System and method for silicon production using a horizontal magnetic field

Abstract

To provide a system for producing a silicon ingot by the horizontal magnetic field Czochralski method.SOLUTION: A system for producing a silicon ingot includes: a natural sand crucible containing a silicon melt in which a cristobalite layer having a thickness between 0.25 mm and 1.25 mm is formed on an inner wall of the crucible; magnetic poles for generating a horizontal magnetic field; and a controller programed to produce a silicon ingot according to the following steps: rotating the natural sand crucible containing a silicon melt at two to five revolutions per minute; applying a horizontal magnetic field to the crucible using the magnetic poles; contacting the silicon melt with a seed crystal; and withdrawing the seed crystal from the silicon melt while rotating the natural sand crucible at two to five revolutions per minute to form a silicon ingot.SELECTED DRAWING: Figure 4

Description

Cross - reference to related applications

[0001] This application claims the priority of U.S. Provisional Patent Application No. 62 / 947,785, filed on December 13, 2019, the entire disclosure of which is incorporated herein by reference.

Technical Field

[0002] The present disclosure generally relates to the manufacture of silicon ingots, and more specifically to methods and systems for achieving a high success rate in manufacturing silicon ingots by the Czochralski process using a horizontal magnetic field.

Background Art

[0003] During the 1990s, at least some high - quality silicon growth was mainly controlled by the thermal conditions of the puller, especially the design of the hot zone (HZ) itself, because the ratio of the pulling speed to the thermal gradient (v / G) was considered the dominant factor. In the late 1990s, in the growth of at least some high - quality silicon, further consideration of the crystal / melt interface at the same v / G was incorporated. Around this time, there were more customers migrating from epi to polish and from 200 mm to 300 mm, and the application of high - quality silicon to memory devices expanded substantially. Shortly thereafter, it was established that the growth of high - quality silicon requires very stable process growth conditions and controlled melt flow to achieve a specific crystal / melt necessary to achieve the desired low crystal defect rate during growth.

[0004] Peripherally, as silicon crystal growth migrated from 200 mm to 300 mm and the corresponding charge size increased to maintain productivity, the need for magnetic field application to stabilize the melt flow in the increased melt volume became recognized as a dominant feature.

[0005] Some silicon manufacturers transitioned to the horizontal magnetic field Czochralski process (HMCZ) in the early 2000s, when high-quality production of 300mm silicon began, in order to effectively control the crystal / melt interface. Other silicon manufacturers used cusp magnetic fields for the production of high-quality 300mm diameter silicon. In either case, the magnetic field in the silicon melt had a dramatic impact on crystal quality and performance, and all manufacturers developed their own techniques from the outset to optimize performance and quality.

[0006] In processes that use cusp zoning (CZ) and magnetic fields to manufacture single-crystal silicon ingots, oxygen can be introduced into the silicon crystal ingot via the melt-solid interface or melt-crystal interface. This oxygen can cause various defects in wafers manufactured from the ingot, potentially reducing the yield of semiconductor devices manufactured using the ingot. For example, low interstitial oxygen concentrations (Oi) are generally required to achieve high resistance in memory devices, insulated-gate bipolar transistors (IGBTs), high-quality radio frequency (RF), high-resistance silicon-on-silicon (HR-SOI), and charge-trapped SOI (CTL-SOI) applications. In the case of HMCZ processes, it was generally thought that very low crucible rotation (C / R) was required to control oxygen in the growing crystal, particularly to control the oxygen content within the desired range applicable to memory devices. Furthermore, compared to processes using cusp magnetic fields, HMCZ has been found to have a higher rate of zero-loss dislocations (LZDs) from crown to body edge.

[0007] Thus, there is a need for a method and system that reduces LZD loss in HMCZ growth and provides an improved ZD success rate for high-quality silicon growth from crown to body.

[0008] This background technology section is intended to introduce to the reader various aspects of the technology that may be relevant to the various aspects of the disclosure described and / or claimed below. This discussion is intended to be useful in providing the reader with background information to facilitate a better understanding of the various aspects of the disclosure. Therefore, these statements should be read in this context and should be understood not as an endorsement of prior art. [Overview of the project]

[0009] In one aspect of this disclosure, a method for producing a silicon ingot by the horizontal magnetic field Czochralski process includes rotating a crucible containing a silicon molten material, applying a horizontal magnetic field to the crucible, bringing the silicon molten material into contact with a seed crystal, and rotating the crucible to form a silicon ingot while simultaneously withdrawing the seed crystal from the silicon molten material. The crucible has a wettable surface on which a cristobalite layer is formed.

[0010] Another embodiment is a wafer produced from a silicon ingot manufactured using the method described above.

[0011] Another embodiment is a system for manufacturing silicon ingots. This system includes a controller configured to manufacture silicon ingots according to the method described above.

[0012] Various improvements exist to the features pointed out in relation to the embodiments described above. Further features can also be incorporated into the embodiments described above. These enhanced and added features may exist individually or in any combination. For example, the various features described below in relation to any of the illustrated embodiments may be incorporated into the embodiments described above, individually or in any combination. [Brief explanation of the drawing]

[0013] [Figure 1] This is a top view of a crucible in one embodiment. [Figure 2] Figure 1 is a side view of the crucible shown. [Figure 3] This is a schematic diagram illustrating the horizontal magnetic field applied to the crucible containing the molten metal in a crystal growth apparatus. [Figure 4] This is a block diagram of a crystal growth system. [Figure 5] This shows the temperature field (mm) of the free surface of the molten metal and the rotation speed (RPM) of the crucible, as measured by MGP. [Figure 6] This is a graph of the crystallization rate as a function of time. [Figure 7] This graph shows the thickness of the formed layer as a function of time at a temperature of 1360°C. [Figure 8] This is a contour plot showing the relationship between the success rate of ZD (Zero Deposition) by crucible rotation in a natural sand crucible and the amount of melt modifier.

[0014] Similar reference numerals in various drawings indicate the same elements. [Modes for carrying out the invention]

[0015] Referring first to Figures 1 and 2, a crucible of one embodiment is shown as 10 in its entirety. The cylindrical coordinate system of the crucible 10 includes the radial direction R12, the angular direction θ14, and the axial direction Z16. The crucible 10 contains a melt 25 having a melt surface 36. A crystal 27 grows from the melt 25. The melt 25 may contain one or more convection cells 17, 18 caused by heating of the crucible 10 and rotation of the crucible 10 and / or the crystal 27 in the angular direction θ14. As will be described in detail below, the structure and interaction of these one or more convection cells 17, 18 are regulated by tuning one or more process parameters and / or applying a magnetic field.

[0016] Figure 3 illustrates the horizontal magnetic field applied to a crucible 10 containing a molten metal 25 in a crystal growth apparatus. As shown in the figure, the crucible 10 contains a silicon molten metal 25 from which a crystal 27 is grown. The transition between the molten metal and the crystal is generally called the crystal-molten interface (alternatively, the molten-crystal, solid-molten, or molten-solid interface) and is typically nonlinear, for example, concave, convex, or gull-wing with respect to the molten surface. Two magnetic poles 29 are positioned opposite each other to generate a magnetic field that is approximately perpendicular to the crystal growth direction and approximately parallel to the molten surface 36. The magnetic poles 29 may be conventional electromagnets, superconducting electromagnets, or any other suitable magnet for generating a horizontal magnetic field of the desired strength. The application of a horizontal magnetic field generates a Lorentz force along the axial direction, opposite to the fluid motion, counteracting the force driving the molten convection. This suppresses the molten convection and increases the axial temperature gradient of the crystal near the interface. Furthermore, the melt-crystal interface moves upward toward the crystal side in accordance with the increasing axial temperature gradient of the crystal near the interface, reducing the contribution of melt convection within the crucible. The horizontal configuration has the advantage of being efficient in attenuating convection at the melt surface 36.

[0017] Figure 4 is a block diagram of the crystal growth system 100. System 100 employs the Czochralski crystal growth method to manufacture semiconductor ingots. In this embodiment, system 100 is configured to manufacture cylindrical semiconductor ingots having an ingot diameter of 150 mm (150 mm), greater than 150 mm (150 mm), more specifically in the range of about 150 mm to 460 mm, and even more specifically about 300 mm (300 mm). In other embodiments, system 100 is configured to manufacture semiconductor ingots having an ingot diameter of 200 mm (200 mm) or 450 mm (450 mm). In addition, in one embodiment, system 100 is configured to manufacture semiconductor ingots having an ingot length of at least 900 mm (900 mm). In some embodiments, the system is configured to manufacture semiconductor ingots having lengths of 1950 mm, 2250 mm, 2350 mm, or longer than 2350 mm. In other embodiments, system 100 is configured to manufacture semiconductor ingots having total ingot lengths in the range of approximately 900 mm to 1200 mm, between approximately 900 mm and approximately 2000 mm, or between approximately 900 mm and approximately 2500 mm. In some embodiments, the system is configured to manufacture semiconductor ingots having total ingot lengths exceeding 2000 mm.

[0018] The crystal growth system 100 includes a vacuum chamber 101 surrounding the crucible 10. A side heater 105, such as a resistance heater, surrounds the crucible 10. A bottom heater 106, such as a resistance heater, is disposed below the crucible 10. During heating and crystal pulling, a crucible drive unit 107 (e.g., a motor) rotates the crucible 10, as indicated by arrow 108, for example, in a clockwise direction. Also, the crucible drive unit 107 may raise and / or lower the crucible 10 as desired during the growth process. Inside the crucible 10 is a silicon melt 25 having a melt level or melt surface 36. In operation, the system 100 pulls a single crystal 27 starting from a seed crystal 115 attached to a pull shaft or cable 117 from the melt 25. One end of the pull shaft or cable 117 is connected via a pulley (not shown) to a drum (not shown), or any other suitable type of lifting mechanism, such as a shaft, and the other end is connected to a chuck (not shown) holding the seed crystal 115 and the crystal 27 grown from the seed crystal 115.

[0019] The crucible 10 and the single crystal 27 have a common symmetry axis 38. When the melt 25 depletes, the crucible drive unit 107 can raise the crucible 10 along the axis 38 to maintain the melt level 36 at a desired height. A crystal drive unit 121 similarly rotates the pull shaft or cable 117 in a direction 110 opposite to the direction in which the crucible drive unit 107 rotates the crucible 10 (e.g., counter - rotation). In an embodiment using co - rotation, the crystal oscillator drive unit 121 may rotate the pull shaft or cable 117 in the same direction (e.g., clockwise direction) as the direction in which the crucible drive unit 107 rotates the crucible 10. Co - rotation is also called co - rotation. Further, the crystal drive unit 121 raises and lowers the crystal 27 relative to the melt level 36 as desired during the growth process.

[0020] According to the Chokralski single crystal growth process, a predetermined amount of polycrystalline silicon, i.e., polysilicon, is loaded into the crucible 10. The heater power supply 123 energizes the resistance heaters 105, 106, and the heat insulating material 125 aligns with the inner wall of the vacuum chamber 101. The gas supply device 127 (e.g., a bottle) supplies argon gas into the vacuum chamber 101 via the gas flow controller 129 when the vacuum pump 131 removes gas from the vacuum chamber 101. The external chamber 133 to which cooling water is supplied from the reservoir 135 surrounds the vacuum chamber 101.

[0021] The cooling water is then discharged into the cooling water return manifold 137. Typically, a temperature sensor such as the optical element 139 (or pyrometer) measures its temperature at the surface of the melt 25, and the diameter transducer 141 measures the diameter of the single crystal 27. In this embodiment, the system 100 does not include an upper heater. The presence or absence of the upper heater changes the cooling characteristics of the crystal 27.

[0022] The magnetic poles 29 are arranged outside the vacuum chamber 101 to generate a horizontal magnetic field (shown in FIG. 3). Although shown approximately at the center of the melt surface 36, the position of the magnetic poles 29 with respect to the melt surface 36 may be changed to adjust the position of the maximum gauss surface (MGP) with respect to the melt surface 36. The reservoir 153 supplies cooling water to the magnetic poles 29 before being drained via the cooling water return manifold 137. The iron shield 155 surrounds the magnetic poles 29, reducing the stray magnetic field and enhancing the strength of the generated magnetic field.

[0023] The control unit 143 is used to adjust several process parameters, including but not limited to at least one of the crystal rotation speed, crucible rotation speed, and magnetic field strength. In various embodiments, the control unit 143 may also control a memory 173 and a processor 144 that process signals received from various sensors of the system 100, including but not limited to the optical element 139 and the diameter transducer 141, as well as one or more devices of the system 100, including but not limited to the crucible drive unit 107, the crystal drive unit 121, the heater power supply 123, the vacuum pump 131, the gas flow controller 129 (e.g., an argon flow controller), the magnetic pole power supply 149, and any combination thereof. The memory 173 may store instructions, when executed by the processor 144, that cause the processor to perform one or more of the methods described herein. That is, the instructions are configured to cause the control unit 143 to perform one or more methods, processes, procedures, etc., described herein.

[0024] The control unit 143 may be a computer system. As described herein, a computer system refers to any known computer device and computer system. As described herein, all such computer systems include a processor and memory. However, any processor in a computer system referred to herein may also refer to one or more processors where the processor is located in one computer device or in multiple computer devices operating in parallel. Furthermore, any memory in a computer device referred to herein may also refer to one or more memories where the memory is located in one computer device or in multiple computer devices operating in parallel. Furthermore, the computer system may be located near system 100 (e.g., in the same room or an adjacent room) or may be located remotely and connected to the rest of the system via a network such as Ethernet or the Internet.

[0025] As used herein, the term "processor" refers to a central processing unit, microprocessor, microcontroller, reduced instruction set circuit (RISC), application-specific integrated circuit (ASIC), logic circuit, and any other circuit or processor capable of performing the functions described herein. These are merely illustrative examples and are not intended to limit in any way the definition and / or meaning of the term "processor." Memory may include, but is not limited to, random access memory (RAM), such as dynamic RAM (DRAM) or static RAM (SRAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), and non-volatile RAM (NVRAM).

[0026] In one embodiment, a computer program is provided to implement the control unit 143, and this program is implemented on a computer-readable medium. The computer-readable medium may include the memory 173 of the control unit 143. In an exemplary embodiment, the computer system runs on a single computer system. Alternatively, the computer system may consist of multiple computer systems, connections to server computers, a cloud computing environment, etc. In some embodiments, the computer system includes multiple components distributed among multiple computer devices. One or more components may be in the form of computer-executable instructions implemented on a computer-readable medium.

[0027] Computer systems and processes are not limited to the specific embodiments described herein. Furthermore, each component of a computer system and each process can be implemented independently and separately from other components and processes described herein. Each component and process can also be used in combination with other assembly packages and processes.

[0028] In one embodiment, the computer system may be configured to receive measurements from one or more sensors, including but not limited to a temperature sensor 139, a diameter transducer 141, and any combination thereof, as described herein and shown in Figure 4, or to control one or more devices of system 100, including but not limited to a crucible drive unit 107, a crystal drive unit 121, a heater power supply 123, a vacuum pump 131, a gas flow controller 129 (e.g., an argon flow controller), a magnetic pole power supply 149, and any combination thereof. The computer system performs all steps used to control one or more devices of system 100, as described herein.

[0029] The loss of zero-dislocation (ZD) structures (quantified by the LZD rate) is generally higher during silicon crystal Cz growth in a horizontal magnetic zone (HMCZ) than during growth in a cusp (or perpendicular) magnetic field. However, growing crystals in a synthetic sand crucible instead of a natural sand crucible dramatically reduces the LZD rate in the HMCZ. While the ZD rate improves, synthetic liner crucibles are more expensive than natural sand crucibles. Furthermore, thin (approximately 2 mm thick) synthetic liners melt in relatively short process times, exposing the backing sand layer to the melt, which can allow quartz particles to enter the melt and collide with the growing crystal. This results in more bubbles within the liner or exposure of bubbles from the backing sand to the melt. To avoid bubble exposure and / or backing sand contamination due to melting when using synthetic liner crucibles, the high-temperature process time is generally limited to less than ~250 hours, which is a much shorter process time than can be achieved with natural sand crucibles (approximately 400-500 hours or more). The high-temperature time required for crystal growth depends on process conditions, HZ configuration, and trials, which can affect refilling capacity and the growth of multiple rods per batch when using synthetic liner crucibles. Therefore, silicon growth using HMCZ in synthetic liner crucibles generally requires optimization of crucible conditions to ensure the highest ZD rate and minimum trials so that maximum refilling capacity can be achieved.

[0030] Furthermore, in Cz growth, the horizontal magnet promotes the flow of the melt by causing irregularly velocating melt waves to continuously strike the crucible wall with strong transient force. In this case, the surface condition of the crucible is crucial for the formation of quartz fragments, which are directly related to the success of ZD. As shown in Figure 5, the melt flow and temperature field are greatly influenced by the strength of the magnetic field and the position of the MGP. As shown in the figure, the circumferential temperature change at the melt surface changes significantly depending on the direction of the magnetic field.

[0031] These and other difficulties can be overcome or mitigated in embodiments of this disclosure by using one or both of two techniques described in detail below. Generally, in the first embodiment, the cristobalite layer is formed on a wettable surface inside the crucible. The wettable surface generally refers to the surface of the crucible that may come into contact with the melt during silicon production. The wettable surface generally includes the inside bottom of the crucible, at least a portion of the inside sidewalls of the crucible, and the internal portion connecting the inside sidewalls and the bottom of the crucible. In Figure 2, the wettable surface is the entire inner surface of the crucible 10 below and including the melt surface 36. The wettable surface may also extend above the melt surface 36. The second technique described in this disclosure is increasing the rotation speed of the crucible.

[0032] Generally, a strong, unsteady molten flow is induced by an HMCZ magnetic field, which can generate strong thermomechanical stress and mechanical shock to the crucible walls, leading to the generation of quartz particles. However, a high crucible rotation speed (C / R) can cause faster convection near the crucible walls, disrupting the magnetic field-driven molten flow. This reduces stress and shock to the walls. As a result, the generation of quartz particles at the walls decreases, and consequently, LZD during crystal growth can be reduced.

[0033] LZD can also be reduced by forming a generally uniform crystalline SiO2 layer (called the cristobalite layer) on the wettable surface of the crucible. This layer is more stable and robust than amorphous quartz itself, and therefore has higher resistance to molten attack due to stress and mechanical impact. Thus, the cristobalite layer reduces the generation of quartz particles.

[0034] There are at least two methods for promoting the growth of crystal layers on a quartz crucible. The first method is to use a crucible pre-coated with a compound such as BaOH that promotes cristobalite growth, and the other is to add a suitable melt modifier (MM) to the melt before crystal growth. Non-limiting examples of suitable MMs include barium (Ba) and strontium (Sr). More specifically, non-limiting examples of preferred MMs include barium carbonate (BaCO3), barium oxide (BaO), and strontium carbonate (SrCO3).

[0035] Cristobalite formation on the inner wall of amorphous quartz crucibles is governed by pressure, Oi concentration, H2O and hydrogen content, temperature, etc. As shown in Figures 6 and 7, the formation and growth of crystalline layers on the crucible walls are governed by the temperature of the crucible walls and the concentration of MM consumed in the crucible. Figure 6 compares the crystallization rates as a function of time for no MM and 8% AL2O3 MM. Figure 7 compares the thickness of the layer formed at a temperature of 1360°C as a function of time for no MM, 8% AL2O3 MM, and barium-based MM. These graphs show that the addition of appropriate MM (e.g., barium-based MM) to the melt or pre-coating with a barium compound generates a uniform and thick crystalline (i.e., cristobalite) layer on the wall surface of natural sand crucibles, resulting in behavior and performance similar to synthetic liner crucibles. Because the cristobalite layer dissolves more slowly than fused quartz, it reduces the generation of quartz fragments due to thermal and mechanical stress, and can suppress the subsequent generation of crystalline LZD. Furthermore, the generation and propagation of secondary bubbles into the molten material in the BFL (Bubble Free Layer) of natural sand crucibles is generally far less than in synthetic sand crucibles due to differences in material properties.

[0036] The formation and growth of a uniform cristobalite layer, either by pre-coating the crucible or post-adding a melt modifier to the melt, is initiated in a stabilization mode before crystal growth (i.e., after or during the melting of polysilicon). In the case of MM post-addition, the rate of cristobalite formation is slower than in the case of pre-coating, because MM is introduced after melting. However, post-addition of MM reduces air pocket (APK) loss because it allows trapped air bubbles formed on the crucible wall to be released to the surface before a stable cristobalite layer is formed. Since it typically takes 3-7 hours from the start of the stabilization step to the start of main growth, the cristobalite thickness is estimated to be approximately 2 mm or more when an appropriate amount of MM is added to increase the rate of cristobalite formation. In practice, it is common for a cristobalite layer of approximately 1.0 mm or less to form on the surface of the crucible (whether natural or synthetic sand). ++ Ya N ++In special cases, such as high-concentration doping processes, thicker cristobalite layers can be obtained by adding large amounts of MM. In such cases, a cristobalite layer of approximately 1.0 mm (±) is formed. The reason these thicknesses differ from 2.0 mm is that, as shown in Figure 7, the cristobalite layer grows by hot time, while the wet cristobalite layer is continuously dissolved in the melt. Other embodiments include cristobalite layers of approximately 2.0 mm, thicker than 1.5 mm, thicker than 1.0 mm, thicker than 0.75 mm, thicker than 0.5 mm, or thicker than 0.25 mm formed on the wettable surface of the crucible. In some embodiments, the cristobalite layer is less than 3.0 mm thick, less than 2.0 mm thick, less than 1.25 mm thick, or less than 1.0 mm thick. In some embodiments, the cristobalite layer falls within the range defined by the minimum and maximum values ​​described above, such as between 0.25 mm and 1.25 mm. In general, if the cristobalite layer is too thin, it may be insufficient to provide the advantages described herein, while if the cristobalite layer is too thick, it may be more likely to break and enter the melt during silicon manufacturing (potentially contributing to LZD).

[0037] The addition of the aforementioned melt modifier forms a uniform crystalline layer on the wettable surface of the crucible, and this crystalline structure has strong resistance to thermomechanical stress caused by the irregular (transient) melt flow that occurs in the HMCZ. The formation of a thick, uniform cristobalite can resist stress and shock from the melt flow, reduce damage to the crucible surface (i.e., resist damage that generates quartz particles in the melt), and increase the success of ZD.

[0038] As mentioned above, the strong convection caused by the high rotation of the crucible suppresses temperature fluctuations at the free surface of the molten metal, as seen in Case 6 of Figure 5, and reduces the molten metal flow caused by the horizontal magnet. The C / R ratio is 0.6 RPM in Cases 1-5 and Case 7, while it is 1.6 RPM in Case 6. This indicates that the force generated from the molten metal to the crucible wall by the magnetic field is reduced or inhibited by the convection associated with the high rotation of the crucible. This reduces the possibility of quartz fragment generation due to damage to the crucible wall and improves the likelihood of successful ZD (Zero Deposition).

[0039] Under horizontal magnetic fields, test conditions were implemented to understand the ZD success rate of synthetic liner crucibles and natural sand crucibles as a function of crucible rotation speed and the amount of melt modifier added. Synthetic crucibles showed high ZD success rates over a wide range of crucible rotation and MM. For natural sand liner crucibles, a total of 96 trials were performed under 19 different conditions, and the results are summarized in the contour plot in Figure 8. The data in Figure 8 ranges from a body length of approximately 200 mm to near the OE of the crystal. In Figure 8, CR is the crucible rotation speed (RPM), success%_3 of 0.0 is ZD failure (LZD), and success%_3 of 1.0 is 100% ZD success. The sign of the crucible rotation indicates the direction of rotation. In some pulls, the crucible rotation sloped between velocities for part of the pull, and arbitrary values ​​of crucible rotation were selected for plotting, with data collected from body lengths of 1000 mm or more to approximately OE of the crystal. The results clearly show that increasing the absolute value of the crucible rotation speed and the amount of melt modifier (in this case, Ba-based) leads to a higher ZD success rate.

[0040] Increasing crucible rotation speed and using the formation of a cristobalite layer have been shown, individually and in combination, to result in improved ZD (zooming out) outcomes. At low C / Rs, such as between approximately 0 and 2 RPM, the addition of appropriate molten metal (MM) (or the use of pre-coated crucibles) is necessary to increase the ZD success rate. As the C / R increases from approximately 2 RPM to approximately 5 RPM, the amount of MM required to achieve a comparable success rate decreases, and MM may even be zero. Within this range, the ZD success rate can be further improved by using at least MM. Above approximately 5 RPM, there is generally no concern about ZD from the crucible, and MM appears unnecessary. However, at such speeds, additional process conditions are likely to be required to control product quality, as there may be other issues such as oxygen control by melt flow rate and melt level control by centrifugal force.

[0041] Accordingly, some embodiments of the present disclosure use more than 1.7 grams of MM per square meter of the crucible's wettable surface during the HMCZ process at any C / R rate to improve the ZD success rate. In some embodiments, 1.7 to 2.0 grams / m of the crucible's wettable surface is used. 2 In other embodiments, MM is used for the wettable surface of the crucible, ranging from 1.7 to 5.4 grams / m². 2 In further embodiments, MM between these values ​​is used. 2 Larger MMs may be used, but excessive amounts of MMs can cause LZD in multiple refilling processes. (e.g., 1.7 grams / m³) 2 This is the case when BaCO3 is used as the moisture manager (MM). Similar embodiments using BaO or SrCO3 include a specific amount of MM that is functionally equivalent to the amount of BaCO3.

[0042] Some embodiments use approximately 0-0.5 g / m². 2 A crucible rotation of MM or more than approximately 2.0 RPM is used. In some such embodiments, the crucible is a natural sand crucible. Alternatively, the crucible may be a synthetic crucible.

[0043] Embodiments of the methods described herein achieve superior results compared to previous methods and systems. For example, the methods described herein facilitate the production of silicon with a higher ZD success rate than several other methods.

[0044] When describing elements of the present invention or its embodiments, the articles "a," "an," "the," and "said" are intended to indicate that there is one or more of those elements. The terms "comprising," "including," and "having" are intended to indicate comprehensiveness and that there may be additional elements other than those described.

[0045] The approximate terms used throughout this specification and the claims may be applied to modify any quantitative expression that may be permissibly varied without resulting in a change in the underlying function of the expression. Thus, values ​​modified by terms such as “about,” “approximately,” and “substantially” are not limited to the specified exact value. In at least some examples, approximate language may correspond to the precision of an instrument used to measure a value. Throughout this specification and the claims, scope limitations may be combined and / or interchangeable, and such scopes identify and include all sub-scopes contained therein unless otherwise indicated by context or language.

[0046] Since various modifications as described above can be made without departing from the scope of the present invention, all matters included in the above description and shown in the accompanying drawings are intended to be interpreted as illustrative and not restrictive.

Claims

1. A system for manufacturing silicon ingots, A crucible made of natural sand for containing molten silicon, wherein a cristobalite layer with a thickness of 0.25 mm to 1.25 mm is formed on the inner wall of the crucible; Magnetic poles for generating a horizontal magnetic field; and, This is a controller, and the procedure is as follows: Rotating a crucible made of natural sand containing molten silicon at a speed of 2 to 5 revolutions per minute; Applying a horizontal magnetic field to a crucible using magnetic poles; Contacting a seed crystal with a silicon melt; and, The process involves rotating a crucible made of natural sand at a speed of 2 to 5 revolutions per minute while drawing seed crystals from molten silicon to form silicon ingots. A controller programmed to manufacture silicon ingots, A system that includes this.

2. The system according to claim 1, wherein the controller is further programmed to add a melt modifier to the natural sand crucible in order to form a cristobalite layer on the inner wall of the natural sand crucible.

3. The system according to claim 2, wherein the controller is programmed to add a melt modifier while heating polycrystalline silicon in a natural sand crucible to form a silicon melt.

4. The system according to claim 2, wherein the controller is programmed to add a melt modifier after the silicon melt has been formed.

5. The melt modifier is barium carbonate (BaCO3). ３ The system according to claim 2, which includes )

6. The system according to claim 5, wherein the controller is programmed to add more than 1.7 grams of barium carbonate per square meter of the inner wall of the natural sand crucible.

7. The controller uses barium oxide (BaO) or strontium carbonate (SrCO) as a melt modifier. ３ The system according to claim 2, which is programmed to add one of the following:

Images