System and method for forming a single crystal silicon ingot using a crucible with a composite liner

Abstract

A method for manufacturing a single crystal silicon ingot from a silicon melt includes providing a crucible within an internal chamber of an ingot pulling apparatus, the crucible including an inner surface and a synthetic liner on the inner surface. The method further includes adding an initial charge of polysilicon to the crucible, melting the initial charge of polysilicon to form a silicon melt within the crucible, dissolving a melt modifier in the silicon melt to devitrify the synthetic liner, and forming a crystallization layer on the crucible. The thickness of the crystallization layer is less than 700 microns. The method further includes pulling a single crystal silicon ingot from the silicon melt.

Description

Cross - Reference to Related Applications

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 369,786, filed on July 29, 2022, the entire disclosure of which is incorporated herein by reference.

Technical Field

[0002] This field relates to systems and methods for forming single - crystal ingots using crucibles with synthetic liners, and more particularly, to systems and methods for forming single - crystal silicon ingots by the Czochralski method or the continuous Czochralski method in which the synthetic liner is used in combination with a melt modifier to promote devitrification of the liner.

Background Art

[0003] Single - crystal silicon, which is the starting material for most processes for manufacturing many electronic components such as semiconductor devices and solar cells, is typically produced by the Czochralski method (CZ method) or the continuous Czochralski method (CCZ method). In these methods, a polycrystalline raw material such as polycrystalline silicon in the form of a solid raw material is filled into a quartz crucible and melted, a single seed crystal is brought into contact with the molten silicon or melt, and a single - crystal silicon ingot is grown by slowly extracting it.

[0004] The quartz crucible includes an inner surface that defines a chamber for containing the melt. Some crucibles include a vitreous silica liner having a coating on the inner surface. Such a liner can be designed to undergo a devitrification process at high temperatures, whereby the liner crystallizes to become cristobalite rather than vitreous silica, and a crystallization layer is formed on the crucible. The devitrification of the liner improves the rigidity of the heated crucible sidewall, reduces the reaction between the melt and the crucible that can cause vibrations in the silicon melt and adversely affect the resulting ingot. Synthetic silica or natural silica liners can be used.

[0005] Synthetic silica liners have the advantage of having a lower impurity content compared to natural silica liners, thereby reducing the potential for melt contamination. However, one problem with synthetic liners within the crucible is that random nucleation occurs on the synthetic liner during the devitrification process, and as these nuclei grow, the resulting devitrification also becomes random. As a result, the surface of the melt becomes a mixed glassy and crystalline structure. As a result of the mixed surface, synthetically lined crucibles need to be operated at shorter intervals to limit the exposure time to the melt, also known as "hot hours", in the absence of other modifications. Furthermore, without modification, the surface roughness of the resulting crystallization layer increases, increasing bubble nucleation in the melt and potentially damaging the ingot.

[0006] To promote devitrification of the synthetic liner, a devitrifying agent or nucleating agent (terms used interchangeably herein), such as barium (Ba), calcium (Ca), or strontium (Sr), can be coated on the synthetic liner. However, when nuclei are formed by such substances, the devitrification of the liner proceeds much faster compared to natural liners, and the thickness of the crystallization layer increases. For example, among the crystallization layers generated from synthetic liners coated with a nucleating agent, when the concentration of the nucleating agent is the same, some have a thickness more than five times that of natural liners. For instance, the thickness of a typical devitrified synthetic liner on a crucible is about 2 millimeters (mm).

[0007] When the thickness of the synthetic devitrified liner becomes thicker compared to the natural liner, the structural stability of the devitrified layer decreases during use. For example, during the growth of an ingot, the devitrified layer is subjected to thermal cycling in a wide temperature range from about 1100 degrees Celsius to 1500 degrees Celsius, resulting in temperature differences in various sections along the thickness of the liner, causing strain in the liner. When the strain becomes large enough, thermal peeling occurs, a part of the liner cracks into the melt, which may generate particles in the melt and defects in the crystalline ingot. Furthermore, since the synthetic liner is thicker than the natural liner, it may move towards and contact the meniscus of the growing ingot before melting. When the unmelted part of the liner contacts the meniscus, the zero dislocation structure is lost, the single crystal dislocates and becomes a polycrystalline material, and as a result, the crystal becomes unusable for device applications.

[0008] This section is intended to introduce readers to various aspects of technologies that may be related to various aspects of the disclosure described and / or claimed below. This description is thought to be helpful in providing readers with background information to better understand the various aspects of the present disclosure. Therefore, it should be understood that these descriptions are not to be read as an identification of prior art from this perspective. Summary

[0009] One aspect of the present disclosure relates to a method for manufacturing a single crystal silicon ingot from a silicon melt held in a crucible disposed within an ingot pulling apparatus. The method includes providing a crucible within an internal chamber of the ingot pulling apparatus, the crucible including an inner surface and a synthetic liner on the inner surface. The method further includes adding an initial fill of polysilicon to the crucible, melting the initial fill of polysilicon to form a silicon melt within the crucible, dissolving a melt modifier in the silicon melt to devitrify the synthetic liner, and forming a crystallization layer on the crucible. The thickness of the crystallization layer is less than 700 microns. The method further includes pulling a single crystal silicon ingot from the silicon melt.

[0010] Another aspect of the present disclosure relates to a crucible for use in an ingot pulling apparatus. The crucible includes a body having an outer surface opposite an inner surface and a synthetic liner provided on the inner surface of the body. The synthetic liner has a composition that devitrifies when exposed to a silicon melt during a crystal growth operation to form a crystallization layer on the crucible. The thickness of the crystallization layer is less than 700 microns.

[0011] Another aspect of the present disclosure relates to a method for manufacturing a single crystal silicon ingot from a silicon melt held in a crucible disposed within an ingot pulling apparatus. The method includes providing a crucible within an internal chamber of the ingot pulling apparatus, the crucible including an inner surface and a synthetic liner on the inner surface. The method further includes adding an initial fill of polysilicon to the crucible and adding a melt modifier precursor to the crucible. The method further includes melting the initial fill of polysilicon to form a silicon melt within the crucible and dissolving the melt modifier precursor in the silicon melt to release a concentration of a melt modifier into the silicon melt. The concentration of the melt modifier causes the synthetic liner to devitrify and a crystallization layer to be formed on the crucible. The thickness of the crystallization layer is less than 700 microns.

[0012] There are various improvements to the features described in connection with the above aspects. Further features may be incorporated into the above aspects. These improvements and additional functions may exist individually or in any combination. For example, the various features described below in connection with any of the illustrated embodiments can be incorporated into any of the above aspects, either alone or in any combination.

Brief Description of the Drawings

[0013]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Embodiments for Carrying Out the Invention

[0014] Throughout the drawings, corresponding reference numerals indicate corresponding parts.

[0015] Referring to FIG. 1, an ingot pulling apparatus, or ingot puller, is schematically shown and generally designated by reference numeral 100. The ingot pulling apparatus 100 is used to produce single crystal (i.e., monocrystalline) ingots of semiconductor or solar cell grade material, such as single crystal silicon ingots for example. The exemplary ingot is grown by a so-called Czochralski (CZ) process that pulls the ingot from a silicon melt 102 held within a crucible 104 of the crystal pulling apparatus 100. The ingot can also be grown by a batch CZ process that fills the crucible 104 with an amount of polycrystalline silicon sufficient to grow one ingot and causes the silicon melt 102 within the crucible 104 to be substantially depleted after the growth of one ingot. In other embodiments, the ingot may be grown by a continuous CZ (CCZ) process in which polycrystalline silicon is continuously or periodically added to the crucible 104 to replenish the silicon melt 102 during the growth process. The CCZ process facilitates the growth of multiple ingots drawn from a single crucible 104.

[0016] The ingot pulling apparatus 100 includes a housing 106 that defines a crystal growth chamber 108 and a pulling chamber 110 having a lateral dimension smaller than that of the growth chamber 108. The growth chamber 108 includes a generally domed upper wall 112 that transitions from the growth chamber 108 to the narrow pulling chamber 110. The ingot pulling apparatus 100 includes an inlet port 114 and an outlet port 116 that can be used to introduce process gas into and discharge process gas from the housing 106 during crystal growth.

[0017] The crucible 104 within the ingot lifting device 100 contains a silicon melt 102 for pulling out a silicon ingot. The crucible 104 is preferably made of quartz or fused silica that has high thermal stability and generally does not react with the molten silicon within the melt 102, although the crucible may slowly dissolve into the melt. The crucible 104 may be made of other materials in addition to quartz. For example, the quartz crucible 104 may be made of a composite material including silica and additional materials such as silicon nitride or silicon carbide. For example, in some applications such as solar cell grade wafers and polycrystalline ingots, crucibles formed of such composite materials may be used.

[0018] The silicon melt 102 is obtained by melting polycrystalline silicon filled in the crucible 104. A supply system (not shown) is used to supply solid raw materials to the crucible assembly 104 and / or the melt 102. The crucible 104 is disposed within a susceptor 118 and supported by the susceptor 118, which is supported by a rotating shaft 120. The susceptor 118 and the rotating shaft 120 serve to rotate the crucible 104 about the central longitudinal axis X of the ingot lifting device 100.

[0019] The heating system 122 (e.g., an electrical resistance heater 122) surrounds the susceptor 118 and the crucible 104, supplies heat by conduction through the susceptor 118 and the crucible 104, melts the silicon fill to produce the melt 102, and / or maintains the melt 102 in a molten state. The heater 122 can also extend below the susceptor 118 and the crucible 104. The heater 122 is controlled by a control system (not shown), and the temperature of the melt 102 is precisely controlled throughout the pulling process. For example, the controller can control the current supplied to the heater 122 to control the amount of thermal energy supplied by the heater 122. The controller can control the heater 122 such that the temperature of the melt 102 is maintained at a temperature higher than the melting point of silicon (e.g., about 1412 °C). For example, the melt 102 may be heated to a temperature of at least about 1425 °C, at least about 1450 °C, or at least about 1500 °C. The amount of heat lost from the housing 106 can be reduced by a heat insulating material (not shown) surrounding the heater 122. The ingot pulling apparatus 100 can be provided with a heat shield door assembly (not shown) on the surface of the melt 102 to protect the ingot from the heat of the crucible 104 and increase the axial temperature gradient at the solid-melt interface.

[0020] The lifting mechanism (not shown) is attached to a pull wire 124 that extends downward from the lifting mechanism. This mechanism can raise and lower the pull wire 124 and can also rotate the pull wire 124. The ingot lifting device 100 may include a pull shaft instead of a wire depending on the type of puller. The pull wire 124 terminates at a pull assembly 126 that includes a seed crystal chuck 128 that holds a seed crystal 130 used in the growth of a silicon ingot. During the growth of the ingot, the lifting mechanism lowers the seed crystal 130 until it contacts the surface of the silicon melt 102. When the seed crystal 130 begins to melt, the lifting mechanism slowly pulls the seed crystal through the growth chamber 108 and the lifting chamber 110 to grow a single crystal ingot. The speed at which the lifting mechanism rotates the seed crystal 130 and the speed at which the lifting mechanism lifts the seed crystal (i.e., the lifting speed V) are controlled by a control system. As the seed crystal 130 is slowly pulled from the melt 102, the silicon atoms of the melt 102 align along the seed crystal 130 and attach to form an ingot.

[0021] Process gas is introduced into the housing 106 through an inlet port 114 and discharged through an outlet port 116. The process gas creates an atmosphere within the housing, and the melt and the atmosphere form a melt gas interface. The outlet port 116 is in fluid communication with an exhaust system (not shown) of the ingot lifting device.

[0022] Referring to FIG. 2, the crucible 104 is schematically shown along with the silicon melt 102 contained therein. The crucible 104 includes a body 140 that includes a bottom wall 142 and side walls 144 that extend upward from the bottom wall 142. The bottom wall 142 and the side walls 144 define a cavity 146 that houses the melt 102. The body 140 further includes an inner surface 148 and an outer surface 150 that extend along the opposite sides of the bottom wall 142 and the side walls 144. The side walls 144 form an open edge 152 at the top of the crucible 104. The body 140 is formed of a natural silica material or quartz (these terms are used interchangeably herein).

[0023] The crucible 104 is further provided on the inner surface 148 of the main body 140 and includes a crystallization layer 154 facing the melt 102. The crystallization layer 154 includes a non-wetted portion 157 extending from the edge 152 to the surface of the melt 102 and a wetted portion 157 extending below the non-wetted portion 157 and below the surface of the melt 102. The wetted portion 157 contacts the melt 102 during at least a part of the ingot growth process, while the non-wetted portion 157 does not contact the melt 102. The crystallization layer 154 on the crucible 104 is formed from a devitrified synthetic liner 158 (shown in FIG. 3) provided on the inner surface 148 of the crucible 104. When the liner 158 is immersed in the melt 102 of the wetted portion 157, the wetted portion 157 of the layer crystallizes, but the non-wetted portion 157 of the layer 154 does not crystallize during the ingot growth process.

[0024] FIGS. 3 and 4 show enlarged views of region A (shown in FIG. 2) of the crucible 104. FIG. 3 shows region A of the crucible 104 with the synthetic liner 158 before the liner is immersed in the melt 102. FIG. 4 shows region A of the crucible 104 after the liner has been immersed in the melt 102 and devitrified to form the crystallization layer 154. The approximate ratios of the liner, the crystallization layer 154, and the crucible 104 are not shown to scale in the figures. In particular, the thicknesses of the liner and the crystallization layer 154 are exaggerated in the figures.

[0025] Referring to FIG. 3, the liner 158 is formed of synthetic silica and is also referred to herein as a "synthetic liner". Table 1 below shows the chemical composition of the synthetic liner 158 in parts per million ("ppm") compared to a natural quartz or "natural sand" liner. As shown in Table 1, the synthetic liner 158 contains significantly reduced impurity concentrations compared to the natural liner.

[0026] Table 1: Examples of impurity compositions of natural and synthetic liners 158 TIFF2025524205000002.tif164152

[0027] The liner 158 extends continuously along the bottom wall 142 along the inner surface 148 of the crucible 104 from the edge 152 and substantially covers the inner surface 148. In other embodiments, at least a portion of the surface of the body 140, for example, the portion of the body 140 close to the non-wetted portion 157, may not be covered by the liner. The crucible 104 is manufactured together with the liner by an arc melting process, but other manufacturing processes suitable for the crucible 104 such as plasma spraying, slip casting, or 3D printing can also be used.

[0028] The liner 158 is designed and adapted for the devitrification process at high temperatures. When the liner 158 is exposed to the high-temperature melt 102, as shown in FIG. 4, the liner 158 crystallizes by the melt 102. To form the crystallization layer 154 (see FIG. 2), an initial filling of polysilicon is added to the crucible 104 and melted to form the melt 102. Further, before or after the polysilicon is melted, a melt modifier precursor is introduced into the crucible 104. When the crucible 104 is heated, the polysilicon melts to form the silicon melt 102. The melt modifier precursor reacts with the melt 102, releases waste gas, and dissolves the melt modifier in the melt 102. The waste gas (for example, carbon dioxide) is discharged from the crucible 104.

[0029] The melt modifier precursor is introduced into the melt 102 using a supply tube (not shown) that extends at least partially into the housing 106 (FIG. 1) and is arranged such that the precursor can flow into the melt 102. In other embodiments, the melt modifier precursor and / or the melt modifier itself can be introduced into the melt 102 by various devices and methods, including but not limited to: transplanting the precursor into the cavity 146 of a seed immersed in the melt 102, placing the precursor on a silicon carriage (not shown) and then into the melt 102, placing a quartz trapdoor container containing the precursor on top of the melt 102 and opening the door to release the precursor into the melt 102, etc.

[0030] The melt modifier reacts with the melt 102, and the modified melt 102 (also referred to as the "doped melt 102") acts as a nucleating agent on the liner, promoting the devitrification of the synthetic liner 158 and forming the crystallization layer 154. The silicon melt 102 is doped with a modifier at a concentration of about 0.013 milligrams per kilogram of silicon.

[0031] The melt 102 may be doped prior to each crystal growth operation or may be doped periodically depending on whether the Cz process used and the doping affect the crystal yield of each Cz process. For example, some Cz processes are less aggressive to heat, so it may not always be necessary to dope every time between the first crystal and subsequent crystals. In other Cz processes, doping between the first crystal and subsequent crystals may be beneficial.

[0032] A suitable melt modifier is barium oxide (BaO). However, barium oxide is unstable at normal atmospheric temperature and pressure. Specifically, at normal atmospheric temperature and pressure, barium oxide reacts with water (H2O) and carbon dioxide (CO2) in the atmosphere to form barium carbonate (BaCO3). When heat is applied to barium carbonate from the melt 102, barium carbonate decomposes into barium oxide and carbon dioxide. Therefore, barium carbonate (also referred to as the "melt modifier precursor") is introduced into the melt 102, decomposes within the melt 102 to release carbon dioxide (CO2), and provides the concentration of barium oxide (BaO) melt modifier within the melt 102.

[0033] Alternatively, barium oxide may be maintained in an unreacted, pure state. For example, the barium oxide can be stored and maintained in a bottle that also contains an inert gas therein to minimize exposure to moisture, and when the barium oxide is transferred to the silicon melt 102, procedures can be followed to minimize contact between the barium oxide and air. In further embodiments, other suitable melt modifier precursors may be used. Suitable melt modifier precursors can include any compound containing suitable ions such as barium, calcium, strontium, magnesium, etc., and suitable counter molecules that decompose the ions and feed them into the melt 102, enabling interaction between the ions and the glass lining surface 149 (shown in FIG. 3) and mediating the nucleation of cristobalite on the lining surface 149. Suitable counter molecules can include, for example, but are not limited to, oxygen, carbonates, oxalates, fluorides, etc. Examples of suitable precursors include, but are not limited to, strontium carbonate (SrCO3), calcium carbonate (CaCO3), and / or magnesium carbonate (MgCO3).

[0034] The liner has an initial thickness T1 that extends from the inner surface 148 of the quartz body 140 of the crucible 104 to the lining surface 149 disposed to contact the melt 102. The initial thickness T1 of the liner is selected based on the desired crystallization thickness T2 of the crystallization layer 154 (shown in FIG. 4) after the liner has devitrified. For example, in this embodiment, the initial thickness T1 is less than 500 microns (μm), and the resulting crystallization thickness T2 is less than 700 microns (μm). The concentration of the melt modifier in the doped melt 102 reacts with the synthetic liner 158 to promote devitrification of the synthetic liner 158 and produce a thin devitrified layer as shown in FIG. 4.

[0035] Referring to FIG. 4, the liner, after being exposed to the doped melt 102, is shown as a crystallized layer or a "devitrified" layer (these terms are used interchangeably herein). In particular, the silica liner devitrifies to cristobalite in this example by being exposed to the modified melt 102. The crystallization layer 154 changes the equilibrium at the melt 102 / crucible 104 interface (i.e., the "cristobalite / silicon / glass phase interface"), avoids bubble nucleation, reduces the vibration of the silicon melt 102, and improves the structure of the resulting ingot.

[0036] The crystallization layer 154 extends radially inward from the inner surface 148 of the quartz crucible 104 body 140 to the crystallization surface 162 that contacts the melt 102. The doped melt 102 not only promotes the devitrification of the liner 158 but also reduces the surface roughness of the crystallization surface 162 compared to the unmodified or "undoped" melt 102. This will be described in more detail with reference to FIGS. 5 and 6. The crystallization layer 154 defines a crystallization thickness T2 between the inner surface 148 of the body 140 and the lining surface 149. Due to the devitrification of the liner, the crystallization thickness T2 increases compared to the starting thickness T1 of the liner. The crystallization thickness T2 is 700 microns (μm) or less, 500 microns (μm) or less, or 100 microns (μm) or less. The crystallization thickness T2 is preferably thick enough so that the starting synthesis liner 158 remains continuous during the arc melting process. In some embodiments, the crystallization thickness T2 may be 50 microns (μm).

[0037] By having the crystallization thickness T2 of 700 microns or less, the resistance of the crystallization layer 154 to thermal exfoliation is improved, at least in part, compared to a crystallization liner having a thickness of 700 microns or more. For example, in experimental tests, keeping the thickness T2 of the crystallization layer 154 below 700 microns results in complete crystallization at the bottom, while using a crucible with a crystallization thickness T2 exceeding 700 microns increases the likelihood of thermal exfoliation, and it was shown that the dislocation structure of the crystal is lost between about 65 percent and 100 percent of the intended length of the crystal 140.

[0038] Figures 5 and 6 show the results of an exemplary experiment demonstrating that the surface roughness of the crystallization layer 154 is reduced when the doped melt 102 is used as compared to the undoped melt 102. It is desirable that the surface of the crystallization layer 154 be smoother because as the roughness of the crystallization layer 154 increases, the instability at the boundary between the melt 102 and the crucible 104 increases, which may damage the resulting ingot.

[0039] For example, the surface roughness of the crystallization surface 162 is correlated with the heat transfer coefficient of the crystallization layer 154. Further, the heat transfer coefficient of the crystallization layer 154 is correlated with bubble nucleation at the melt 102 / crucible 104 boundary. In other words, as the surface roughness of the crystallization layer 154 increases, the nucleation of bubbles at the boundary between the melt 102 and the crucible 104 also increases. Thus, it can be understood that the surface roughness of the crystallization layer 154 directly affects the nucleation potential of the crystallization surface 162, and for a smoother or less rough surface, the nucleation potential of the gas nucleating at the melt 102 / crucible 104 boundary is reduced.

[0040] Due to bubble nucleation, gas bubbles may be released into the melt 102, causing the melt 102 to vibrate and potentially damaging the structure of the ingot. Further, bubble nucleation at the melt 102 / crucible 104 boundary may create fine voids within the silicon melt 102, which may remain in the grown ingot when the crystalline ingot is formed. The presence of these voids (commonly also referred to as "air pockets" or "pinholes") in the grown ingot may weaken the structure of the resulting ingot and make the ingot unsuitable for use in electronic devices.

[0041] In the experiment, quartz samples with surfaces lined with synthetic silica having a thickness of 2 millimeters (mm) or more were each immersed in a silicon melt 102 at a temperature of 1525 °C for 2 hours. In the control test, the melt 102 was not doped, meaning that the melt 102 contained silicon and no nucleating agent melt modifier was used. In the second test, the melt 102 was doped with approximately 0.013 milligrams of Ba per kilogram of silicon. An undoped quartz sample was produced in the control test, and a doped quartz sample was produced in the second test. The samples were each removed from the melt 102 after 2 hours of immersion, photographed and measured, and the surface roughness between the melts was compared.

[0042] Referring to FIG. 5, the first chart shown at 502 shows a comparison of the root mean square (RMS) deviation in microns between the doped sample (indicated by BaCO3) and the undoped sample. The second graph shown at 504 shows a comparison of the area ratio of the doped sample and the undoped sample. The area ratio was determined by measuring the area of the peaks and valleys of the crystallized surface and normalizing the measured values to the area of the plan view.

[0043] As shown in the first graph, the doped sample had a reduced RMS deviation compared to the doped sample. Specifically, the RMS deviation of the doped sample was about 1 micron, and the detected deviation was in the range of 0.8 micron or more and 1.2 microns or less. In comparison, the RMS deviation of the undoped sample was about 1.24 microns, and the range of the deviation was about 1.15 or more and about 1.35 or less. Furthermore, as shown in the second graph, the doped sample also had a reduced area ratio compared to the undoped sample. In particular, the area ratio of the doped sample was about 1.0575, while the area ratio of the undoped sample was about 1.08. Therefore, it was found that in the undoped sample, the difference between the valleys and peaks of the surface was larger compared to the doped sample.

[0044] Figure 6 shows micrographs of the undoped sample 602 and the doped sample 604 after removal from the melt 102. The doped sample 602 was clearly less rough than the undoped sample 604. For example, as shown in the photograph, in the photograph of the undoped sample, on the crystallized surface 162, a more intense swirling pattern is seen, such as shown at 603a and 603b, which coincides with steeper mountains and valleys. In contrast, the image of the doped sample shows a much more monotonic and consistent crystallized surface 162, indicating a reduction in surface roughness.

[0045] Figure 7 is a flowchart of a method 700 for manufacturing a single crystal silicon ingot from a silicon melt 102 held in a crucible 104 disposed within an ingot pulling apparatus. The method includes a step 702 of providing a crucible 104 within an internal chamber of the ingot pulling apparatus, the crucible 104 including an inner surface 148 and a synthetic liner 158 on the inner surface 148. The method further includes adding 704 an initial fill of polysilicon to the crucible 104. The method also includes a step 706 of melting the initial fill of polysilicon to form a silicon melt 102 within the crucible 104. The method further includes dissolving 708 a melt modifier in the silicon melt 102 to devitrify the synthetic liner 158 and form a crystallization layer 154 on the crucible 104, the thickness T2 of the crystallization layer 154 being less than 700 microns. The method also includes a step 710 of pulling a single crystal silicon ingot from the silicon melt 102.

[0046] As described herein, advantages of using a synthetic liner 158 with a reduced thickness in combination with a melt 102 to which a high concentration of nucleating agent has been added include an improved resistance to thermal delamination of the crystallization layer 154. As a result, the crucible 104 is exposed to the silicon melt 102 for a longer time or the high-temperature time is lengthened, thereby reducing costs and improving throughput and yield. As throughput and yield increase, the ingot is produced more efficiently and at a lower cost. Further advantages include a reduced surface roughness of the crystallization surface 162, a reduced bubble nucleation in the melt 102, thereby improving the resulting structure and reducing defects in the grown crystal ingot.

[0047] As used herein, the terms "about", "substantially", "essentially" and "approximately" when used in connection with a range of dimensions, concentrations, temperatures or other physical or chemical properties or characteristics are intended to cover variations that may exist at the upper and / or lower limits of the range of the property or characteristic, including for example variations resulting from rounding, measurement methods or other statistical variations.

[0048] When introducing elements of the present disclosure or embodiments thereof, the articles "a", "an", "the" and "said" shall be taken to mean that there is one or more of the elements. The terms "comprising", "including", "containing" and "having" are intended to be inclusive and mean that there may be additional elements other than the elements described. The use of terms indicating a particular direction (such as "up", "down", "sideways", etc.) is for convenience of explanation and does not require a particular direction of the item being described.

[0049] Since the above-described configurations and methods can be variously modified without departing from the scope of the present invention, all matters included in the above description and shown in the accompanying drawings should be construed as illustrative and not in a limiting sense.

Images