum standard deviation was 0.92, and the maximum was 1.18. In particular, comparative examples 2-4, which did not adopt the correct differentiation strategy, all had standard deviations exceeding 1.0 × 10⁻⁶. 8 pcs / cm 3 The uneven distribution of BMD along the crystal ingot axis directly affects the consistency of the electrical performance of the subsequent wafer.

[0112] Regarding the wafer dislocation density and dislocation-free rate, the wafer dislocation density is less than or equal to 5.0 × 10⁻⁶. 2 cm -2 The requirement of a dislocation-free rate greater than or equal to 99.5% fundamentally guarantees the wafer yield. The dislocation density in all embodiments is less than or equal to 2.5 × 10⁻⁶. 2 cm -2 The dislocation-free rate is greater than or equal to 99.6%, far exceeding the limiting standard, while the dislocation density of all comparative examples is greater than 5.0 × 10⁻⁶. 2 cm -2 The dislocation-free rates were all below 99.5%, with the dislocation density of Comparative Example 4 reaching 2.0 × 10⁻⁶. 3 cm -2 ~3.0×10 3 cm -2 The dislocation-free rate is only 92.5%, indicating excessive thermal stress due to improper cooling strategies, which introduces a large number of dislocation defects. Furthermore, this solution establishes an optimal process range for the cooling medium: first zone flow rate 10~20 L / h, pressure 0.2~0.4 MPa; second zone flow rate 20~30 L / h, pressure 0.3~0.5 MPa; third zone flow rate 15~25 L / h, pressure 0.2~0.4 MPa. All embodiments use this parameter range, demonstrating that this parameter range is the optimal process window for achieving the core numerical indicators.

[0113] Furthermore, as can be seen from Comparative Examples 6 and 7, even when slight overcooling occurs in the first or third region, both still adopt the basic cooling strategy of greater cooling intensity in the middle region than in the two end regions, and the technical effect obtained is still significantly better than that of other comparative examples.

[0114] The average bulk micro-defect density of Comparative Examples 6 and 7 is within the acceptable range. The dislocation density and dislocation-free rate are also significantly better than traditional single-region cooling, three-region equal-intensity cooling, and strong-at-the-ends-weak-in-the-middle cooling schemes. Only the axial standard deviation slightly exceeds the standard due to the slightly higher cooling intensity at both ends. This demonstrates that as long as a core cooling distribution of "strong in the middle and weak at both ends" is adopted, even if the parameters are not within the optimal range, significantly better crystal quality can still be obtained compared to existing technologies, effectively improving bulk micro-defect uniformity, reducing dislocations, and increasing the dislocation-free rate.

[0115] Figure 10 This is a comparison curve showing the distribution of bulk micro-defect density along the crystal pulling axis between traditional cooling methods and the cooling control scheme of this disclosure. Figure 10 As can be seen, the bulk micro-defect density of the cooling control scheme disclosed herein is relatively evenly distributed along the crystal pulling axis, and is maintained at 1.40 × 10⁻⁶. 8 pcs / cm 3 ~2.0×10 8 pcs / cm3 Within the specified range, the axial standard deviation is ≤0.8×10. 8 pcs / cm 3 The novel cooling method exhibits better axial uniformity. In contrast, the bulk microdefect density distribution of traditional cooling methods fluctuates significantly along the axial direction, resulting in poor uniformity. Obvious peaks and valleys of BMD density appear at different axial positions of the crystal rod, with significant differences between the head and tail. This comparative result visually demonstrates the significant improvement effect of the proposed partitioned independent cooling scheme on the axial uniformity of bulk microdefect density, verifying the technical advantages of the "strong in the middle, weak at both ends" differentiated cooling strategy in precisely controlling the BMD nucleation density and its axial distribution.

[0116] This disclosure also provides a computer-readable storage medium storing at least one instruction, which is executed by a processor to implement the Czochralski single-crystal silicon growth apparatus and cooling control method as described in the above embodiments.

[0117] This disclosure also provides a computer program product including computer instructions stored in a computer-readable storage medium; a processor of a computing device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computing device to perform the Czochralski single-crystal silicon growth apparatus and cooling control method described in the above embodiments.

[0118] Those skilled in the art will recognize that the functions described in this disclosure in one or more of the examples above can be implemented using hardware, software, firmware, or any combination thereof. When implemented in software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media include computer storage media and communication media, wherein communication media include any medium that facilitates the transfer of a computer program from one place to another. Storage media can be any available medium accessible to a general-purpose or special-purpose computer.

[0119] It should be noted that the technical solutions described in this disclosure can be combined arbitrarily as long as they do not conflict.

[0120] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the scope of the claims.

Claims

1. A zoned cooling system for Czochralski single-crystal silicon growth, characterized in that, include: The partitioned independent heat exchange components are arranged into multiple independent heat exchange levels along the vertical direction of crystal pulling. Each heat exchange level applies local cooling intensity to multiple height spatial domains. The thermal state sensing unit is deployed along the growth path of the single crystal silicon ingot and is configured to acquire real-time thermodynamic state data corresponding to multiple height spatial domains on the growth path of the silicon ingot during the single crystal growth process. as well as The control unit is communicatively coupled to the partition-independent heat exchange component and the thermal state sensing unit, and is used to receive real-time thermodynamic state data and independently control the partition-independent heat exchange component.

2. The zoned cooling system for Czochralski single-crystal silicon growth according to claim 1, characterized in that, The heat exchange level includes a fluid cooling sleeve, and multiple fluid cooling sleeves are stacked along the vertical direction of crystal pulling to form the partitioned independent heat exchange assembly.

3. The zoned cooling system for Czochralski single-crystal silicon growth according to claim 2, characterized in that, The multiple height spatial domains are sequentially divided into a first region, a second region, and a third region along the vertical direction of crystal pulling; The plurality of fluid cooling sleeves include at least a first cooling sleeve disposed in the first region, a second cooling sleeve disposed in the second region, and a third cooling sleeve disposed in the third region; The control unit is configured to control the local cooling intensity applied by the second cooling jacket to be greater than the local cooling intensity applied by the first cooling jacket and the third cooling jacket.

4. The zoned cooling system for Czochralski single-crystal silicon growth according to claim 2, characterized in that, The fluid cooling sleeve is provided with an inlet and an outlet for the cooling medium at both ends. The fluid cooling sleeve has multiple baffles arranged alternately inside, and the multiple baffles extend inward from the inner walls of opposite sides of the fluid cooling sleeve. The multiple baffles define a tortuous continuous flow channel connecting the inlet and the outlet within the fluid cooling sleeve.

5. The zoned cooling system for Czochralski single-crystal silicon growth according to claim 1, characterized in that, The thermal state sensing unit includes multiple sets of temperature measurement sensors deployed along the crystal pulling axis. Each group of temperature measurement sensors corresponds to multiple height spatial domains.

6. A cooling control method for Czochralski single-crystal silicon growth, characterized in that, The method includes: Acquire real-time thermodynamic state data corresponding to multiple height spatial domains along the growth path of a single-crystal silicon ingot; Based on the real-time thermodynamic state data, multiple independent heat exchange levels in the partitioned independent heat exchange components are controlled to modulate the local cooling intensity of each of the height spatial domains respectively. The multiple independent heat exchange layers are arranged vertically around the growth path of the single-crystal silicon ingot.

7. The cooling control method for Czochralski single-crystal silicon growth according to claim 6, characterized in that, The step of controlling multiple independent heat exchange levels in the partitioned independent heat exchange components based on the real-time thermodynamic state data to modulate the local cooling intensity of each of the height spatial domains includes: The plurality of height spatial domains are sequentially divided into a first region, a second region, and a third region along the vertical direction; When the temperature of the second region is determined to be within the preset bulk micro-defect sensitive nucleation temperature range based on the real-time thermodynamic state data, the heat exchange layer configured in the first region is controlled to apply a first local cooling intensity, the heat exchange layer in the second region is controlled to apply a second local cooling intensity, and the heat exchange layer in the third region is controlled to apply a third local cooling intensity. The second local cooling intensity is greater than the first local cooling intensity and the third local cooling intensity.

8. The cooling control method for Czochralski single-crystal silicon growth according to claim 6, characterized in that, The method further includes: After maintaining the adjusted working state and waiting for a preset delay time, real-time thermodynamic state data is acquired again to trigger the next control adjustment cycle until the single crystal silicon ingot pulling growth is detected to be completed or the preset process termination condition is met.

9. A crystal rod, characterized in that, The crystal rod is prepared using the cooling control method for Czochralski single crystal silicon growth as described in any one of claims 6 to 8; The average density of bulk microdefects along the entire pulling axis of the crystal rod is 1.40 × 10⁻⁶. 8 pcs / cm 3 ~2.0×10 8 pcs / cm 3 Between; and the axial standard deviation of the bulk microdefect density of the crystal rod is less than or equal to 0.8 × 10⁻⁶. 8 pcs / cm 3 .

10. A wafer, characterized in that, The wafer is obtained by cutting, grinding, polishing and cleaning the crystal rod as described in claim 9; Wherein, the dislocation density of the wafer is ≤5.0×10 2 cm -2 And the dislocation rate is ≥99.5%.

Images