Crystal pulling methods, apparatus, systems, and single crystal silicon

Abstract

The present disclosure provides a crystal pulling method, device, system and single crystal silicon. The crystal pulling method comprises: in a temperature stabilization stage before the start of the crystal pulling after the completion of the material melting, performing the following power control steps: determining an initial optimized power value; reducing the heating power from the initial optimized power value to a temperature stabilization starting power value; increasing the heating power from the temperature stabilization starting power value to a temperature stabilization target power value higher than the temperature stabilization starting power value in a stepwise temperature recovery manner, so as to stabilize the silicon melt surface state. The crystal pulling method, device, system and single crystal silicon provided by the present disclosure can effectively reduce the pinhole defect density in the single crystal silicon.

Description

Technical Field

[0001] This disclosure relates to the field of semiconductor manufacturing technology, and in particular to a crystal pulling method, apparatus, system, and single-crystal silicon. Background Technology

[0002] The Czochralski method is the mainstream method for preparing semiconductor-grade single-crystal silicon. In the crystal pulling process, polycrystalline silicon raw material is first heated and melted in a quartz crucible, followed by a critical temperature stabilization stage to ensure that the silicon melt reaches sufficient thermal equilibrium and uniform composition. Finally, through steps such as crystal pulling, shoulder formation, and constant diameter growth, a high-quality single-crystal silicon rod is obtained.

[0003] However, in the silicon wafers produced from the single-crystal silicon rods obtained through crystal pulling, tiny pinhole defects are sometimes observed in the device region. These defects can lead to serious problems such as short circuits and leakage, significantly reducing the final yield and long-term reliability of integrated circuits. The source of pinhole defects is tiny bubbles trapped during crystal growth, or heterogeneous particles such as SiO2 (silicon dioxide) and SiC (silicon carbide). These tiny bubbles or particles are exposed in subsequent processes, forming holes.

[0004] In traditional Czochralski processes, the control of heating power during the temperature stabilization phase often relies on the operator's personal experience. A common practice is to set the power to a fixed, empirical value to maintain the molten silicon in a liquid state. However, this approach has the following drawbacks:

[0005] If the power setting during the stabilization stage is too high, the melt temperature will be too high, which will intensify thermal convection within the melt. This not only disrupts the thermal field stability but also accelerates the etching of the quartz crucible wall, promotes silicon volatilization, and leads to an increase in oxygen content in the melt, generating a large number of SiO2 and SiC particles. Simultaneously, the high-temperature environment may also cause the protective gas to precipitate and form bubbles, directly increasing the concentration of heterogeneous particles and bubbles. Conversely, if the power setting during the stabilization stage is too low, it may result in uneven melt temperature distribution, leading to incompletely melted silicon or the precipitation of high-melting-point impurity particles.

[0006] Therefore, excessively high power during the temperature stabilization stage will introduce a large number of bubbles and reaction particles; while insufficient power will lead to unmelted residue and uneven composition. Regardless of whether the power is too high or too low during the temperature stabilization stage, it will introduce the source of pinhole defects. Summary of the Invention

[0007] To address at least one technical problem in the related art, embodiments of this disclosure provide a crystal pulling method, apparatus, system, and monocrystalline silicon that can effectively reduce the density of pinhole defects in monocrystalline silicon.

[0008] The technical solutions provided in this disclosure are as follows:

[0009] In a first aspect, embodiments of this disclosure provide a crystal pulling method, comprising:

[0010] Material preparation stage: The polycrystalline silicon in the crucible is heated until it is completely melted to form a silicon melt;

[0011] Temperature stabilization phase: During the temperature stabilization phase, the following power control steps are performed:

[0012] Based on the historical process database, the initial optimized power value is determined;

[0013] Reduce the heating power from the initial optimized power value to a stable initial power value;

[0014] The heating power is increased from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, incremental temperature recovery manner, so as to stabilize the state of the silica molten liquid surface.

[0015] Crystal growth stage: crystal seeding and single crystal growth are carried out.

[0016] For example, determining the initial optimized power value based on the historical process database includes:

[0017] Obtain the instantaneous power P_melt of complete material processing under the current process conditions;

[0018] The proportionality coefficient k is determined based on the historical process database, and the initial optimized power value P_opt is calculated according to the formula P_opt = k × P_melt, wherein the value of the proportionality coefficient k ranges from 0.85 to 0.95.

[0019] For example, the instantaneous power P_melt of the fully oxidized material is 120~150kW;

[0020] The initial optimized power value P_opt is 108~135kW, the initial power value for stabilizing the temperature is 10~40kW, and the target power value for stabilizing the temperature is 110kW~160kW;

[0021] The stepped heating method includes 5 to 10 power steps, with a total increase in heating power of 100 kW to 120 kW and a total duration of 1 to 3 hours;

[0022] In the stepped temperature recovery method, the duration of each power step is 5 to 10 minutes.

[0023] For example, the step of increasing the heating power from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, incremental temperature recovery manner, so as to stabilize the state of the molten silica surface, specifically includes:

[0024] The state parameters of the silicon melt surface are monitored in real time, and the heating power is dynamically adjusted based on the monitoring results.

[0025] For example, the real-time monitoring of the silicon melt surface state parameters and the dynamic adjustment of the heating power based on the monitoring results specifically include:

[0026] When the fluctuation parameter of the liquid level is detected to exceed the preset threshold, the heating power is gradually adjusted within the power range corresponding to the current power step according to the preset adjustment ratio.

[0027] For example, after the heating power is increased from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step temperature recovery manner to stabilize the state of the silica molten liquid, the heating power is maintained to operate stably for at least 60 minutes, and the fluctuation range of the heating power is ±0.3%.

[0028] Secondly, this disclosure provides a single-crystal silicon, which is prepared by the crystal pulling method described above, wherein the pinhole defect size caused by bubbles or heterogeneous particles in the single-crystal silicon is 30~50μm.

[0029] Thirdly, embodiments of this disclosure provide a control device, including:

[0030] At least one processor;

[0031] At least one memory, on which a computer program is stored;

[0032] When the computer program is executed by the processor, the control device performs the temperature control method as described above.

[0033] Fourthly, embodiments of this disclosure provide a crystal pulling system, including:

[0034] A heater is used to heat the silicon material inside the crucible;

[0035] A power controller, electrically connected to the heater, is used to set and adjust the heating power; and

[0036] The control device described above is communicatively connected to the power controller and is used to send control commands to the power controller.

[0037] The embodiments disclosed herein have at least the following beneficial effects:

[0038] In the above scheme, the heating power during the stabilization stage after material preparation and before crystal pulling is optimized and controlled. After determining an initial optimized power value, the power is reduced from this initial optimized power value to a stable initial power value. This avoids the overheating of the silicon melt caused by empirically setting excessively high power in traditional processes, effectively suppressing violent convection in the silicon melt, slowing down the etching rate of the crucible and silicon volatilization, reducing the oxygen content, SiO2, and SiC heteroparticle generation in the silicon melt, and also reducing the risk of protective gas precipitation and bubble formation. Furthermore, by using a stepped temperature recovery method, the temperature is gradually increased from the stable initial power value to the stable target power value, which helps maintain a stable silicon melt surface and ensures that the silicon melt can smoothly reach the ideal thermal equilibrium state, avoiding defects such as thermal field fluctuations and melt convection turbulence caused by insufficient power or excessively rapid heating. Therefore, this embodiment of the present disclosure, by optimizing and controlling the heating power during the stabilization stage, can effectively suppress the generation of bubbles and heteroparticles and reduce the density of pinhole defects in single-crystal silicon. Attached Figure Description

[0039] Figure 1 A schematic diagram showing the structure of a single crystal furnace;

[0040] Figure 2 A flowchart illustrating the crystal pulling method according to an embodiment of this disclosure;

[0041] Figure 3 This diagram illustrates the heating power in the crystal pulling method according to an embodiment of the present disclosure.

[0042] Figure 4 This is a comparison chart showing the incidence of pinhole defects in single-crystal silicon compared to the crystal pulling method of this disclosure and the conventional crystal pulling direction;

[0043] Figure 5 This is a comparison chart showing the COP defect incidence rate of monocrystalline silicon compared to the crystal pulling method of this disclosure and the conventional crystal pulling direction. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0045] Unless otherwise defined, the technical or scientific terms used in this disclosure shall have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms “first,” “second,” and similar terms used in this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an,” “a,” or “the,” and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms “including,” “comprising,” or “containing,” and similar terms mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. The terms “connected,” “linked,” or similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms “upper,” “lower,” “left,” and “right,” etc., are used only to indicate relative positional relationships, and these relative positional relationships may change accordingly when the absolute position of the described objects changes.

[0046] Before providing a detailed description of the crystal pulling method, apparatus, system, and single-crystal silicon provided in the embodiments of this disclosure, it is necessary to explain the related technologies as follows:

[0047] The single-crystal silicon rods obtained by crystal pulling contain tiny pinhole defects. The source of these defects is tiny bubbles or heterogeneous particles such as SiO2 (silicon dioxide) and SiC (silicon carbide) that occurred during the crystal growth process. These tiny bubbles or particles are exposed in subsequent processes, forming pores.

[0048] The main causes of pinhole defects are twofold: First, gas issues during crystal growth. In Czochralski crystal growth, improper furnace pressure control or inappropriate argon flow rate can prevent timely venting of gases from the melt, leading to the formation of pores in the crystal and ultimately pinhole defects. Second, wafer surface contamination and particles. Particulate contaminants on the wafer surface can cause blockage or shielding effects in subsequent processes (such as thin film deposition), resulting in pinholes and micropores.

[0049] The inventors of this application have discovered that in traditional Czochralski processes, the control of heating power during the temperature stabilization phase often relies on the operator's personal experience. A common practice is to set the power to a fixed value determined by the operator's experience, the purpose of which is to maintain the molten silicon in a stable liquid state.

[0050] However, this method can lead to pinhole defects for the following reasons: If the heating power during the stabilization stage is set too high, the melt temperature will be too high, which will intensify thermal convection within the melt. This not only disrupts the thermal field stability but also accelerates the etching of the crucible wall, promotes silicon volatilization, and increases the oxygen content in the melt, generating a large number of SiO2 and SiC particles. Simultaneously, the high-temperature environment may also cause the protective gas to precipitate and form bubbles, directly increasing the concentration of heterogeneous particles and bubbles. Conversely, if the power during the stabilization stage is set too low, it may result in uneven melt temperature distribution, leading to incompletely melted silicon or the precipitation of high-melting-point impurity particles.

[0051] Therefore, excessively high heating power during the temperature stabilization stage will introduce a large number of bubbles and reaction particles; while insufficient heating power will lead to unmelted residue and uneven composition. Regardless of whether the heating power during the temperature stabilization stage is too high or too low, it will introduce defect sources (such as gas or impurity particles) that can form pinhole defects.

[0052] Based on this, in order to solve the problem of pinhole defects, this application provides a crystal pulling method, apparatus, system and single crystal silicon. By optimizing and controlling the heating power during the temperature stabilization stage, pinhole defects can be effectively reduced.

[0053] Figure 1 The diagram shown is a schematic representation of the Czochralski single crystal furnace used in the crystal pulling method provided in this embodiment of the disclosure. Figure 1 As shown, a Czochralski single crystal furnace may include a furnace body 100, a crucible 200 disposed within the furnace body 100, and a heater 300 for heating the crucible 200. The entire Czochralski crystal pulling process may include a material preparation stage, a temperature stabilization stage, and a crystal growth stage. In the material preparation stage, the heater is activated to heat the polycrystalline silicon in the crucible to complete melting according to a preset program, forming a silicon melt. The temperature stabilization stage is performed on the silicon melt after the material preparation stage. The crystal growth stage is performed after the temperature stabilization stage, involving crystal introduction and single crystal growth. For example, the crystal growth stage may include a crystal introduction stage, a necking stage, a shoulder formation stage, a constant diameter growth stage, a finishing stage, and a cooling stage.

[0054] The crystal pulling method provided in this disclosure mainly focuses on optimizing and controlling the heating power during the temperature stabilization stage to effectively eliminate pinhole defects. For example... Figure 2 and Figure 3 As shown, the crystal pulling method provided in this embodiment includes:

[0055] Material preparation stage: The polycrystalline silicon in the crucible is heated until it is completely melted to form a silicon melt;

[0056] Temperature stabilization phase: During the temperature stabilization phase, the following power control steps are performed: Step S01: Determine the initial optimized power value;

[0057] Step S02: Reduce the heating power from the initial optimized power value to a stable initial power value;

[0058] Step S03: Increase the heating power from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, incremental temperature recovery manner, so as to stabilize the state of the silica molten liquid surface.

[0059] Crystal growth stage: crystal seeding and single crystal growth are carried out.

[0060] In the above scheme, during the temperature stabilization stage, the initial optimized power value is determined in step S01, and in step S02, the temperature is rapidly reduced from the initial optimized power value to a stable initial power value. This can quickly quell the violent flow that may occur at the end of the melting process, allowing the silicon melt to enter a controllable initial state, laying the foundation for subsequent staged temperature recovery. Furthermore, during crystal growth, rapid temperature changes generate enormous thermal stress, which is a major cause of various crystal defects. In this application, in step S03, the staged temperature recovery method refers to not directly increasing the heating power to the maximum in the initial stage of crystal growth (e.g., the crystal pulling stage), but rather increasing the temperature gradually and progressively. Each small increase in heating power reduces thermal convection and disturbance at the melt surface, thereby achieving a stable melt surface state.

[0061] Specifically, stepped temperature recovery improves crystal quality and indirectly affects pinhole formation in the following ways: First, this stepped temperature recovery method reduces thermal shock and thermal stress. During crystal growth, rapid temperature changes generate enormous thermal stress, a major cause of various crystal defects. Stepped heating, through a gradual heating process, helps reduce thermal shock and thermal stress, laying the foundation for high-quality crystal preparation. Second, it stabilizes the melt and reduces defects. A gradual heating process helps maintain a stable silicon melt surface, which is beneficial for the stable execution of subsequent crystal pulling operations. A stable growth start helps reduce defects caused by thermal field fluctuations and melt convection turbulence at the defect source, suppressing the generation of gas-related defects (including pinholes).

[0062] Therefore, in the crystal pulling method provided in this disclosure, the heating power during the stabilization stage after the material preparation and before crystal pulling is optimized and controlled. By determining an initial optimized power value and then decreasing it to a stable initial power value, the overheating of the silicon melt caused by empirically setting excessively high power, as in traditional processes, is avoided. This effectively suppresses violent convection in the silicon melt, slows down the etching rate of the crucible and silicon volatilization, reduces the oxygen content, SiO2, and SiC heteroparticles in the silicon melt, and also reduces the risk of protective gas precipitation and bubble formation. Furthermore, by using a stepped temperature recovery method, the temperature is gradually increased from the stable initial power value to the stable target power value, which helps maintain the stability of the silicon melt surface and ensures that the silicon melt can smoothly reach the ideal thermal equilibrium state. This avoids defects such as thermal field fluctuations and melt convection turbulence caused by insufficient power or excessively rapid heating, thereby effectively suppressing the generation of bubbles and heteroparticles and reducing the pinhole defect density in single-crystal silicon.

[0063] It should be noted that the inventors of this application have discovered through research that, in the crystal pulling process, the effect of heating power on pinhole defects is not a simple linear relationship, but rather there exists an optimal power window.

[0064] Specifically, the heating power can be divided into three regions: a low-power region, an optimized power window region, and a high-power region. In the low-power region, incomplete cleaning or treatment of the wafer surface leaves residual contaminants, oxides, or particles that can become defect sources in subsequent processes, resulting in a high pinhole defect density. In the optimized power window region, the process energy is sufficient to effectively remove surface contaminants or achieve uniform material modification without causing severe surface damage or redeposition, minimizing the measured pinhole defect density. This optimized power window region represents the target operating window for the process. In the high-power region, excessive energy can lead to surface physical damage or induce unnecessary chemical reactions such as redeposition. These negative effects can introduce new defects or deteriorate the surface condition, causing the pinhole defect density to rise again. Therefore, accurately and stably controlling the process parameters within this optimized power window region is crucial for reducing pinhole defects.

[0065] The crystal pulling method provided by this invention is based on the above analysis, achieving precise control and stable maintenance of the optimized power window region. In this embodiment, the stable temperature target power value set in step S03 and the determination of the power of each step are intended to ensure that the process ultimately operates stably within the optimized power window, thereby theoretically ensuring the lowest defect density. The more steps in the stepped temperature recovery method, the more of the above-mentioned optimal range will be. More steps can help reduce or eliminate pinhole defects and stabilize the process.

[0066] In some exemplary embodiments, in step S01 above, the initial optimized power value is set based on the historical process database.

[0067] In the above scheme, when the system is put into use, a large amount of historical crystal pulling process data can be collected and accumulated. This historical process data includes successful process data that can produce high-quality, low-defect crystal rods. Within the same single crystal furnace, when process conditions such as thermal field parameters and key initial conditions such as feed rate and polycrystalline silicon type are the same, the thermodynamic behavior is highly repeatable. Therefore, successful process data that can produce high-quality, low-defect crystal rods under the current process conditions can be determined from the historical process database to determine the initial optimized power value. In this way, compared with the traditional technology that relies on fixed experience values ​​set by operators, obtaining the initial optimized power value from the historical process database can convert individual experience into a standardized data model, eliminate the experience differences between different operators, and ensure that the starting conditions of each temperature stabilization process are consistent, thereby improving the repeatability and batch stability of single crystal silicon rod quality.

[0068] In some exemplary embodiments, setting the initial optimized power value based on a historical process database specifically includes: obtaining the instantaneous power P_melt at full material reaction under the current process conditions; determining a proportionality coefficient k based on the historical process database, and calculating the initial optimized power value P_opt according to the formula P_opt = k × P_melt, wherein the proportionality coefficient k ranges from 0.85 to 0.95. For example, the instantaneous power P_melt at full material reaction can be 120~150kW; the proportionality coefficient k can be 0.9, and the calculated initial optimized power value P_opt can be 108~135kW. The fluctuation range of the set initial optimized power value P_opt can be controlled within ±0.5%.

[0069] In the above scheme, at the end of the current material preparation stage, the instantaneous power P_melt of complete material preparation can be automatically recorded and obtained. The control unit can query the corresponding optimization ratio coefficient K from the historical process database based on the current process conditions, and calculate the initial optimized power value P_opt according to the above formula. This initial optimized power value is used as the starting point for the cooling process in the stabilization stage. Since the initial optimized power value P_opt is calculated based on the instantaneous power P_melt of complete material preparation, it is not a fixed empirical value, giving the crystal pulling method a certain degree of adaptive compensation function.

[0070] Furthermore, in some exemplary embodiments, the stepped temperature recovery method includes 5 to 10 power steps, with a total increase in heating power of 100 kW to 120 kW and a total duration of 1 to 3 hours. In the stepped temperature recovery method, the duration of each power step is 5 to 10 minutes. For example, the initial power value for temperature stabilization can be 10 to 40 kW, and the target power value for temperature stabilization is 110 kW to 160 kW.

[0071] In the above scheme, the number of steps and the heating amplitude and duration of each power step are defined based on the process running time. First, the total power increase for step-by-step reheating is determined to be 100~120KW (e.g., from 10~40KW to 110KW~160KW). The time during the step-by-step reheating process is mainly affected by factors such as production capacity, and can be controlled within 1~3 hours. Therefore, the heating amplitude and the number of steps can be controlled based on the heating time. If there are too many steps, the reheating process will require a longer time, resulting in wasted production capacity. If there are too few steps, the heating process will be too fast, leading to poor stability of the silicon melt surface and more crystal defects, such as pinholes, which cannot remain in the optimal process window area and cannot expel gas in time.

[0072] For example, as a specific numerical embodiment, in the production of 300mm monocrystalline silicon with a feed rate of 400~425kg, the following parameters can be used: set 8 power steps, starting from the initial power value of 35 kW for temperature stabilization, increasing by approximately 15 kW per step, gradually increasing to the target power value of 155 kW for temperature stabilization, with a total temperature increase of 120 kW. Each power step is maintained for 8 minutes, and the total duration of temperature stabilization is approximately 64 minutes.

[0073] Monocrystalline silicon prepared using the traditional crystal pulling method was used as a control example, while monocrystalline silicon prepared using the crystal pulling method provided in this disclosure was used as a test example. The pinhole occurrence rate of monocrystalline silicon obtained in the control example and the test example was statistically analyzed.

[0074] Comparison Example

[0075] The traditional crystal pulling method is used to prepare single-crystal silicon, and the process is as follows:

[0076] First, 400-425 kg of polycrystalline silicon is put into a single crystal furnace, and a certain amount of dopant is added. Then, a vacuum is drawn and the temperature is raised for material processing.

[0077] Then, after the material is processed, a temperature stabilization process is carried out. During the temperature stabilization stage, the heating power is increased directly from 10~40KW to 1010~160KW in a linear manner, and the heating time is within 1 hour.

[0078] Then, after the temperature stabilization stage ends, the crystal growth stage begins, resulting in single-crystal silicon.

[0079] Example

[0080] The crystal pulling method disclosed herein is used to prepare single-crystal silicon, and the process is as follows:

[0081] First, 400-425 kg of polycrystalline silicon is put into a single crystal furnace, and a certain amount of dopant is added. Then, a vacuum is drawn and the temperature is raised for material processing.

[0082] Then, after the material is melted, the temperature stabilization stage begins. In this stage, an initial optimized power value is first determined based on historical success data. Then, the heating power is reduced from the initial optimized power value to a temperature stabilization starting power value. Then, a stepped temperature recovery method is used to increase the heating power from the temperature stabilization starting power value to a temperature stabilization target power value higher than the initial temperature stabilization power value, so as to stabilize the state of the silica molten liquid.

[0083] Then, the crystal growth stage begins, involving crystal introduction and single crystal growth.

[0084] In Example 1, the initial optimized power value during the temperature stabilization stage is 108~135 KW, the initial power value for temperature stabilization is 10~40 KW, and the target power value for temperature stabilization is 110~160 KW. In the stepped temperature recovery method, two power steps are set. First, the heating power is increased from 10~40 KW to 60~80 KW in a linear heating method for 10 minutes. After maintaining 60~80 KW for 20 minutes, the power is increased from 60~80 KW to 110~160 KW for 10 minutes and maintained for 20 minutes.

[0085] In Example 2, the initial optimized power value during the temperature stabilization stage is 108~135 KW, the starting power value for temperature stabilization is 10~40 KW, and the target power value for temperature stabilization is 110~160 KW. In the stepped temperature recovery method, four power steps are set. Starting from the starting power value of 10~40 KW, the power is gradually increased by 25-30 kW per step until the target power value of 110~160 kW is reached. The heating time for each power step is 2 minutes, and the temperature of each step is maintained for 8 minutes.

[0086] In Example 3, the initial optimized power value during the temperature stabilization stage is 108~135 KW, the starting power value for temperature stabilization is 10~40 KW, and the target power value for temperature stabilization is 110~160 KW. In the stepped temperature recovery method, eight power steps are set. Starting from the starting power value of 10~40 KW, the power is gradually increased to the target power value of 110~160 KW in increments of 12.5-15 kW per step. The heating time for each power step is 1 minute, and the temperature of each step is maintained for 6 minutes.

[0087] Figure 4 This diagram shows a comparison of the single-crystal silicon pinhole defect rate between the crystal pulling method of this disclosure and the conventional crystal pulling method. In this diagram, Case 1 represents the single-crystal silicon pinhole defect rate in the control example, Case 2 represents the single-crystal silicon pinhole defect rate in Example 1, Case 3 represents the single-crystal silicon pinhole defect rate in Example 2, and Case 4 represents the single-crystal silicon pinhole defect rate in Example 3. Figure 5 This is a comparison chart showing the COP defect rate of monocrystalline silicon compared to the crystal pulling method of the present disclosure and the conventional crystal pulling method, where a represents the COP defect rate of monocrystalline silicon in the control example and b represents the COP defect rate of monocrystalline silicon in Examples 1-3.

[0088] In the detection section, the pinhole defect rate in silicon wafers obtained from the monocrystalline silicon of the above-mentioned comparative examples and Examples 1-3 was collected, and the pinhole occurrence rate and COP defect rate in the monocrystalline silicon were statistically analyzed. Figure 4 As shown, the pinhole incidence rate of monocrystalline silicon in the control example was 3.2%. The pinhole incidence rate of monocrystalline silicon in Example 1 was 2%. The pinhole incidence rate of monocrystalline silicon in Example 2 was 1%. The pinhole incidence rate of monocrystalline silicon in Example 3 was less than 0.3%. Figure 5 As shown, the COP occurrence rate of monocrystalline silicon in the control example was 0.3%, while the COP occurrence rate of monocrystalline silicon in Examples 1-3 was less than 0.03%.

[0089] Therefore, it can be seen that the pinhole defect rate of the monocrystalline silicon rod produced by the method of this embodiment can be reduced by more than 25% compared with the traditional fixed power temperature stabilization process, and the COP rate can be reduced by more than 29% compared with the traditional fixed power temperature stabilization process.

[0090] Furthermore, in some exemplary embodiments, the step of increasing the heating power from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, incremental temperature recovery manner to stabilize the state of the silicon melt surface specifically includes: real-time monitoring of the silicon melt surface state parameters and dynamic adjustment of the heating power based on the monitoring results.

[0091] In the above scheme, the entire stepped reheating process is not executed in an open-loop program. Instead, by monitoring the state parameters of the silicon melt surface in real time, a correction signal can be calculated based on the silicon melt surface parameters to dynamically correct the heating power and realize a closed-loop control process.

[0092] For example, the real-time monitoring of the state parameters of the silicon melt surface and the dynamic adjustment of the heating power based on the monitoring results specifically include: when the fluctuation parameters of the melt surface are detected to exceed a preset threshold, the heating power is gradually increased or decreased within the power range corresponding to the current power step according to a preset adjustment ratio.

[0093] Specifically, the adjustment amount can be calculated based on the magnitude of the difference exceeding the preset threshold, according to a preset adjustment ratio algorithm. The adjustment ratio can be optimized to ensure that each adjustment is gentle and avoids liquid surface oscillation caused by excessive adjustment. Within the upper and lower limits of the current power step setting, for example, if the current step target is 80kW, the allowable adjustment range is 78-82kW, gradually increasing or decreasing the power to the new set value at a small rate.

[0094] For example, the crystal pulling apparatus may include a monitoring unit 400, which can acquire image data of the liquid surface in real time and feed it back to the control unit. The control unit receives and analyzes the image data from the monitoring unit to obtain liquid surface state parameters, and generates adjustment commands based on the liquid surface state parameters to adjust the heating power.

[0095] For example, the monitoring unit can utilize image recognition to monitor the state of the liquid surface. Specifically, it can employ a combination of infrared imaging technology and computer vision / artificial intelligence algorithms. The raw data acquired from the infrared thermal imager contains a matrix of temperature values ​​for each pixel. First, the raw data is converted into a pseudo-color image for display, while retaining the original temperature matrix for analysis. One or more monitoring areas are delineated in the image, for example, monitoring only the liquid surface at the center of the reactor, ignoring the edges of the reactor wall. Analyzing the rate of temperature change within the monitoring area can be used to detect local overheating or uneven cooling. If the average temperature > T_max or the highest temperature > T_critical, an alarm is triggered. If the temperature gradient suddenly increases, it may indicate a violent chemical reaction or local boiling. Monitoring the temperature change curve over time can also provide early warning if the heating rate is abnormal. Of course, it is understood that the monitoring unit is not limited to this.

[0096] Furthermore, in some exemplary embodiments, after the heating power is increased from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, incremental temperature recovery manner to stabilize the state of the silica molten liquid, the heating power is maintained at a stable operation for at least 60 minutes, and the fluctuation range of the heating power is ±0.3%.

[0097] In the above scheme, after the heating power is increased to the target stable temperature power value, the heating power is maintained at that target stable temperature power value for at least 60 minutes. During this period, the actual fluctuation range of the heating power is strictly controlled within ±0.3%. In this way, under the stable thermal environment of heating power for at least 60 minutes, defect sources in the silicon melt, such as gases and impurity particles, have sufficient time to be effectively eliminated, and the thermal field is homogenized, providing a better environment for subsequent crystal growth and ensuring the quality of single crystal silicon.

[0098] Furthermore, embodiments of this disclosure provide a control device, including:

[0099] At least one processor;

[0100] At least one memory, on which a computer program is stored;

[0101] When the computer program is executed by the processor, the control device performs the temperature control method as described above.

[0102] Furthermore, embodiments of this disclosure provide a crystal pulling system, including:

[0103] A heater is used to heat the silicon material inside the crucible;

[0104] A power controller, electrically connected to the heater, is used to set and adjust the heating power; and

[0105] The control device of this disclosure embodiment is communicatively connected to the power controller and is used to send control commands to the power controller.

[0106] Furthermore, this disclosure provides a monocrystalline silicon prepared by the crystal pulling method described above, wherein the pinhole defects caused by bubbles or foreign particles in the monocrystalline silicon are within the range of 30-50 μm in size. Thus, the monocrystalline silicon prepared by the crystal pulling method provided in this application does not contain pinhole defects larger than 50 μm. In other words, the density of pinhole defects larger than 50 μm is 0 per square centimeter. Clearly, the control device and crystal pulling system provided in this disclosure also possess the beneficial effects of the crystal pulling method of this disclosure, and will not be elaborated further here.

[0107] The following points need to be explained:

[0108] (1) The accompanying drawings of the embodiments of this disclosure only involve the structures involved in the embodiments of this disclosure. Other structures can be referred to the general design.

[0109] (2) For clarity, the thickness of layers or regions is enlarged or reduced in the drawings used to describe embodiments of the present disclosure, i.e., these drawings are not drawn to scale. It will be understood that when an element such as a layer, film, region or substrate is referred to as being “above” or “below” another element, the element may be “directly” located “above” or “below” the other element or there may be intermediate elements.

[0110] (3) Where there is no conflict, the embodiments of this disclosure and the features in the embodiments can be combined with each other to obtain new embodiments.

[0111] The above are merely specific embodiments of this disclosure, but the scope of protection of this disclosure is not limited thereto. The scope of protection of this disclosure shall be determined by the scope of the claims.

Claims

1. A crystal pulling method, characterized in that, include: Material preparation stage: The polycrystalline silicon in the crucible is heated until it is completely melted to form a silicon melt; Temperature stabilization phase: During the temperature stabilization phase, the following power control steps are performed: Based on the historical process database, the initial optimized power value is determined; Reduce the heating power from the initial optimized power value to a stable initial power value; The heating power is increased from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, incremental temperature recovery manner, so as to stabilize the state of the silica molten liquid surface. Crystal growth stage: crystal seeding and single crystal growth are carried out.

2. The crystal pulling method according to claim 1, characterized in that, The process of determining the initial optimized power value based on the historical process database includes: Obtain the instantaneous power P_melt of complete material processing under the current process conditions; The proportionality coefficient k is determined based on the historical process database, and the initial optimized power value P_opt is calculated according to the formula P_opt = k × P_melt, wherein the value of the proportionality coefficient k ranges from 0.85 to 0.

95.

3. The crystal pulling method according to claim 2, characterized in that, The instantaneous power P_melt of the fully oxidized material is 120~150kW; The initial optimized power value P_opt is 108~135kW, the initial power value for stabilizing the temperature is 10~40kW, and the target power value for stabilizing the temperature is 110kW~160kW; The stepped reheating method includes 5 to 10 power steps, with a total increase in heating power of 100 kW to 120 kW and a total duration of 1 to 3 hours; In the stepped temperature recovery method, the duration of each power step is 5 to 10 minutes.

4. The crystal pulling method according to claim 1, characterized in that, The step of increasing the heating power from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, stepped temperature recovery manner to stabilize the state of the silica molten surface specifically includes: The state parameters of the silicon melt surface are monitored in real time, and the heating power is dynamically adjusted based on the monitoring results.

5. The crystal pulling method according to claim 4, characterized in that, The real-time monitoring of the silicon melt surface state parameters and the dynamic adjustment of the heating power based on the monitoring results specifically include: When the fluctuation parameter of the liquid level is detected to exceed the preset threshold, the heating power is gradually adjusted within the power range corresponding to the current power step according to the preset adjustment ratio.

6. The crystal pulling method according to claim 1, characterized in that, After the heating power is increased from the initial stable temperature power value to a target stable temperature power value higher than the initial stable temperature power value in a step-by-step, gradual temperature recovery manner to stabilize the state of the silica molten liquid, the heating power is maintained at a stable operation for at least 60 minutes, and the fluctuation range of the heating power is ±0.3%.

7. A type of monocrystalline silicon, characterized in that, The single-crystal silicon is prepared by the crystal pulling method according to any one of claims 1 to 6, and the pinhole defect size caused by bubbles or heterogeneous particles in the single-crystal silicon is 30~50μm.

8. A control device, characterized in that, include: At least one processor; At least one memory, on which a computer program is stored; When the computer program is executed by the processor, the control device performs the crystal pulling method as described in any one of claims 1 to 6.

9. A crystal pulling system, characterized in that, include: A heater is used to heat the silicon material inside the crucible; A power controller, electrically connected to the heater, is used to set and adjust the heating power; and The control device as described in claim 8 is communicatively connected to the power controller and is used to send control commands to the power controller.

Images