WO2023125206A1 - 单晶体的制备方法及硅晶体 - Google Patents

单晶体的制备方法及硅晶体 Download PDF

Info

Publication number
WO2023125206A1
WO2023125206A1 PCT/CN2022/140778 CN2022140778W WO2023125206A1 WO 2023125206 A1 WO2023125206 A1 WO 2023125206A1 CN 2022140778 W CN2022140778 W CN 2022140778W WO 2023125206 A1 WO2023125206 A1 WO 2023125206A1
Authority
WO
WIPO (PCT)
Prior art keywords
stage
temperature
solid
liquid interface
ingot
Prior art date
Application number
PCT/CN2022/140778
Other languages
English (en)
French (fr)
Inventor
王双丽
陈俊宏
Original Assignee
中环领先半导体材料有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN202111619092.8A external-priority patent/CN114481302B/zh
Priority claimed from CN202111619103.2A external-priority patent/CN114438585A/zh
Application filed by 中环领先半导体材料有限公司 filed Critical 中环领先半导体材料有限公司
Publication of WO2023125206A1 publication Critical patent/WO2023125206A1/zh

Links

Images

Classifications

    • CCHEMISTRY; METALLURGY
    • C30CRYSTAL GROWTH
    • C30BSINGLE-CRYSTAL GROWTH; UNIDIRECTIONAL SOLIDIFICATION OF EUTECTIC MATERIAL OR UNIDIRECTIONAL DEMIXING OF EUTECTOID MATERIAL; REFINING BY ZONE-MELTING OF MATERIAL; PRODUCTION OF A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; SINGLE CRYSTALS OR HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; AFTER-TREATMENT OF SINGLE CRYSTALS OR A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; APPARATUS THEREFOR
    • C30B15/00Single-crystal growth by pulling from a melt, e.g. Czochralski method
    • C30B15/20Controlling or regulating
    • CCHEMISTRY; METALLURGY
    • C30CRYSTAL GROWTH
    • C30BSINGLE-CRYSTAL GROWTH; UNIDIRECTIONAL SOLIDIFICATION OF EUTECTIC MATERIAL OR UNIDIRECTIONAL DEMIXING OF EUTECTOID MATERIAL; REFINING BY ZONE-MELTING OF MATERIAL; PRODUCTION OF A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; SINGLE CRYSTALS OR HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; AFTER-TREATMENT OF SINGLE CRYSTALS OR A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; APPARATUS THEREFOR
    • C30B29/00Single crystals or homogeneous polycrystalline material with defined structure characterised by the material or by their shape
    • C30B29/02Elements
    • C30B29/06Silicon

Definitions

  • the disclosure belongs to the field of semiconductors, and in particular relates to a method for preparing a single crystal and a silicon crystal.
  • Voronkov proposed the V/G theory specifically the ratio of the pulling speed to the temperature gradient G in the pulling direction near the crystal growth interface, that is, there is a critical value of V/G in the CZ crystal growth process.
  • the V/G ratio is smaller than the critical value, the crystal lattice
  • the concentration of interstitial point defects (type I point defects) is higher than that of void type point defects (type V point defects), and there are too many type I point defects in the crystal, which is called I-type silicon crystal
  • the V/G ratio is greater than the critical
  • the concentration of lattice gap type point defects (I-type point defects) is lower than that of void type point defects (V-type point defects), and there are too many V-type point defects in the crystal, which is called V-type silicon crystal
  • the concentration of I-type point defects and V-type point defects remaining in the crystal is very small, and the difference is not large, and the crystal with few point defects grows, that is,
  • V/G value it is very difficult to strictly control the V/G value to be at a critical value, that is, to control the concentration of type I point defects and type V point defects at the solid-liquid interface of the crystal growth to be equal.
  • an object of the present disclosure is to provide a method for preparing a single crystal and a silicon crystal.
  • the method allows the point defects in the crystal to fully convect and diffuse by controlling the axial temperature gradient of the crystal rod in the region from the solid-liquid interface to the reference plane. and recombination, reducing the concentration of free type V point defects and type I point defects.
  • the present disclosure provides a method for preparing a single crystal. According to an embodiment of the present disclosure, the method includes:
  • the crystal preparation method of the embodiment of the present disclosure by adjusting the axial temperature gradient of the crystal rod in the region from the solid-liquid interface to the reference plane, the I-type point defects and V-type point defects formed during the crystal growth process are in the vicinity of the solid-liquid interface. Sufficient diffusion and recombination, thereby reducing the formation concentration of free point defects.
  • the method for preparing a single crystal according to the above-mentioned embodiments of the present disclosure may also have the following additional technical features:
  • ⁇ G c is controlled to be 0.2-1K/cm, ⁇ G e is 5-10K/cm, and G r is not greater than 6K/cm;
  • the method further includes adjusting the width of the temperature zone on the ingot, the temperature zone includes a first temperature zone, a second temperature zone and a third temperature zone, the first temperature zone
  • the range is 1685K to 1605K
  • the range of the second temperature range is 1605K to 1355K
  • the range of the third temperature range is 1355K to 955K.
  • the I-type point defects and V-type point defects formed during the crystal growth process can diffuse and recombine at the solid-liquid interface, reducing the formation of point defects concentration, controls the size of micro-defects, and increases the width of the perfect crystal window.
  • this method is especially suitable for the growth of large-diameter ingots, and can also improve the yield of perfect crystals in large-diameter ingots.
  • the pulling speed of the ingot is v 1
  • the distance between liquid ports is d 1
  • the A 1 , the B 1 , the k 1 , the v 1 and the d 1 satisfy: 0.03 ⁇ A 1 ⁇ 0.05, -0.3 ⁇ B 1 ⁇ -0.1, -0.12 ⁇ k 1 ⁇ -0.1, 0.4 ⁇ v 1 ⁇ 0.8mm/min, 50 ⁇ d 1 ⁇ 52mm;
  • the temperature gradient change ⁇ G c in the axial direction at the central position of the boundary layer is controlled to be 2-2 3K/cm, and the temperature gradient variation ⁇ G e in the axial direction of the edge position of the control boundary layer is 3.5 ⁇ 5.5K/cm;
  • the temperature gradient change ⁇ G c in the axial direction at the central position of the control boundary layer is 4-6 K/cm, and the temperature gradient change ⁇ G e in the axial direction at the edge position of the control boundary layer is not greater than 5.4K/cm;
  • the temperature gradient change ⁇ G c in the axial direction at the center of the boundary layer is controlled to be 2 to 3 K/cm, and the temperature gradient change ⁇ G e in the axial direction at the edge of the boundary layer is controlled to be 3.5 ⁇ 5.5K/cm;
  • the boundary layer is located in the region from the solid-liquid interface to the second reference plane.
  • the radial temperature gradient G r of the boundary layer is controlled to be no greater than 2K/cm;
  • the radial temperature gradient G r of the boundary layer is controlled to be not greater than 4K/cm;
  • the radial temperature gradient G r of the boundary layer is controlled to be no greater than 2K/cm.
  • the area near the ingot is the central area between the wall of the ingot and the inner wall of the water cooling jacket.
  • the present disclosure provides a silicon crystal.
  • the silicon crystal is prepared by the above-mentioned method.
  • Fig. 1 is the curve diagram (from the solid-liquid interface to the first reference surface) of the axial temperature gradient G h of the crystal rod and the height h relative to the solid-liquid interface at the initial stage of the isodiametric growth stage (the length of the crystal rod is 400mm);
  • Fig. 2 is a curve diagram (from the solid-liquid interface to the first reference surface) of the axial temperature gradient G h of the crystal rod and the height h relative to the solid-liquid interface in the middle stage of the equal-diameter growth stage (the length of the crystal rod is 700 mm);
  • Fig. 3 is the curve diagram (from the solid-liquid interface to the first reference surface) of the axial temperature gradient G h of the crystal rod and the height h relative to the solid-liquid interface in the late stage of the equal-diameter growth stage (the length of the crystal rod is 1000 mm);
  • Fig. 4 is a curve diagram (from the solid-liquid interface to the first reference surface) of the axial temperature gradient G h of the crystal rod and the height h relative to the solid-liquid interface in the late stage of the equal-diameter growth stage (the length of the crystal rod is 1300 mm);
  • Fig. 5 is a graph of the temperature band width in the initial stage and middle stage of the isometric growth stage in the 1685K ⁇ 1665K temperature band in the first temperature band;
  • Fig. 6 is a curve diagram of the temperature band width in the later period of the equal-diameter growth stage in the 1685K ⁇ 1665K temperature band of the first temperature band;
  • Fig. 7 is the temperature band width curve diagram of the initial stage and the middle stage of the equal-diameter growth stage in the 1665K ⁇ 1645K temperature band of the first temperature band;
  • Fig. 8 is a curve diagram of the temperature band width in the later period of the equal-diameter growth stage in the 1665K ⁇ 1645K temperature band of the first temperature band;
  • Fig. 9 is a graph showing the temperature band width curves in the initial stage and middle stage of the equal-diameter growth stage in the 1645K ⁇ 1625K temperature band in the first temperature band;
  • Fig. 10 is a graph of the temperature band width in the later stage of the equal-diameter growth stage in the 1645K-1625K temperature band of the first temperature band;
  • Fig. 11 is a temperature band width curve diagram of the initial stage and middle stage of the equal-diameter growth stage in the 1625K ⁇ 1605K temperature band of the first temperature band;
  • Fig. 12 is a curve diagram of the temperature band width in the later period of the equal-diameter growth stage in the 1625K-1605K temperature band of the first temperature band;
  • Fig. 13 is a schematic diagram of solid-liquid interface and boundary layer in the crystal growth process
  • Figure 14 is the initial stage of the equal-diameter growth stage (the length of the crystal rod is 400 mm), from the solid-liquid interface to the second reference plane, the axial temperature gradient G H near the solid-liquid interface and the height H relative to the solid-liquid interface;
  • Fig. 15 is the mid-stage of the equal-diameter growth stage (the length of the crystal rod is 700 mm), from the solid-liquid interface to the second reference plane, the axial temperature gradient G H near the solid-liquid interface and the height H relative to the solid-liquid interface;
  • Fig. 16 is a late stage of the equal-diameter growth stage (the length of the crystal rod is 1000 mm), from the solid-liquid interface to the second reference surface, the axial temperature gradient G H near the solid-liquid interface and the height H curve relative to the solid-liquid interface;
  • Fig. 17 is a late stage of the equal-diameter growth stage (ingot length 1300mm), from the solid-liquid interface to the second reference plane, the axial temperature gradient G H near the solid-liquid interface and the height H relative to the solid-liquid interface;
  • Fig. 18 is the initial stage of equal-diameter growth stage (ingot length 400mm), the temperature T L of the area near the ingot at the distance L from the solid-liquid interface and the curve diagram of the height L relative to the solid-liquid interface;
  • Fig. 19 is the mid-stage of the equal-diameter growth stage (the length of the crystal rod is 700mm), the temperature T L of the area near the crystal rod at a distance L from the solid-liquid interface and the curve diagram of the height L relative to the solid-liquid interface;
  • Fig. 20 is the late stage of the equal-diameter growth stage (the length of the crystal rod is 1000mm), the temperature T L of the area near the crystal rod at the distance L from the solid-liquid interface and the curve diagram of the height L relative to the solid-liquid interface;
  • Fig. 21 is the late stage of the equal-diameter growth stage (the length of the crystal rod is 1300mm), the temperature T L of the area near the crystal rod at a distance L from the solid-liquid interface and the curve diagram of the height L relative to the solid-liquid interface;
  • Fig. 22 is a schematic diagram of the specific operation of the copper decoration characterization method.
  • the liquid opening distance is the distance between the lower end of the draft tube and the solid-liquid interface.
  • the term "perfect crystal” does not mean an absolutely perfect crystal or a crystal without any defects, but allows for the existence of a small amount of one or more defects, which is not enough to make a crystal or the resulting wafer. A large change in electrical or mechanical properties can lead to deterioration of the performance of electronic devices made of it.
  • the Czochralski method to prepare crystals is to put the raw material polysilicon block into a quartz crucible, then heat the quartz crucible, melt the polysilicon in the quartz crucible, and then insert the seed crystal into the surface of the melt for welding, while rotating the seed crystal, and then reverse
  • the crucible is turned, the seed crystal is slowly lifted upwards, and the ingot is prepared through processes such as seeding, amplification, shoulder rotation, equal-diameter growth, and finishing.
  • the heat history in the crystal rod is directly related to the distribution, type and size of crystal defects, and this application is to make the point defects in the crystal fully convective diffusion and recombination, and reduce the number of free V-type points. Defect and type I point defect concentrations.
  • the present disclosure proposes a method for preparing a single crystal.
  • the I-type point defects and V-type point defects formed during the crystal growth process are fully diffused and recombined in the region near the solid-liquid interface, Thereby reducing the formation concentration of point defects.
  • the present application in order to precisely control the temperature gradient of the ingot in the region from the solid-liquid interface to the first reference plane, separately controls the initial stage of the equal diameter growth stage, the middle stage of the equal diameter growth stage, and the temperature gradient of the equal diameter growth stage. In the later stage, the corresponding temperature gradient is optimized.
  • the "initial stage of isodiametric growth stage” is defined as the stage in which the length of the grown crystal is not greater than 400mm;
  • the “late stage of the equal diameter growth stage” is defined as the stage after the liquid level of the silicon melt in the crucible enters the R angle of the crucible.
  • the distance between liquid ports is d 1 , wherein, the above k 1 , v 1 and d 1 satisfy: -0.12 ⁇ k 1 ⁇ -0.1, 0.4 ⁇ v 1 ⁇ 0.8mm/min, 50 ⁇ d 1 ⁇ 52mm, that is, according to the linear
  • G h k*h+G 0
  • the actual value of h1 is equal to the theoretical value of G h1 .
  • the adjustment of the temperature gradient G h in each stage is realized, so that the I-type points formed during the growth process Defects and V-type point defects fully diffuse and recombine in the region near the solid-liquid interface, reducing the concentration of free V-type point defects and I-type point defects.
  • the variation range of the pulling speed in the crystal growth stage can be increased by 10%, and the variation range of the liquid gap can be increased by 5%, thereby expanding the adjustable window of process parameters and improving the crystal quality and the yield of perfect crystals.
  • the yield is the ratio of the amount of molten material to grow a perfect crystal to the amount of input material.
  • the axial temperature gradient of the ingot at different growth stages in the region from the solid-liquid interface to the first reference plane is plotted relative to the height of the solid-liquid interface, as shown in
  • the height h relative to the crystal growth interface is taken as the abscissa, and the temperature gradient G h in the axial direction of the ingot is used as the ordinate to draw the curve.
  • the fitting degree of each of the above linear equations is greater than 0.998, indicating that the fitted linear equations are very consistent with the actual situation. Therefore, according to the linear equations, adjusting the pulling speed and liquid opening distance can accurately control the solid-liquid interface.
  • the axial temperature gradient in the region to the first reference plane effectively reduces the formation concentration of point defects.
  • the temperature gradient change ⁇ G c in the axial direction at the center of the ingot is controlled to be 0.2-1K/cm, Control the temperature gradient change amount ⁇ G e in the axial direction of the edge position of the ingot to be 5-10K/cm, and control the radial temperature gradient G r of the ingot to not exceed 6K/cm;
  • the temperature gradient change in the axial direction of the center position ⁇ G c is 2-6K/cm, the temperature gradient change in the axial direction of the edge position of the control crystal rod is 5-10K/cm, and the radial temperature of the crystal rod is controlled
  • the gradient G r is not greater than 10K/cm; in the later stage of the equal-diameter growth stage, the temperature gradient change ⁇ G c in the axial direction of the central position of the crystal rod is controlled to be 0.2-1K/c
  • the shape of the solid-liquid interface is similar to a plane, so that the axial temperature in the same section
  • the gradient is approximately uniform, and according to the V/G theory, the window for growing perfect crystals can be enlarged, and the yield of perfect crystals can be improved.
  • the above method of the present application further includes adjusting the width of the 1685K-955K temperature zone on the ingot, wherein the temperature zone includes the first temperature zone, the second temperature zone and the third temperature zone, and the first temperature zone
  • the range is 1685K to 1605K
  • the second temperature zone ranges from 1605K to 1355K
  • the third temperature zone ranges from 1355K to 955K.
  • the first temperature zone includes 1685K-1665K, 1665K-1645K, 1645K-1625K, and 1625K-1605K temperature zone
  • the above-mentioned first temperature zone includes several small temperature zones, and each of the above-mentioned small temperature zones
  • the difference in width is less than 1.0 mm, and the change in width of the band at the same temperature during the growth process is less than 0.5 mm.
  • the width of the first temperature zone is adjusted according to the above linear equation, so that at the same period of the equal-diameter growth stage, the difference between the widths of each small temperature zone in the first temperature zone is less than 1.0 mm, that is to say, the width of each small temperature zone Relatively similar, at the same time, in each period of the equal-diameter growth stage, the difference in width of each small temperature zone is less than 0.5mm, that is to say, the change in width of each small temperature zone is small, and the first temperature zone on the ingot can be precisely controlled width position.
  • the temperature gradient G0 at the solid-liquid interface can be adjusted by adjusting the height of the guide tube from the solid-liquid interface, and then the adjustment of D1 can be realized.
  • the increase of the height H of the guide cylinder from the solid-liquid interface is realized by the rotation of the crucible shaft to lower the quartz crucible, and the quartz crucible is closer to the heater, so that G 0 decreases, and then the temperature band width D 1 increase; similarly, the height H of the guide tube to the solid-liquid interface decreases because the rotation of the crucible axis makes the quartz crucible rise, and the quartz crucible is farther away from the heater, so that G 0 increases, and the temperature band width D 1 decreases.
  • the temperature gradient G0 at the solid-liquid interface can be adjusted by adjusting the height H of the guide cylinder from the solid-liquid interface, and then the adjustment of D2 can be realized.
  • the increase of the height H of the guide tube from the solid-liquid interface is due to the rotation of the crucible shaft to make the quartz crucible drop, and the quartz crucible is closer to the heater, so that G 0 decreases, and then the temperature band width D 2 increases;
  • the reduction of the height H of the guide tube from the solid-liquid interface is due to the rotation of the crucible shaft to make the quartz crucible rise, and the quartz crucible is farther away from the heater, so that G 0 increases, and the temperature band width D 2 decreases.
  • the temperature band width D3 in the middle stage is between the temperature band width D1 in the early stage and the temperature band width D2 in the later stage.
  • the temperature band width D 3 in the middle stage is equal to the temperature band width D 1 in the early stage.
  • the temperature band width D3 is equal to the later temperature band width D2 , and the temperature band width D3 changes between D1 and D2 in the middle period.
  • the width of the first temperature zone is adjusted in time, so that it is finally close to the theoretical value, and the temperature distribution of the ingot is effectively controlled.
  • the temperature gradient G 0 at the solid-liquid interface is taken as the abscissa, and the temperature band width D x is taken as the ordinate to draw a curve.
  • the temperature gradient G 0 at the solid-liquid interface is taken as the abscissa
  • the temperature band width D x is taken as the ordinate.
  • the temperature gradient G 0 at the solid-liquid interface is taken as the abscissa
  • the temperature band width D x is taken as the ordinate.
  • the temperature gradient G 0 at the solid-liquid interface is taken as the abscissa
  • the temperature band width D x is taken as the ordinate.
  • the fitting degrees of the linear equations of each small temperature zone width in the first temperature zone are all greater than 0.95, indicating that the fitted linear equations are very consistent with the actual situation. Therefore, according to the linear equation of the first temperature zone, it is instructive to adjust the height of the guide tube from the solid-liquid interface, which can accurately adjust the width of the first temperature zone, control the thermal history of the ingot and suppress the growth of point defects, and control the defect. size, increasing the yield of growing perfect crystals.
  • the second temperature zone includes 1605K-1555K, 1555K-1505K, 1505K-1455K, 1455K-1405K, and 1405K-1355K temperature zone, and the above-mentioned second temperature zone includes a plurality of small temperature zones, and each The difference in the width of the small temperature zone is less than 1.5mm, and the change in the width of the same temperature zone during the growth process is less than 1.0mm.
  • the width of the second temperature zone is adjusted according to the above linear equation, so that at the same period of the equal-diameter growth stage, the difference between the widths of each small temperature zone in the second temperature zone is less than 1.5 mm, that is to say, the width of each small temperature zone Almost the same, at the same time, in each period of the equal-diameter growth stage, the difference in the width of each small temperature zone is less than 1.0 mm, that is to say, the change in the width of each small temperature zone is small, and the width of the first temperature zone on the ingot can be precisely controlled s position.
  • the temperature band width D5 becomes smaller and closer to the theoretical value; In the middle stage of the radial growth stage, the temperature band width D6 in the middle stage is between the initial temperature band width D4 and the later temperature band width D5 .
  • the temperature band width D 6 in the middle stage is equal to the temperature band width D 4 in the early stage, and the moment of transition from the middle stage to the late stage, At this time, the temperature band width D 6 in the middle period is equal to the temperature band width D 5 in the later period, and the temperature band width D 6 changes between D 4 and D 5 during the middle term.
  • the fitting degree of the linear equations of each small temperature band width in the second temperature band is greater than 0.95, and will not be repeated here. Therefore, according to the linear equation of the second temperature zone, it is instructive to adjust the distance between the water cooling jacket and the outer wall of the ingot or the cooling water flow in the water cooling jacket, which can accurately adjust the width of the second temperature zone and control the thermal history of the ingot and the size of defects. , improving the yield of growing perfect crystals.
  • the temperature band width D 8 is equal to the temperature band width D 7 in the early stage at the moment from the early stage to the middle stage, and the temperature band width D 8 remains unchanged in the middle and late stages.
  • the fitting degree of the linear equation of the third temperature zone is also greater than 0.95, which will not be repeated here. Therefore, according to the linear equation of the third temperature zone, it is instructive to adjust the cooling water flow in the water cooling jacket, which can accurately adjust the width of the third temperature zone, control the thermal history of the ingot and the size of the defect, and improve the yield of perfect crystal growth.
  • the boundary layer when the liquid phase changes to the solid phase and grows crystals, there is a boundary layer between the solid-liquid interface and the melt, and the boundary layer has a transport effect on the interface.
  • the temperature distribution of the boundary layer determines the size of the crystal growth driving force and
  • the boundary layer is in the region from the solid-liquid interface to the second reference plane (the second reference plane is the interface 10 mm above the solid-liquid interface).
  • Voronkov's theory points out that lattice interstitial point defects (type I point defects) and void type point defects (type V point defects) are formed near the solid-liquid interface of long crystals, and point defects undergo convective diffusion and recombination in the boundary layer.
  • the thermal field structure of the boundary layer is determined by the crystal growth rate and the liquid port distance. Therefore, for different thermal field conditions, the growth rate and the liquid port distance can be adjusted at the same time to realize or optimize the thermal field structure to achieve boundary
  • the temperature gradient distribution of the layers meets the requirements for preparing perfect crystals. And at present, it is difficult to grow perfect crystals from large-sized crystal rods (such as 450mm and above), because the size of the crystal rods is relatively large.
  • the inventors found that the temperature distribution in the vicinity of the ingot and the temperature gradient distribution near the solid-liquid interface have a direct impact on the growth of large-sized ingots.
  • the inventor proposes adjusting the region of the solid-liquid interface to the second reference plane (the second reference plane being the interface above the solid-liquid interface) while adjusting the temperature gradient in the region from the solid-liquid interface to the first reference plane temperature gradient inside.
  • the temperature gradient in the region from the solid-liquid interface to the first reference surface and the temperature gradient in the region from the solid-liquid interface to the second reference surface that is, the solid-liquid interface from the first reference surface to the second reference surface is precisely controlled.
  • the temperature gradient in the area of the two reference surfaces, and the temperature gradient in the area above and below the solid-liquid interface directly affects the formation concentration of point defects.
  • the application in order to precisely control the temperature gradients in the regions from the solid-liquid interface to the first reference surface and from the solid-liquid interface to the second reference surface, the application corresponds to the initial stage, middle stage, and late stage of the isometric growth stage.
  • the temperature gradient is optimized.
  • G H2 and G h2 are greater than the theoretical values, increase the liquid opening distance d 2 so that G 0 decreases, then G H2 and G h2 are close to the theoretical values, and at the same time in the above pulling speed range
  • G h3 and G H3 are respectively smaller than the theoretical values, and the distance d 3 is reduced to increase G 0 , then G h3 and G H3 are close to the theoretical values. Decrease the pulling speed v 3 , so that the point defects formed in the crystal have a certain time to fully diffuse and recombine, and then reduce the formation concentration of point defects.
  • the I-type point defects and V-type point defects formed in the crystal growth process can be Sufficient diffusion and recombination at the solid-liquid interface reduces the concentration of point defect formation, controls the size of micro-defects, and increases the width of the perfect crystal window.
  • this method is especially suitable for the growth of large-diameter ingots, and can also improve the yield of perfect crystals in large-diameter ingots.
  • the axial temperature gradient near the solid-liquid interface and the height relative to the solid-liquid interface were plotted at different growth stages from the solid-liquid interface to the second reference surface region, as shown in Figure 14
  • the height H relative to the solid-liquid interface is taken as the abscissa
  • the axial temperature near the solid-liquid interface is Gradient G H draws a curve for the ordinate
  • the fitting degree of this equation R 2 0.9915; length 700mm)
  • taking the axial temperature gradient G H near the solid-liquid interface as the ordinate to draw a curve
  • the equation The fitting degree R 2 0.9
  • the ingot length is 1000 mm
  • the ingot length is The equation for 1300mm
  • G H 0.0163H 2 -0.0924H+2.4074
  • the fitting degree of each of the above equations is greater than 0.99, indicating that the fitted equation is very consistent with the actual situation. Therefore, adjusting the pulling speed and liquid opening distance according to the equation can accurately control the solid-liquid interface to the first
  • the axial temperature gradient from the reference plane region and the solid-liquid interface to the second reference plane region can effectively reduce the concentration of point defects.
  • the value of the height H relative to the solid-liquid interface is a negative number, it means that it is located below the solid-liquid interface and its height from the solid-liquid interface is
  • the number means that it is located above the solid-liquid interface and its height from the solid-liquid interface is h.
  • the axial temperature gradient at the central position of the boundary layer is controlled The amount ⁇ G c is 2 ⁇ 3K/cm, the change amount ⁇ G e of the axial temperature gradient at the edge position of the control boundary layer is 3.5 ⁇ 5.5K/cm, and the radial temperature gradient G r of the control boundary layer is not greater than 2K/cm;
  • the axial temperature gradient change ⁇ G c at the center of the boundary layer is controlled to be 4-6K/cm, and the axial temperature gradient change ⁇ G e at the edge of the boundary layer is not greater than 5.4K/cm cm, the radial temperature gradient G r of the control boundary layer is not greater than 4K/cm; in the later stage of the isometric growth stage, the axial temperature gradient change ⁇ G c at the center of the boundary layer is controlled to be 4-6K/cm, and the axial temperature gradient change ⁇ G e at the edge of the boundary layer is not greater than 5.4K/cm cm, the radial temperature gradient G
  • the boundary layer is located in the region from the solid-liquid interface to the second reference surface, at different stages of the isometric growth stage, by controlling the axial temperature gradient change and the radial temperature gradient at the center and edge positions of the boundary layer, the below the solid-liquid interface
  • the uniform temperature gradient distribution in the region ensures that the shape of the solid-liquid interface is close to a plane, which can improve the yield of perfect crystals, wherein the boundary layer is located in the region from the solid-liquid interface to the second reference plane.
  • the temperature of the area near the rod, in K Therefore, by quantifying the temperature distribution in the vicinity of the crystal ingot, the thermal field distribution near the crystal ingot can be optimized, the growth of micro-defects can be suppressed, and the yield of perfect crystals can be improved.
  • the area near the crystal ingot is the central area between the ingot wall and the inner wall of the water cooling jacket.
  • T L g*L 3 +f*L 2 +q*L+T 0
  • g g 1
  • the ingot is adjusted according to T L1 Nearby gas flow rate (the gas flow range is 90 ⁇ 120sL/pm), where the value of g 1 is -8e -0.5 ⁇ -7e -0.5 , the value of f 1 is 0.0411 ⁇ 0.0477, and the value of q 1 is -10.309 ⁇ - 9.6535
  • T L g*L 3 +f*L 2 +q*L+T 0
  • g g 1
  • T L1 when the actual value of T L1 is greater than the theoretical value, increase the gas flow rate near the ingot so that T L1 is close to the theoretical value; on the contrary, when the actual value of T L1 is smaller than the theoretical value, reduce the gas flow near the ingot flow, making T L1 close to the theoretical value.
  • the temperature in the vicinity of the ingot at different crystal growth stages in the region from the solid-liquid interface to the second reference plane is plotted relative to the height of the solid-liquid interface, as shown in Figure 18,
  • the height L relative to the solid-liquid interface of the long crystal is taken as the abscissa
  • the temperature T L in the vicinity of the crystal rod is used as the ordinate to draw a curve.
  • the height L is the abscissa, and the temperature T L in the vicinity of the ingot is used as the ordinate to draw the curve.
  • the fitting degree of each of the above equations is greater than 0.99, indicating that the fitted equation is very consistent with the actual situation. Therefore, according to the equation to adjust the gas flow near the ingot, the temperature distribution in the vicinity of the ingot can be accurately controlled. This further affects the temperature zone distribution of the ingot, further suppresses the size of micro-defects, and improves the yield of perfect crystals.
  • the present disclosure provides a silicon crystal.
  • the above-mentioned silicon crystal is prepared by the above-mentioned method.
  • the silicon crystal has a higher quality. It can be understood that the proportion of windows belonging to perfect crystals in the silicon crystal is large.
  • the gradient ⁇ G c is 0.2-1K/cm
  • the temperature gradient ⁇ G e in the axial direction of the edge position of the crystal rod is controlled to be 5-10K/cm
  • the radial temperature gradient G r of the crystal rod is not greater than 6K/cm.
  • the ingot is pulled out, that is, the control of the temperature gradient and temperature zone is not performed compared with the process of the embodiment.
  • Example 1 and Comparative Example 1 the wafers at the ingot lengths of 180mm, 340mm, 650mm, 950mm, 1200mm and 1350mm were respectively intercepted, and the defects were characterized by the copper decoration method, and the defect characterization results As shown in Table 1, the ingots in the examples and comparative examples are silicon ingots.
  • the specific operation of the copper decoration method includes: firstly, the test piece is cleaned with tap water, and then the surface of the test piece is cleaned with a surfactant to remove surface particles on the surface of the test piece, and then the surface of the test piece is cleaned by chemical polishing.
  • the test piece is a silicon wafer. Micro-defects on wafers can be observed under a microscope due to the formation of many copper deposits and dislocations around the micro-defects, forming a larger area. The black part in the micro-defect map is the perfect area, and the white part is the defect area.
  • Table 1 The detection results of wafer defects at various positions of the ingot obtained in Example 1 and Comparative Example 1 using the above-mentioned copper decoration method
  • Example 1 Wafer position(mm) Example 1 Comparative example 1 180 Type I deficiency V-shaped defect 340 perfect crystal Type I deficiency 650 perfect crystal Type I deficiency 950 perfect crystal Type I deficiency 1200 perfect crystal Type I deficiency 1350 Type I deficiency V-shaped defect
  • Comparative Example 1 adopts the prior art method, and the defects of the grown ingot are V-type defects at the beginning, I-type defects in the middle, and V-type defects at the end, and no perfect crystals appear in the entire ingot. .
  • Example 1 adopts the preparation method of the present application, and the grown ingots have perfect crystals at lengths of 340mm, 650mm, 950mm, and 1200mm, and the windows are large (it can be seen directly from the copper decoration defect map) , according to the V/G theory, within a certain range of pulling speed (0.4 ⁇ v ⁇ 0.8mm/min), there are perfect crystals in the range of ingot length from 340mm to 1200mm. Then adopting the preparation method of the present application does improve the yield of perfect crystals.
  • G H A*H 2 +B*H+G 0
  • G h1 and G H1 adjust the pulling speed of the ingot to be v 1 and the distance between liquid ports to be d 1
  • a 1 , B 1 , k 1 , v 1 and d 1 satisfy: 0.03 ⁇ A 1 ⁇ 0.05, -0.3 ⁇ B 1 ⁇ -0.1, -0.12 ⁇ k 1 ⁇ -0.1, 0.4 ⁇ v 1 ⁇ 0.8mm/min, 50 ⁇ d 1 ⁇ 52mm, at the same time, control the center position of the boundary layer in the area from the solid-liquid interface to the second reference surface
  • the axial temperature gradient change ⁇ G c is 2 ⁇ 3K/cm
  • the axial temperature gradient change ⁇ G e at the edge position of the control boundary layer is 3.5 ⁇ 5.5
  • T L m*L 3 +n*L 2 +q*L+T 0
  • m m 1
  • Example 2 and Comparative Example 2 the wafers at the ingot lengths of 100mm, 400mm, 700mm, 1000mm, and 1300mm were respectively intercepted, and the defects were characterized by the copper decoration method.
  • the defect characterization results are shown in the table 2, wherein the ingots in the examples and comparative examples are silicon ingots.
  • Comparative Example 2 adopts the prior art method, and the defects of the grown ingot are V-type defects at the beginning, I-type defects in the middle, and V-type defects in the middle, and no perfect crystals appear in the entire ingot.
  • Example 2 adopts the preparation method of the present application, and the grown ingots have perfect crystals at lengths of 400mm, 700mm, 1000mm, and 1300mm, and the windows are large (it can be seen directly from the copper decoration defect map) , according to the V/G theory, within a certain range of pulling speed (0.4 ⁇ v ⁇ 0.8mm/min), there are perfect crystals in the range of ingot length from 400mm to 1300mm. Then adopting the preparation method of the present application does improve the yield of perfect crystals.

Abstract

一种单品体的制备方法及硅晶体,该制备方法包括:在等径生长阶段,根据线性方程G h=k*h+G 0得到G h,其中,G 0为固液界面处温度梯度,取值为35-55K/cm, k为常数,h为相对于固液界面高度,取值为0-10mm, G h为距离固液界面高度h处品棒轴向温度梯度,单位为K/cm:在固液界面至第一参考面的区域内调整温度梯度至G h,其中,所述第一参考面为固液界面上方10mm的界面。该方法通过调控固液界面至第一参考面的区域内晶棒轴向温度梯度,使晶体中点缺陷充分地对流扩散和再复合,减少游离V型点缺陷和Ⅰ型点缺陷浓度。

Description

单晶体的制备方法及硅晶体
优先权信息
本公开请求于2021年12月27日向中国国家知识产权局提交的、专利申请号为202111619092.8,申请名称均为“单晶体的制备方法和硅晶体”以及申请号为202111619103.2,申请名称均为“单晶体的制备方法及硅晶体”的中国专利申请的优先权,并且其全部内容通过引用结合在本公开中。
技术领域
本公开属于半导体领域,具体涉及一种单晶体的制备方法及硅晶体。
背景技术
Voronkov提出V/G理论,具体为提拉速度与长晶界面附近提拉方向的温度梯度G比值即V/G在CZ长晶过程存在临界值,当V/G比值小于临界值时,晶格间隙型点缺陷(I型点缺陷)浓度高于空孔型点缺陷(V型点缺陷),晶体中残留过多I型点缺陷,称为I-型硅晶体;当V/G比值大于临界值时,晶格间隙型点缺陷(I型点缺陷)浓度低于空孔型点缺陷(V型点缺陷),晶体中则残留过多V型点缺陷,称为V-型硅晶体;当V/G比值等于临界值时,残留在晶体中的I型点缺陷和V型点缺陷浓度很小,相差也不大,生长为点缺陷很少的晶体,即完美晶体。
但严格控制V/G值处于临界值,即控制长晶固液界面处I型点缺陷和V型点缺陷浓度相当,其难度较高。
因此,现有制备完美晶体的技术有待探究。
公开内容
本公开旨在至少在一定程度上解决相关技术中的技术问题之一。为此,本公开的一个目的在于提出一种单晶体的制备方法及硅晶体,该方法通过调控固液界面至参考面的区域内晶棒轴向温度梯度,使晶体中的点缺陷充分地对流扩散和再复合,减少游离V型点缺陷和I型点缺陷的浓度。
在本公开的一个方面,本公开提出了一种单晶体的制备方法。根据本公开的实施例,所述方法包括:
在等径生长阶段,根据线性方程G h=k*h+G 0得到G h,其中,G 0为固液界面处温度梯度,取值为35~55K/cm,k为常数,h为相对于固液界面高度,取值为0~10mm,G h为距离固液界面高度h处晶棒轴向温度梯度,单位为K/cm;
在固液界面至第一参考面的区域内调整温度梯度至G h,其中,所述第一参考面为固液界面上方10mm的界面。
根据本公开实施例的晶体的制备方法,通过调控固液界面至参考面的区域内晶棒轴向温度梯度,使长晶过程形成的I型点缺陷和V型点缺陷在固液界面附近区域充分扩散和再复合,从而减少游离的点缺陷的形成浓度。
另外,根据本公开上述实施例的单晶体的制备方法还可以具有如下附加的技术特征:
在本公开的一些实施例中,在等径生长阶段的初期,取k=k 1计算得到G h1,根据G h1调节晶棒的提拉速度为v 1、液口距为d 1,其中,所述k 1、所述v 1和所述d 1满足:-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm;在等径生长阶段的中期,取k=k 2计算得到G h2,根据G h2调节晶棒的提拉速度为v 2、液口距为d 2,其中,所述k 2、所述v 2和所述d 2满足:-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm;在等径生长阶段的后期,取k=k 3计算得到G h3,根据G h3调节晶棒的提拉速度为v 3、液口距为d 3,其中,所述k 3、所述v 3和所述d 3满足:-0.16≤k 3≤0.14, 0.6≤v 3≤0.8mm/min,54≤d 3≤55mm。
在本公开的一些实施例中,所述等径生长阶段的初期,控制△G c为0.2~1K/cm,△G e为5~10K/cm,G r不大于6K/cm;所述等径生长阶段的中期,控制△G c为2~6K/cm,△G e为5~10K/cm,G r不大于10K/cm;所述等径生长阶段的后期,控制△G c为0.2~1K/cm,△G e为5~10K/cm,G r不大于6K/cm,其中,晶棒中心位置的轴向温度梯度变化量为△G c,晶棒边缘位置的轴向温度梯度变化量为△G e,晶棒的径向温度梯度为G r
在本公开的一些实施例中,所述方法还包括调节晶棒上温度带的宽度,所述温度带包括第一温度带、第二温度带和第三温度带,所述第一温度带的范围为1685K至1605K,所述第二温度带的范围为1605K至1355K,所述第三温度带的范围为1355K至955K。
在本公开的一些实施例中,在等径生长阶段,所述第一温度带范围内,根据线性方程D x=m*G 0+b调整温度带宽度,其中G 0为固液界面处温度梯度,取值为35~55K/cm,D x为温度带宽度,单位为mm。
在本公开的一些实施例中,在等径生长阶段的初期,取m=m 1,b=b 1,计算得到D 1,根据D 1调节导流筒距固液界面的高度,其中-0.1≤m 1≤-0.09,9≤b 1≤9.1;在等径生长阶段的后期,取m=m 2,b=b 2,计算得到D 2,根据D 2调节导流筒距固液界面的高度,其中-0.09≤m 2≤-0.05,7≤b 2≤8.5;在等径生长阶段的中期,所述中期的温度带宽度D 3处于所述初期的温度带宽度D 1和所述后期的温度带宽度D 2之间。
在本公开的一些实施例中,在等径生长阶段,所述第二温度带范围内,根据线性方程D y=m*G 0+b调整温度带宽度,其中G 0为固液界面处温度梯度,取值为35~55K/cm,D y为温度带宽度,单位为mm。
在本公开的一些实施例中,在等径生长阶段的初期,取m=m 3,b=b 3,计算得到D 4,根据D 4调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.27≤m 3≤-0.24,23.5≤b 3≤25.5;在等径生长阶段的后期,取m=m 4,b=b 4,计算得到D 5,根据D 5调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.1≤m 4≤-0.06,16.5≤b 4≤17.5;在等径生长阶段的中期,所述中期的温度带宽度D 6处于所述初期的温度带宽度D 4和所述后期的温度带宽度D 5之间。
在本公开的一些实施例中,在等径生长阶段,所述第三温度带范围内,根据线性方程D z=n*L+c调整温度带宽度,D z为温度带宽度,单位为mm,L为晶棒长度,单位为mm。
在本公开的一些实施例中,在等径生长阶段的初期,取n=n 1,c=c 1,计算得到D 7,根据D 7调节水冷套内冷却水流量,其中-0.035≤n 1≤-0.015,35≤c 1≤45;在等径生长阶段的中期和后期,控制温度带宽度保持不变且所述中期的温度带宽度D 8等于所述初期的温度带宽度D 7
在本公开的一些实施例中,在等径生长阶段,根据方程G H=A*H 2+B*h+G 0得到G H,其中,A和B为常数,G 0为固液界面处温度梯度,取值为35~55K/cm,H为相对于固液界面高度,取值为-10~0mm,G H为距离固液界面高度H处边界层的轴向温度梯度,单位为K/cm;在固液界面至第二参考面的区域内调整温度梯度至G H,其中,所述第二参考面为固液界面下方10mm的界面。由此,通过量化控制固液界面上方区域和下方区域内轴向温度梯度,使长晶过程形成的I型点缺陷和V型点缺陷在固液界面处扩散和再复合,减小点缺陷形成浓度,控制微缺陷的尺寸,增加完美晶体窗口宽度。另外,该方法特别适合大直径晶棒的生长,也可提高大尺寸晶棒的完美晶体的良率。
在本公开的一些实施例中,所述等径生长阶段的初期,取A=A 1,B=B 1计算得到G H1,取k=k 1计算得到G h1,根据G H1和G h1调节晶棒的提拉速度为v 1、液口距为d 1,其中,所述A 1、所述B 1、所述k 1、所述v 1和所述d 1满足:0.03≤A 1≤0.05,-0.3≤B 1≤-0.1,-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm;
所述等径生长阶段的中期,取A=A 2,B=B 2计算得到G H2,取k=k 2计算得到G h2,根据 G H2和G h2调节晶棒的提拉速度为v 2、液口距为d 2,其中,所述A 2、所述B 2、所述k 2、所述v 2和所述d 2满足:0.01≤A 2≤0.03,-0.2≤B 2≤-0.1,-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm;
所述等径生长阶段的后期,取A=A 3,B=B 3计算得到G H3,取k=k 3计算得到G h3,根据G H3和G h3调节晶棒的提拉速度为v 3、液口距为d 3,其中,所述A 3、所述B 3、所述k 3、所述v 3和所述d 3满足:0.01≤A 3≤0.03,-0.2≤B 3≤-0.1,-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm。
在本公开的一些实施例中,在固液界面至第二参考面的区域内:所述等径生长阶段的初期,控制边界层的中心位置轴向方向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向方向温度梯度变化量△G e为3.5~5.5K/cm;
所述等径生长阶段的中期,控制边界层的中心位置轴向方向温度梯度变化量△G c为4~6K/cm,控制边界层的边缘位置轴向方向温度梯度变化量△G e不大于5.4K/cm;
所述等径生长阶段的后期,控制边界层的中心位置轴向方向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向方向温度梯度变化量△G e为3.5~5.5K/cm;
其中,所述边界层位于固液界面至第二参考面区域内。
在本公开的一些实施例中,在固液界面至第二参考面的区域内:所述等径生长阶段的初期,控制边界层的径向温度梯度G r不大于2K/cm;
所述等径生长阶段的中期,控制边界层的径向温度梯度G r不大于4K/cm;
所述等径生长阶段的后期,控制边界层的径向温度梯度G r不大于2K/cm。
在本公开的一些实施例中,根据方程T L=g*L 3+f*L 2+q*L+T 0计算得到T L,g、f和q为常数,T 0为三相点处的平均温度,单位为K,L为相对于固液界面高度,取值为0~300mm,T L为距离固液界面距离L处晶棒附近区域温度,单位为K,
其中,所述晶棒附近区域是晶棒壁与水冷套内壁之间的中心区域。
在本公开的一些实施例中,取g=g 1,f=f 1和q=q 1计算得到T L1,根据T L1调节晶棒附近气体流量,其中,g 1取值为-8e -0.5~-7e -0.5,f 1取值为0.0411~0.0477,q 1取值为-10.309~-9.6535。
在本公开的再一个方面,本公开提出了一种硅晶体。根据本公开的实施例,所述硅晶体采用上述的方法制备得到。
本公开的附加方面和优点将在下面的描述中部分给出,部分将从下面的描述中变得明显,或通过本公开的实践了解到。
附图说明
本公开的上述和/或附加的方面和优点从结合下面附图对实施例的描述中将变得明显和容易理解,其中:
图1是等径生长阶段初期(晶棒长度400mm)晶棒轴向温度梯度G h和相对于固液界面高度h的曲线图(在固液界面至第一参考面);
图2是等径生长阶段中期(晶棒长度700mm)晶棒轴向温度梯度G h和相对于固液界面高度h的曲线图(在固液界面至第一参考面);
图3是等径生长阶段后期(晶棒长度1000mm)晶棒轴向温度梯度G h和相对于固液界面高度h的曲线图(在固液界面至第一参考面);
图4是等径生长阶段后期(晶棒长度1300mm)晶棒轴向温度梯度G h和相对于固液界面高度h的曲线图(在固液界面至第一参考面);
图5是第一温度带的1685K~1665K温度带内等径生长阶段初期和中期的温度带宽度曲线图;
图6是第一温度带的1685K~1665K温度带内等径生长阶段后期的温度带宽度曲线图;
图7是第一温度带的1665K~1645K温度带内等径生长阶段初期和中期的温度带宽度曲 线图;
图8是第一温度带的1665K~1645K温度带内等径生长阶段后期的温度带宽度曲线图;
图9是第一温度带的1645K~1625K温度带内等径生长阶段初期和中期的温度带宽度曲线图;
图10是第一温度带的1645K~1625K温度带内等径生长阶段后期的温度带宽度曲线图;
图11是第一温度带的1625K~1605K温度带内等径生长阶段初期和中期的温度带宽度曲线图;
图12是第一温度带的1625K~1605K温度带内等径生长阶段后期的温度带宽度曲线图;
图13是长晶过程中固液界面、边界层的示意图;
图14是等径生长阶段的初期(晶棒长度400mm),在固液界面至第二参考面,固液界面附近轴向温度梯度G H和相对于固液界面高度H曲线图;
图15是等径生长阶段的中期(晶棒长度700mm),在固液界面至第二参考面,固液界面附近轴向温度梯度G H和相对于固液界面高度H曲线图;
图16是等径生长阶段的后期(晶棒长度1000mm),在固液界面至第二参考面,固液界面附近轴向温度梯度G H和相对于固液界面高度H曲线图;
图17是等径生长阶段的后期(晶棒长度1300mm),在固液界面至第二参考面,固液界面附近轴向温度梯度G H和相对于固液界面高度H曲线图;
图18是等径生长阶段的初期(晶棒长度400mm),距离固液界面距离L处晶棒附近区域温度T L和相对于固液界面高度L曲线图;
图19是等径生长阶段的中期(晶棒长度700mm),距离固液界面距离L处晶棒附近区域温度T L和相对于固液界面高度L曲线图;
图20是等径生长阶段的后期(晶棒长度1000mm),距离固液界面距离L处晶棒附近区域温度T L和相对于固液界面高度L曲线图;
图21是等径生长阶段的后期(晶棒长度1300mm),距离固液界面距离L处晶棒附近区域温度T L和相对于固液界面高度L曲线图;
图22是铜饰法表征法的具体操作示意图。
具体实施方式
下面详细描述本公开的实施例,旨在用于解释本公开,而不能理解为对本公开的限制。
本公开中,如无特殊说明,下述属于以及符号所表示的意义如下所定义的:液口距为导流筒的下端与固液界面处的间隔距离。本文中所使用的术语“完美晶体”并不意指绝对完美的晶体或没有任何缺陷的晶体,而是容许存在极少量的一种或多种缺陷,其不足以使晶体或得到的晶圆的某种电学或机械学特性产生大的变化而致使其制成电子器件的性能劣化。
直拉法制备晶体是把原料多晶硅块放入石英坩埚中,然后对石英坩埚加热,将装在石英坩埚中的多晶硅熔化,然后将籽晶插入熔体表面进行熔接,同时转动籽晶,再反转坩埚,籽晶缓慢向上提升,经过引晶、放大、转肩、等径生长、收尾等过程制备晶棒的过程。并且在直拉法制备晶体过程中,晶棒中热历史对晶体缺陷分布、种类和尺寸直接相关,而本申请正是为了使晶体中点缺陷充分对流扩散和再结合,减小游离V型点缺陷和I型点缺陷浓度。
为此,在本公开的一个方面,本公开提出了一种单晶体的制备方法。根据本公开的实施例,该方法包括:在等径生长阶段根据线性方程G h=k*h+G 0得到G h,其中,G 0为固液界面处温度梯度,取值为35~55K/cm,k为常数,h为相对于固液界面高度,取值为0~10mm,G h为距离固液界面高度h处晶棒轴向温度梯度,单位为K/cm,在固液界面至第一参考面的区域内调整温度梯度至G h,其中,上述第一参考面为固液界面上方10mm的界面。由此, 通过控制固液界面至第一参考面的区域内晶棒轴向温度梯度,使长晶过程形成的I型点缺陷和V型点缺陷在固液界面附近区域充分扩散和再复合,从而减少点缺陷的形成浓度。
根据本公开的实施例,为了对固液界面至第一参考面的区域内晶棒温度梯度精确控制,本申请分别对等径生长阶段的初期、等径生长阶段的中期和等径生长阶段的后期对应温度梯度进行优化。需要说明的是,“等径生长阶段的初期”定义为生长的晶体长度不大于400mm的阶段;“等径生长阶段的中期”定义为生长的晶体长度大于400mm且坩埚中硅熔体液面进入坩埚R角前的阶段;“等径生长阶段的后期”定义为坩埚中硅熔体液面进入坩埚R角后的阶段。
具体的,在等径生长阶段的初期,根据上述线性方程G h=k*h+G 0中k=k 1计算得到G h1,根据计算得到的G h1调节晶棒的提拉速度为v 1、液口距为d 1,其中,上述k 1、v 1和d 1满足:-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm,即根据线性方程G h=k*h+G 0中k=k 1计算得到G h1的理论值,然后在上述v 1和d 1对应范围内通过调整提拉速度v 1和液口距d 1,使得G h1的实际值等于G h1的理论值。可以的理解是当G h1的实际值小于理论值时,减小液口距d 1,使得G 0增加,则G h1靠近理论值,同时在上述提拉速度v 1范围内可降低提拉速度,使得晶体中形成点缺陷有一定时间充分扩散和再复合,则降低点缺陷的形成浓度;在等径生长阶段的中期,根据上述线性方程G h=k*h+G 0中k=k 2计算得到G h2,根据计算得到的G h2调节晶棒的提拉速度为v 2、液口距为d 2,其中,上述k 2、v 2和d 2满足:-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm,即根据线性方程G h=k*h+G 0中k=k 2计算得到G h2的理论值,然后在上述v 2和d 2对应范围内通过调整提拉速度v 2和液口距d 2,使得G h2的实际值等于G h2的理论值。可以的理解是当G h2的实际值大于理论值时,增大液口距d 2,使得G 0减小,则G h2靠近理论值,同时在上述提拉速度范围内可降低提拉速度v 2,使得晶体中形成的点缺陷有一定时间充分扩散和再复合,则降低点缺陷的形成浓度;在等径生长阶段的后期,根据上述线性方程G h=k*h+G 0中k=k 3计算得到G h3,根据计算得到的G h3调节晶棒的提拉速度为v 3、液口距为d 3,其中,上述k 3、v 3和d 3满足:-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm,即根据线性方程G h=k*h+G 0中k=k 3计算得到G h3的理论值,然后在上述v 3和d 3对应范围内通过调整拉速度v 3和液口距d 3,使得G h3的实际值等于G h3的理论值。可以的理解是当G h3的实际值小于G h3的理论值,减小液口距d 3,使得G 0增加,则G h3靠近理论值,同时在上述提拉速度范围内可降低提拉速度v 3,使得晶体中形成点缺陷有一定时间充分扩散和再复合,则降低点缺陷的形成浓度。由此,依据上述的制备方法,通过分别控制等径生长阶段的初期、中期和后期的提拉速度和液口距,实现对各时期温度梯度G h的调整,使生长过程形成的I型点缺陷和V型点缺陷在固液界面附近区域充分扩散和再复合,减少游离V型点缺陷和I型点缺陷浓度。则长晶生长阶段的提拉速度变化范围可增加10%,液口距变化范围可增加5%,进而扩大了工艺参数可调节的窗口,提高晶体品质和完美晶体的良率。其中,良率为生长出完美晶体的熔料量与投料量的比值。
进一步地,为了验证上述线性方程的拟合性,对在长晶固液界面至第一参考面的区域内不同生长阶段时期的晶棒轴向温度梯度和相对于固液界面高度进行绘图,如图1所示,在等径生长阶段的初期(晶棒长度400mm),以相对于长晶界面高度h为横坐标,以晶棒轴向温度梯度G h为纵坐标绘制曲线,线性方程为G h=-0.10h+41.361,该方程的拟合度R 2=0.9994;如图2所示,在等径生长阶段的中期(晶棒长度700mm),以相对于长晶界面高度h为横坐标,以晶棒轴向温度梯度G h为纵坐标绘制曲线,线性方程为G h=-0.24.h+47.031,该方程的拟合度R 2=0.9988;如图3和4所示,在等径生长阶段后期(晶棒长度分别为1000mm和1300mm),以相对于长晶界面高度h为横坐标,以晶棒轴向温度梯度G h为纵坐标绘制曲线,晶棒长度为1000mm对应的线性方程为G h=-0.2463h+49.155,该方程的拟合度R 2=0.999,晶棒长度为1300mm对应的线性方程为G h=-0.1507h+45.003,该方程的拟合度R 2=0.999。总之,上述各个线性方程的拟合度都大于0.998,说明拟合后的线性方程与实际情况非常符合,因 此根据该线性方程对应去调节提拉速度和液口距,可精确地控制固液界面到第一参考面的区域内轴向温度梯度,有效地减少点缺陷的形成浓度。
根据本公开的实施例,为了进一步提高晶体品质和完美晶体的良率,在等径生长阶段的初期,控制晶棒的中心位置轴向方向温度梯度变化量△G c为0.2~1K/cm,控制晶棒的边缘位置轴向方向温度梯度变化量△G e为5~10K/cm,控制晶棒的径向温度梯度G r不大于6K/cm;在等径生长阶段的中期,控制晶棒的中心位置轴向方向温度梯度变化量△G c为2~6K/cm,控制晶棒的边缘位置轴向方向温度梯度变化量△G e为5~10K/cm,控制晶棒的径向温度梯度G r不大于10K/cm;在等径生长阶段的后期,控制晶棒的中心位置轴向方向温度梯度变化量△G c为0.2~1K/cm,控制晶棒的边缘位置轴向方向温度梯度变化量△G e为5~10K/cm,控制晶棒的径向温度梯度G r不大于6K/cm。则在等径生长阶段的各个不同时期,通过控制晶棒的中心位置和边缘位置的轴向温度梯度变化量及径向温度梯度,保证固液界面形状近似于平面,使得同一截面内轴向温度梯度近似均匀,根据V/G理论可增大生长完美晶体的窗口,提高完美晶体的良率。
根据本公开的实施例,本申请上述方法还包括调节晶棒上1685K~955K温度带的宽度,其中,温度带包括第一温度带、第二温度带和第三温度带,第一温度带的范围为1685K至1605K,第二温度带的范围为1605K至1355K,第三温度带的范围为1355K至955K。
根据本公开的一个实施例,第一温度带包括1685K~1665K、1665K~1645K、1645K~1625K、1625K~1605K温度带,上述第一温度带包括若干个的小温度带,且上述各小温度带宽度差异小于1.0mm,且生长过程中同一温度带宽度变化小于0.5mm。在生长过程中并且在等径生长阶段第一温度带范围内,根据线性方程D x=m*G 0+b调整温度带宽度,其中G 0为固液界面处温度梯度,取值为35~55K/cm,D x为温度带宽度,单位为mm。可以的理解是依照上述线性方程来调整第一温度带宽度,使得在等径生长阶段的同一时期,第一温度带内各个小温度带宽度间的差异小于1.0㎜,也就是说各个小温度宽度比较相近,同时在等径生长阶段的各个时期,每个小温度带宽度差异变化小于0.5㎜,也就是说每个小温度带宽度变化量较小,可精确地控制晶棒上第一温度带宽度的位置。
具体的,在等径生长阶段的初期,根据上述线性方程D x=m*G 0+b中m=m 1,b=b 1,计算得到D 1,根据D 1调节导流筒距固液界面的高度,其中-0.1≤m 1≤-0.09,9≤b 1≤9.1,即根据线性方程D x=m*G 0+b中m=m 1,b=b 1,计算得到D 1的理论值,然后通过调节导流筒距固液界面的高度,使得D 1的实际值等于D 1的理论值。可以的理解是,通过调节导流筒距固液界面的高度实现对固液界面处温度梯度G 0的调节,进而实现对D 1的调整。根据实际生产情形,导流筒距固液界面的高度H增大是通过坩埚轴旋转使得石英坩埚下降而实现的,则石英坩埚较靠近加热器,使得G 0减小,进而温度带宽度D 1增大;同理导流筒距固液界面的高度H减小是通过坩埚轴旋转使得石英坩埚上升,则石英坩埚较远离加热器,使得G 0增加,进而温度带宽度D 1减小。在等径生长阶段的后期,根据上述线性方程D x=m*G 0+b中m=m 2,b=b 2,计算得到D 2,根据D 2调节导流筒距固液界面的高度,其中-0.09≤m 2≤-0.05,7≤b 2≤8.5,即根据线性方程D x=m*G 0+b中m=m 2,b=b 2,计算得到D 2的理论值,然后通过调节导流筒距固液界面的高度,使得D 2的实际值等于D 2的理论值。可以理解为,通过调节导流筒距固液界面的高度H实现对固液界面处温度梯度G 0的调节,进而实现对D 2的调整。根据实际生产情形,导流筒距固液界面的高度H增大是通过坩埚轴旋转使得石英坩埚下降,则石英坩埚较靠近加热器,使得G 0减小,进而温度带宽度D 2增大;同理导流筒距固液界面的高度H减小是通过坩埚轴旋转使得石英坩埚上升,则石英坩埚较远离加热器,使得G 0增加,进而温度带宽度D 2减小。同时在等径生长阶段的中期,中期的温度带宽度D 3处于初期的温度带宽度D 1和后期的温度带宽度D 2之间。在从初期经过中期到后期的过程中,从初期转换为中期的时刻,此时中期的温度带宽度D 3与初期的温度带宽度D 1相等,从中期转换为后期的时刻,此时中期的温度带宽度D 3与后期的温度带宽度D 2相等,在中期的期间内温度 带宽度D 3在D 1与D 2之间变化。总之,依据各个时期的线性方程,通过调节导流筒距固液界面的高度,及时地调整第一温度带宽度,使其最终接近理论值,有效地控制晶棒的温度分布。
具体的,在第一温度带的1685K~1665K温度带内,如图5为等径生长阶段初期和中期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0998G 0+9.0432,该方程的拟合度R 2=0.9923;如图6为在等径生长阶段后期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0879G 0+8.5236,该方程的拟合度R 2=0.9963,图中以固液界面处温度梯度G 0为横坐标,以温度带宽度D x为纵坐标,绘制曲线。
具体的,在第一温度带的1665K~1645K温度带内,如图7为在等径生长阶段的初期和中期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0971G 0+9.0516,该方程的拟合度R 2=0.989;如图8为在等径生长阶段的后期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0688G 0+7.7369,该方程的拟合度R 2=0.9977,图中以固液界面处温度梯度G 0为横坐标,以温度带宽度D x为纵坐标。
具体的,在第一温度带的1645K~1625K温度带内,如图9为在等径生长阶段的初期和中期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0909G 0+8.872,该方程的拟合度R 2=0.9763;如图10为在等径生长阶段的后期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0705G 0+7.8889,该方程的拟合度R 2=0.9906,图中以固液界面处温度梯度G 0为横坐标,以温度带宽度D x为纵坐标。
具体的,在第一温度带的1625K~1605K温度带内,如图11为在等径生长阶段的初期和中期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.093G 0+9.0792,该方程的拟合度R 2=0.9562;如图12为在等径生长阶段的后期的温度带宽度曲线,该温度带宽度的线性方程为D x=-0.0582G 0+7.3744,方程的拟合度R 2=0.9979,图中以固液界面处温度梯度G 0为横坐标,以温度带宽度D x为纵坐标。
总之,第一温度带内各个小温度带宽度的线性方程的拟合度都大于0.95以上,说明拟合后的线性方程与实际情况非常符合。因此根据第一温度带的线性方程有指导性去调节导流筒距离固液界面的高度,可精确地调节第一温度带宽度,控制晶棒的热历史并抑制点缺陷长大,控制缺陷的尺寸,提高生长完美晶体的良率。
根据本公开的再一个实施例,第二温度带包括1605K~1555K、1555K~1505K、1505K~1455K、1455K~1405K、1405K~1355K温度带,上述第二温度带包括多个小温度带,且各个小温度带宽度差异小于1.5mm,且生长过程中同一温度带宽度变化小于1.0mm。并且在等径生长阶段第二温度带范围内,根据线性方程D y=m*G 0+b调整温度带宽度,其中G 0为固液界面处温度梯度,取值为35~55K/cm,D y为温度带宽度,单位为mm。可以的理解是依照上述线性方程来调整第二温度带宽度,使得在等径生长阶段的同一时期,第二温度带内各个小温度带宽度间的差异小于1.5㎜,也就是说各个小温度宽度几乎相同,同时在等径生长阶段的各个时期,每个小温度带宽度差异变化小于1.0㎜,也就是说每个小温度带宽度变化量小,可精确地控制晶棒上第一温度带宽度的位置。
具体的,在等径生长阶段的初期,根据线性方程D y=m*G 0+b中m=m 3,b=b 3,计算得到D 4,根据D 4调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.27≤m 3≤-0.24,23.5≤b 3≤25.5,即根据线性方程D=m*G 0+b中m=m 3,b=b 3,计算得到D 4的理论值,然后通过调节水冷套距晶棒外壁的距离,使得D 4的实际值等于D 4的理论值。可以的理解是若D 4的实际值小于理论值,降低水冷套中冷却水流量或增大水冷套与晶棒外壁的距离,则温度带宽度D 4增大,较靠近理论值;在等径生长阶段的后期,根据线性方程D=m*G 0+b中m=m 4,b=b 4,计算得到D 5,根据D 5调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.1≤m 4≤-0.06,16.5≤b 4≤17.5,即根据线性方程D=m*G 0+b中m=m 4,b=b 4,计算得到D 5的理论值,然后通过调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,使得D 5 的实际值等于D 5的理论值。可以的理解是若D 5的实际值大于理论值,增大水冷套中冷却水流量或减小水冷套与晶棒外壁的距离,则温度带宽度D 5变小,较靠近理论值;在等径生长阶段的中期,中期的温度带宽度D 6处于初期的温度带宽度D 4和后期的温度带宽度D 5之间。可以的理解是在从初期经过中期到后期的过程中,从初期转换为中期的时刻,此时中期的温度带宽度D 6与初期的温度带宽度D 4相等,从中期转换为后期的时刻,此时中期的温度带宽度D 6与后期的温度带宽度D 5相等,在中期的期间内温度带宽度D 6在D 4与D 5之间变化。
同理采用第二温度宽度的模拟方法,可得到第二温度带内各个小温度带宽度的线性方程的拟合度都大于0.95以上,在此不再赘述。因此根据第二温度带的线性方程有指导性去调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,可精确地调节第二温度带宽度,控制晶棒的热历史及缺陷的尺寸,提高生长完美晶体的良率。
根据本公开的又一个实施例,在等径生长阶段第三温度带范围内,根据线性方程D z=n*L+c调整温度带宽度,D z为温度带宽度,单位为mm,L为晶棒长度,单位为mm。具体的,在等径生长阶段的初期,根据线性方程D=n*L+c中n=n 1,c=c 1,计算得到D 7,根据D 7调节水冷套内冷却水流量,其中-0.035≤n 1≤-0.015,35≤c 1≤45,即根据线性方程D=n*L+c中n=n 1,c=c 1,计算得到D 7的理论值,然后通过调节水冷套内冷却水流量,使得D 7的实际值等于D 7的理论值。可以的理解是若D 7的实际值大于理论值,则增大水冷套中冷却水流量,则温度带宽度D 7增大,较靠近理论值;在等径生长阶段的中期和后期,控制温度带宽度保持不变且中期的温度带宽度D 8等于初期的温度带宽度D 7。可以的理解是在等径生长阶段,从初期到中期的时刻,温度带宽度D 8等于此刻初期的温度带宽度D 7,在中期及后期的阶段,温度带宽度D 8一直保持不变。
同理采用第三温度宽度的模拟方法,可得到第三温度带的线性方程的拟合度也大于0.95以上,在此不再赘述。因此根据第三温度带的线性方程有指导性去调节水冷套内冷却水流量,可精确地调节第三温度带宽度,控制晶棒的热历史及缺陷的尺寸,提高生长完美晶体的良率。
总之,通过对晶棒温度带宽度的精确控制,有效地抑制了微缺陷的形核和尺寸,生长完美晶体的良率提高了10%-20%。
进一步地,依据晶体生长理论,液相变为固相生长出晶体,则固液界面与熔液间存在边界层,边界层对界面具有输送作用,边界层温度分布决定长晶驱动力的大小和固液界面温度梯度,具体参考图13。可以的理解是边界层处于固液界面至第二参考面(第二参考面为固液界面上方10mm的界面)的区域内。Voronkov理论指出长晶固液界面附近形成晶格间隙型点缺陷(I型点缺陷)和空孔型点缺陷(V型点缺陷),点缺陷在边界层进行对流扩散和再结合。而边界层的热场结构由长晶提拉速度和液口距决定,因此对不同的热场条件,可同时调整长晶提拉速度和液口距来实现或优化热场结构,以实现边界层温度梯度分布达到制备完美晶体的要求。且目前大尺寸(如450㎜及以上)晶棒的生长出完美晶体比较困难,因为晶棒的尺寸比较大,根据V/G理论,仅仅对晶棒温度梯度进行控制,很难控制晶棒的轴向温度均匀化,造成长晶过程中完美晶体的良率比较低。对此,发明人发现晶棒附近区域内温度分布及固液界面附近的温度梯度分布对大尺寸晶棒的生长有直接影响。
为此,发明人提出了对在固液界面至第一参考面的区域内调整温度梯度的同时调整固液界面至第二参考面(第二参考面为固液界面上方10mm的界面)的区域内的温度梯度。
根据本发明的实施例,在等径生长阶段,根据方程G H=A*H 2+B*H+G 0得到G H,其中,A和B为常数,G 0为固液界面处温度梯度,取值为35~55K/cm,H为相对于固液界面高度,取值为-10~0mm,G H为距离固液界面高度H处边界层的轴向温度梯度,单位为K/cm;
在固液界面至第二参考面的区域内调整温度梯度至G H,其中,第二参考面为固液界面下方10mm的界面。
由此,通过量化控制固液界面至第一参考面的区域内温度梯度和固液界面至第二参考 面的区域内温度梯度,也就是精确控制了固液界面处从第一参考面到第二参考面区域内的温度梯度,而固液界面上下方区域处的温度梯度直接影响到点缺陷的形成浓度。
根据本发明的实施例,为了对固液界面至第一参考面区域和固液界面至第二参考面区域内温度梯度进行精确控制,本申请分别对等径生长阶段的初期、中期和后期对应温度梯度进行优化。
具体的,在等径生长阶段的初期,根据方程G H=A*H 2+B*H+G 0中A=A 1,B=B 1计算得到G H1,同时根据方程G h=k*h+G 0中k=k 1计算得到G h1,根据G h1和G H1调节晶棒的提拉速度为v 1、液口距为d 1,其中,上述A 1、B 1、k 1、v 1和d 1满足:0.03≤A 1≤0.05,-0.3≤B 1≤-0.1,-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm,即根据方程G H=A*H 2+B*H+G 0中A=A 1,B=B 1计算得到G H1的理论值,根据方程G h=k*h+G 0中k=k 1计算得到G h1的理论值,然后在上述v 1和d 1对应范围内通过调整提拉速度v 1和液口距d 1,使得G H1的实际值等于G H1的理论值,G h1的实际值等于G h1的理论值。可以的理解是当G H1和G h1的实际值分别小于理论值时,减小液口距d 1,使得G 0增加,则G H1和G h1靠近理论值,同时在上述提拉速度v 1范围内可降低提拉速度,使得晶体中形成点缺陷有一定时间充分扩散和再复合,则降低点缺陷的形成浓度;在等径生长阶段的中期,根据方程G H=A*H 2+B*H+G 0中A=A 2,B=B 2计算得到G H2,同时根据方程G h=k*h+G 0中k=k 2计算得到G h2,根据G h2和G H2调节晶棒的提拉速度为v 2、液口距为d 2,其中,上述A 2、B 2、k 2、v 2和d 2满足:0.01≤A 2≤0.03,-0.2≤B 2≤-0.1,-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm,即根据方程G H=A*H 2+B*H+G 0中A=A 2,B=B 2计算得到G H2的理论值,根据方程G h=k*h+G 0中k=k 2计算得到G h2的理论值,然后在上述v 2和d 2对应范围内通过调整提拉速度v 2和液口距d 2,使得G h2的实际值等于G h2的理论值,G H2的实际值等于G H2的理论值。可以的理解是当G H2和G h2的实际值分别大于理论值时,增大液口距d 2,使得G 0减小,则G H2和G h2靠近理论值,同时在上述提拉速度范围内可降低提拉速度v 2,使得晶体中形成的点缺陷有一定时间充分扩散和再复合,则降低点缺陷的形成浓度;在等径生长阶段的后期,根据方程G H=A*H 2+B*H+G 0中A=A 3,B=B 3计算得到G H3,同时根据方程G h=k*h+G 0中k=k 3计算得到G h3,根据G h3和G H3调节晶棒的提拉速度为v 3、液口距为d 3,其中,上述A 3、B 3、k 3、v 3和d 3满足:0.01≤A 3≤0.03,-0.2≤B 3≤-0.1,-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm,即根据方程G H=A*H 2+B*H+G 0中A=A 3,B=B 3计算得到G H3的理论值,根据方程G h=k*h+G 0中k=k 3计算得到G h3的理论值,然后在上述v 3和d 3对应范围内通过调整提拉速度v 3和液口距d 3,使得G h3的实际值等于G h3的理论值,G H3的实际值等于G H3的理论值。可以的理解是当G h3和G H3的实际值分别小于理论值,减小液口距d 3,使得G 0增加,则G h3和G H3靠近理论值,同时在上述提拉速度范围内可降低提拉速度v 3,使得晶体中形成点缺陷有一定时间充分扩散和再复合,则降低点缺陷的形成浓度。由此,通过量化控制固液界面至第一参考面的区域内温度梯度和固液界面至第二参考面的区域内温度梯度,使长晶过程形成的I型点缺陷和V型点缺陷在固液界面处充分扩散和再复合,减小点缺陷形成浓度,控制微缺陷的尺寸,增加完美晶体窗口宽度。另外,该方法特别适合大直径晶棒的生长,也可提高大尺寸晶棒的完美晶体的良率。
进一步地,为了验证上述方程的拟合性,对在固液界面至第二参考面区域的不同生长阶段时期的固液界面附近轴向温度梯度和相对于固液界面高度进行绘图,如图14所示,在固液界面至第二参考面区域内,在等径生长阶段的初期(晶棒长度400mm),以相对于固液界面高度H为横坐标,以固液界面附近的轴向温度梯度G H为纵坐标绘制曲线,方程为G H=0.039H 2-0.2006H+4.1184,该方程的拟合度R 2=0.9915;如图15所示,在等径生长阶段的中期(晶棒长度700mm),以相对于长晶界面高度H为横坐标,以固液界面附近的轴向温度梯度G H为纵坐标绘制曲线,方程为G H=0.0188H 2-0.0946H+3.1755,该方程的拟合度R 2=0.9952;如图16和17所示,在等径生长阶段的后期(晶棒长度分别为1000mm和1300mm),以相对于固液界面高度H为横坐标,以固液界面附近的轴向温度梯度G H为纵坐标绘制曲线, 晶棒长度为1000mm其方程为G H=0.0136H 2+0.0076H+2.0733,该方程的拟合度R 2=0.9995,晶棒长度为1300mm其方程为G H=0.0163H 2-0.0924H+2.4074,该方程的拟合度R 2=0.9983。总之,上述各个方程的拟合度都大于0.99,说明拟合后的方程与实际情况非常符合,因此根据该方程对应去调节提拉速度和液口距,可精确地控制固液界面到第一参考平面区域和固液界面到第二参考平面区域内的轴向温度梯度,有效地减少点缺陷浓度。
需要说明的是,若相对于固液界面高度H数值为负数,则指其位于固液界面下方且其距离固液界面高度为|H|的位置;反之相对于固液界面高度h数值为正数,则指其位于固液界面上方且其距离固液界面高度为h的位置。
根据本发明的实施例,为了进一步提高晶体品质和完美晶体的良率,在固液界面至第二参考面区域内:在等径生长阶段的初期,控制边界层的中心位置轴向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向温度梯度变化量△G e为3.5~5.5K/cm,控制边界层的径向温度梯度G r不大于2K/cm;在等径生长阶段的中期,控制边界层的中心位置轴向温度梯度变化量△G c为4~6K/cm,控制边界层的边缘位置轴向温度梯度变化量△G e不大于5.4K/cm,控制边界层的径向温度梯度G r不大于4K/cm;在等径生长阶段的后期,控制边界层的中心位置轴向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向温度梯度变化量△G e为3.5~5.5K/cm,控制边界层的径向温度梯度G r不大于2K/cm。则在固液界面至第二参考面区域内,等径生长阶段的各个不同时期,通过控制边界层的中心位置和边缘位置的轴向温度梯度变化量及径向温度梯度,使得固液界面下方区域的温度梯度分布均匀,保证固液界面形状近似于平面,可提高完美晶体的良率,其中,边界层位于固液界面至第二参考面区域内。
根据本发明的实施例,本申请上述方法还包括在固液界面至其上方的区域内,根据方程T L=g*L 3+f*L 2+q*L+T 0计算得到T L,g、f和q为常数,T 0为三相点处的平均温度,单位为K,L为相对于固液界面高度,取值为0~300mm,T L为距离固液界面距离L处晶棒附近区域温度,单位为K。由此,通过量化晶棒附近区域温度分布,优化晶棒附近的热场分布,抑制微缺陷的长大,提高制备完美晶体的良率,并且需要说明的是,本文中,“晶棒附近区域”为晶棒壁与水冷套内壁之间的中心区域。
具体的,根据方程T L=g*L 3+f*L 2+q*L+T 0中g=g 1,f=f 1和q=q 1计算得到T L1,根据T L1调节晶棒附近气体流量(气体流量范围为90~120sL/pm),其中,g 1取值为-8e -0.5~-7e -0.5,f 1取值为0.0411~0.0477,q 1取值为-10.309~-9.6535,即根据方程T L=g*L 3+f*L 2+q*L+T 0中g=g 1,f=f 1和q=q 1计算得到T L1的理论值,然后通过调节晶棒附近的气体流量,使得T L1的实际值等于T L1的理论值。可以理解为当T L1的实际值大于理论值时,增大晶棒附近的气体流量,使得T L1靠近理论值;反之,当T L1的实际值小于理论值时,减小晶棒附近的气体流量,使得T L1靠近理论值。
进一步地,为了验证上述方程的拟合性,在固液界面至第二参考面区域内的不同长晶阶段的晶棒附近区域温度和相对于固液界面高度进行绘图,如图18所示,在等径生长阶段的初期(晶棒长度400mm),以相对于长晶固液界面高度L为横坐标,以晶棒附近区域温度T L为纵坐标绘制曲线,方程为T L=-7e -0.5L 3+0.0433L 2-9.9252L+1532.4,该方程的拟合度R 2=0.9943;如图19所示,在等径生长阶段的中期(晶棒长度700mm),以相对于固液界面高度L为横坐标,以晶棒附近区域温度T L为纵坐标绘制曲线,方程为T L=-8e -0.5L 3+0.0477L 2-10.309L+1542.4,该方程的拟合度R 2=0.994;如图20和21所示,在等径生长阶段后期(晶棒长度分别为1000mm和1300mm),以相对于固液界面高度L为横坐标,以晶棒附近区域温度T L为纵坐标绘制曲线,晶棒长度为1000mm对应方程为T L=-7e -0.5L 3+0.0411L 2-9.6535L+1558,该方程的拟合度R 2=0.9977,晶棒长度为1300mm对应方程为T L=-8e -0.5L 3+0.0474L 2-10.266L+1555.2,该方程的拟合度R 2=0.9937。同理,上述各个方程的拟合度都大于0.99,说明拟合后的方程与实际情况非常符合,因此根据该方程对 应去调节晶棒附近的气体流量,可以精确控制晶棒附近区域温度分布,进而影响到晶棒的温度带分布,进一步抑制微缺陷的尺寸,提高制备完美晶体的良率。
总之,通过对固液界面上下方区域的温度梯度的精确控制及调节晶棒附近的温度分布,有效地降低了点缺陷的形成浓度,也进一步抑制了微缺陷的尺寸,生长完美晶体的良率提高了10%-20%。
在本公开的再一个方面,本公开提出了一种硅晶体。根据本公开的实施例,上述硅晶体采用上述的方法制备得到。由此,该硅晶体具有较高的品质。可以的理解是该硅晶体中属于完美晶体的窗口比例大。
下面参考具体实施例,对本公开进行描述,需要说明的是,这些实施例仅仅是描述性的,而不以任何方式限制本公开。
实施例1
在等径生长阶段,按照如下方式调整固液界面至参考面的区域内温度梯度至G h及晶棒的温度带宽度,其中G 0取值为35~55K/cm:
在等径生长阶段的初期,根据线性方程G h=k*h+G 0中k=k 1计算得到G h1,根据G h1调节晶棒的提拉速度为v 1、液口距为d 1,其中,-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm,同时控制晶棒的中心位置轴向方向温度梯度△G c为0.2~1K/cm,控制晶棒的边缘位置轴向方向温度梯度△G e为5~10K/cm,控制晶棒的径向温度梯度G r不大于6K/cm,另外在等径生长阶段第一温度带1685K至1605K范围内,根据线性方程D x=m*G 0+b中m=m 1,b=b 1,计算得到D 1,根据D 1调节导流筒距固液界面的高度,其中-0.1≤m 1≤-0.09,9≤b 1≤9.1;在等径生长阶段第二温度带1605K至1355K范围内,根据线性方程D y=m*G 0+b调整中m=m 3,b=b 3,计算得到D 4,根据D 4调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.27≤m 3≤-0.24,23.5≤b 3≤25.5,在等径生长阶段第三温度带1355K至955K范围内,根据线性方程D z=n*L+c调整温度带宽度,其中取n=n 1,c=c 1,计算得到D 7,根据D 7调节水冷套内冷却水流量,其中-0.035≤n 1≤-0.015,35≤c 1≤45;
在等径生长阶段的中期,根据线性方程G h=k*h+G 0中k=k 2计算得到G h2,根据G h2调节晶棒的提拉速度为v 2、液口距为d 2,其中,-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm,同时控制晶棒的中心位置轴向方向温度梯度△G c为2~6K/cm,控制晶棒的边缘位置轴向方向温度梯度△G e为5~10K/cm,控制晶棒的径向温度梯度G r不大于10K/cm;同时,该中期第一温度带1685K至1605K范围内,控制温度带宽度D 3处于初期的温度带宽度D 1和后期的温度带宽度D 2之间;控制第二温度带1605K至1355K的温度带宽度D 6处于初期的温度带宽度D 4和后期的温度带宽度D 5之间;控制第三温度带1355K至955K的温度带宽度D z与等径生长阶段前期的第三温度带范围内温度带宽度D 7保持不变。
在等径生长阶段的后期,根据线性方程G h=k*h+G 0中k=k 3计算得到G h3,根据G h3调节晶棒的提拉速度为v 3、液口距为d 3,其中,上述k 3、v 3和d 3满足:-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm,同时控制晶棒的中心位置轴向方向温度梯度△G c为0.2~1K/cm,控制晶棒的边缘位置轴向方向温度梯度△G e为5~10K/cm,控制晶棒的径向温度梯度G r不大于6K/cm。在第一温度带1685K至1605K范围内,取m=m 2,b=b 2,计算得到D 2,根据D 2调节导流筒距固液界面的高度,其中-0.09≤m 2≤-0.05,7≤b 2≤8.5;并且在第二温度带1605K至1355K范围内,取m=m 4,b=b 4,计算得到D 5,根据D 5调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.1≤m 4≤-0.06,16.5≤b 4≤17.5;在第三温度带1355K至955K范围内,控制温度带宽度D 8与等径生长阶段前期的第三温度带范围内温度带宽度D 7保持不变。
上述实施例中涉及到调整工艺参数的部分,请参照前文详细描述的方式来调节,此处不再赘述。
对比例1
采用现有的生长装置及制备方法,提拉出晶棒,即与实施例工艺相比不进行温度梯度和温度带的控制。
针对实施例1和对比例1,分别截取晶棒长度180mm、340㎜、650㎜、950㎜、1200㎜和1350mm处的晶圆片,分别采用铜饰法对其缺陷进行表征,其缺陷表征结果如表1所示,其中实施例和对比例中的晶棒为硅晶棒。参考图22,铜饰法表征的具体操作包括:首先用自来水对试片进行清洗,再用表面活性剂清洗试片表面,以去除试片表面的表面颗粒,然后采用化学抛光对试片表面进行抛光和清洁,再在清洁后的试片表面涂布硝酸铜,待热处理后在试片表面形成铜析出物,然后在对试片表面进行抛光和清洁,最后进行刻蚀显影,由于铜沉淀会在微缺陷的周围区域产生压缩应力,这些应力会通过硅原子的释出而得以缓解,释出的硅原子会在铜周围形成新的位错,铜沉淀随后会在这些位错上形成。其中,试片为硅晶圆片。由于微缺陷周围形成许多铜沉淀及位错,形成较大的区域,因此在显微镜下可以观察到晶圆片的微缺陷。该微缺陷图中黑色部分为完美区,白色部分为缺陷区。
表1采用上述铜饰法对实施例1和对比例1所得晶棒各个位置处的晶圆片缺陷检测结果
晶圆片位置(㎜) 实施例1 对比例1
180 I型缺陷 V型缺陷
340 完美晶体 I型缺陷
650 完美晶体 I型缺陷
950 完美晶体 I型缺陷
1200 完美晶体 I型缺陷
1350 I型缺陷 V型缺陷
从上述表1检测结果可知,对比例1采用现有技术方法,其生长出的晶棒的缺陷开始为V型缺陷,中间为I型缺陷,最后为V型缺陷,整个晶棒没有出现完美晶体。实施例1采用本申请的制备方法,其生长出的晶棒在长度340㎜、650㎜、950㎜、1200㎜处都存在完美晶体且窗口较大(从铜饰缺陷图中可直接看出),根据V/G理论在提拉速度一定范围内(0.4≤v≤0.8mm/min),晶棒长度340㎜到长度1200㎜范围内都存在完美晶体。则采用本申请的制备方法确实提高了完美晶体的良率。
实施例2
生长450㎜的晶棒,并在固液界面至第一参考面的区域和固液界面至第二参考面的区域内,参考下列方程调整温度梯度G h和G H,其中G 0取值为35~55K/cm:
等径生长阶段的初期,根据方程G H=A*H 2+B*H+G 0取A=A 1,B=B 1计算得到G H1,取k=k 1计算得到G h1,根据G h1和G H1调节晶棒的提拉速度为v 1、液口距为d 1,其中,A 1、B 1、k 1、v 1和d 1满足:0.03≤A 1≤0.05,-0.3≤B 1≤-0.1,-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm,同时在固液界面至第二参考面区域,控制边界层的中心位置轴向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向温度梯度变化量△G e为3.5~5.5K/cm,控制边界层的径向温度梯度G r不大于2K/cm;另外在固液界面至第一参考面的区域,控制该区域内的中心位置轴向温度梯度变化量△G c为0.2~1K/cm,控制该区域内的边缘位置轴向温度梯度变化量△G e为5~10K/cm,控制该区域内的径向温度梯度G r不大于6K/cm;
等径生长阶段的中期,根据方程G H=A*H 2+B*H+G 0取A=A 2,B=B 2计算得到G H2,取k=k 2计算得到G h2,根据G h2和G H2调节晶棒的提拉速度为v 2、液口距为d 2,其中,A 2、B 2、k 2、v 2和d 2满足:0.01≤A 2≤0.03,-0.2≤B 2≤-0.1,-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm,同时在固液界面至第二参考面的区域,控制边界层的中心位置轴向温度梯度变化量△G c为 4~6K/cm,控制边界层的边缘位置轴向温度梯度变化量△G e不大于5.4K/cm,控制边界层的径向温度梯度G r不大于4K/cm;另外在固液界面至第一参考面的区域,控制该区域内的中心位置轴向温度梯度变化量△G c为2~6K/cm,控制该区域内的边缘位置轴向温度梯度变化量△G e为5~10K/cm,控制该区域内的径向温度梯度G r不大于10K/cm;
等径生长阶段的后期,根据方程G H=A*H 2+B*H+G 0取A=A 3,B=B 3计算得到G H3,取k=k 3计算得到G h3,根据G h3和G H3调节晶棒的提拉速度为v 3、液口距为d 3,其中,A 3、B 3、k 3、v 3和d 3满足:0.01≤A 3≤0.03,-0.2≤B 3≤-0.1,-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm,同时在固液界面至第二参考面的区域,控制边界层的中心位置轴向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向温度梯度变化量△G e为3.5~5.5K/cm,控制边界层的径向温度梯度G r不大于2K/cm;另外在固液界面至第一参考面的区域,控制该区域内的中心位置轴向温度梯度变化量△G c为0.2~1K/cm,控制该区域内的边缘位置轴向温度梯度变化量△G e为5~10K/cm,控制该区域内的径向温度梯度G r不大于6K/cm。
并且在固液界面至其上方的区域内,根据方程T L=m*L 3+n*L 2+q*L+T 0中m=m 1,n=n 1和q=q 1计算得到T L1,根据T L1调节晶棒附近气体流量,其中,m 1取值为-8e -0.5~-7e -0.5,n 1取值为0.0411~0.0477,q 1取值为-10.309~-9.6535。
对比例2
采用现有的生长装置及制备方法,提拉出450㎜的晶棒,即与实施例2工艺相比不进行固液界面附近的温度梯度及晶棒附近温度的控制。
针对实施例2和对比例2,分别截取晶棒长度100mm、400㎜、700㎜、1000㎜、1300㎜处的晶圆片,分别采用铜饰法对其缺陷进行表征,其缺陷表征结果如表2所示,其中,实施例和对比例中的晶棒为硅晶棒。
表2采用上述铜饰法对实施例2和对比例2所得晶棒各个位置处的晶圆片缺陷检测结果
晶圆片位置(㎜) 实施例2 对比例2
100 I型缺陷 V型缺陷
400 完美晶体 I型缺陷
700 完美晶体 I型缺陷
1000 完美晶体 I型缺陷
1300 完美晶体 V型缺陷
从上述表2检测结果可知,对比例2采用现有技术方法,其生长出的晶棒的缺陷开始为V型缺陷,中间为I型缺陷、最后为V型缺陷,整个晶棒没有出现完美晶体。实施例2采用本申请的制备方法,其生长出的晶棒在长度400㎜、700㎜、1000㎜、1300㎜处都存在完美晶体且窗口较大(从铜饰缺陷图中可直接看出),根据V/G理论在提拉速度一定范围内(0.4≤v≤0.8mm/min),晶棒长度400㎜到长度1300㎜范围内都存在完美晶体。则采用本申请的制备方法确实提高了完美晶体的良率。
在本说明书的描述中,参考术语“一个实施例”、“一些实施例”、“示例”、“具体示例”、或“一些示例”等的描述意指结合该实施例或示例描述的具体特征、结构、材料或者特点包含于本公开的至少一个实施例或示例中。在本说明书中,对上述术语的示意性表述不必须针对的是相同的实施例或示例。而且,描述的具体特征、结构、材料或者特点可以在任一个或多个实施例或示例中以合适的方式结合。此外,在不相互矛盾的情况下,本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。
上面结合附图对本申请的实施例进行了描述,但是本申请并不局限于上述的具体实施方式,上述的具体实施方式仅仅是示意性的,而不是限制性的,本领域的普通技术人员在本申请的启示下,在不脱离本申请宗旨和权利要求所保护的范围情况下,还可做出很多形式,均属于本申请的保护之内。

Claims (17)

  1. 一种单晶体的制备方法,其中,包括:
    在等径生长阶段,根据线性方程G h=k*h+G 0得到G h,其中,G 0为固液界面处温度梯度,取值为35~55K/cm,h为相对于固液界面高度,取值为0~10mm,k为常数,G h为距离固液界面高度h处晶棒轴向温度梯度,单位为K/cm;
    在固液界面至第一参考面的区域内调整温度梯度至G h,其中,所述第一参考面为固液界面上方10mm的界面。
  2. 根据权利要求1所述的方法,其中,在等径生长阶段的初期,取k=k 1计算得到G h1,根据G h1调节晶棒的提拉速度为v 1、液口距为d 1,其中,所述k 1、所述v 1和所述d 1满足:-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm;
    在等径生长阶段的中期,取k=k 2计算得到G h2,根据G h2调节晶棒的提拉速度为v 2、液口距为d 2,其中,所述k 2、所述v 2和所述d 2满足:-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm;
    在等径生长阶段的后期,取k=k 3计算得到G h3,根据G h3调节晶棒的提拉速度为v 3、液口距为d 3,其中,所述k 3、所述v 3和所述d 3满足:-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm。
  3. 根据权利要求2所述的方法,其中,所述等径生长阶段的初期,控制△G c为0.2~1K/cm,△G e为5~10K/cm,G r不大于6K/cm;
    所述等径生长阶段的中期,控制△G c为2~6K/cm,△G e为5~10K/cm,G r不大于10K/cm;
    所述等径生长阶段的后期,控制△G c为0.2~1K/cm,△G e为5~10K/cm,G r不大于6K/cm;
    其中,晶棒中心位置的轴向温度梯度变化量为△G c,晶棒边缘位置的轴向温度梯度变化量为△G e,晶棒的径向温度梯度为G r
  4. 根据权利要求1所述的方法,其中,所述方法还包括调节晶棒上温度带的宽度,所述温度带包括第一温度带、第二温度带和第三温度带,所述第一温度带的范围为1685K至1605K,所述第二温度带的范围为1605K至1355K,所述第三温度带的范围为1355K至955K。
  5. 根据权利要求4所述的方法,其中,在等径生长阶段,所述第一温度带范围内,根据线性方程D x=m*G 0+b调整温度带宽度,其中G 0为固液界面处温度梯度,取值为35~55K/cm,D x为温度带宽度,单位为mm。
  6. 根据权利要求5所述的方法,其中,在等径生长阶段的初期,取m=m 1,b=b 1,计算得到D 1,根据D 1调节导流筒距固液界面的高度,其中-0.1≤m 1≤-0.09,9≤b 1≤9.1;
    在等径生长阶段的后期,取m=m 2,b=b 2,计算得到D 2,根据D 2调节导流筒距固液界面的高度,其中-0.09≤m 2≤-0.05,7≤b 2≤8.5;
    在等径生长阶段的中期,所述中期的温度带宽度D 3处于所述初期的温度带宽度D 1和所述后期的温度带宽度D 2之间。
  7. 根据权利要求4所述的方法,其中,在等径生长阶段,所述第二温度带范围内,根据线性方程D y=m*G 0+b调整温度带宽度,其中G 0为固液界面处温度梯度,取值为35~55K/cm,D y为温度带宽度,单位为mm。
  8. 根据权利要求7所述的方法,其中,在等径生长阶段的初期,取m=m 3,b=b 3,计算得到D 4,根据D 4调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.27≤m 3≤-0.24,23.5≤b 3≤25.5;
    在等径生长阶段的后期,取m=m 4,b=b 4,计算得到D 5,根据D 5调节水冷套距晶棒外壁的距离或水冷套内冷却水流量,其中-0.1≤m 4≤-0.06,16.5≤b 4≤17.5;
    在等径生长阶段的中期,所述中期的温度带宽度D 6处于所述初期的温度带宽度D 4和所述后期的温度带宽度D 5之间。
  9. 根据权利要求4所述的方法,其中,在等径生长阶段,所述第三温度带范围内,根据线性方程D z=n*L+c调整温度带宽度,D z为温度带宽度,单位为mm,L为晶棒长度,单位为mm。
  10. 根据权利要求9所述的方法,其中,在等径生长阶段的初期,取n=n 1,c=c 1,计算得到D 7,根据D 7调节水冷套内冷却水流量,其中-0.035≤n 1≤-0.015,35≤c 1≤45;
    在等径生长阶段的中期和后期,控制温度带宽度保持不变,且所述中期的温度带宽度D 8等于所述初期的温度带宽度D 7
  11. 根据权利要求1所述的方法,其中,在等径生长阶段,根据方程G H=A*H 2+B*H+G 0得到G H,其中,A和B为常数,G 0为固液界面处温度梯度,取值为35~55K/cm,H为相对于固液界面高度,取值为-10~0mm,G H为距离固液界面高度H处边界层的轴向温度梯度,单位为K/cm;
    在固液界面至第二参考面的区域内调整温度梯度至G H,其中,所述第二参考面为固液界面下方10mm的界面。
  12. 根据权利要求11所述的方法,其中,所述等径生长阶段的初期,取A=A 1,B=B 1计算得到G H1,取k=k 1计算得到G h1,根据G H1和G h1调节晶棒的提拉速度为v 1、液口距为d 1,其中,所述A 1、所述B 1、所述k 1、所述v 1和所述d 1满足:0.03≤A 1≤0.05,-0.3≤B 1≤-0.1,-0.12≤k 1≤-0.1,0.4≤v 1≤0.8mm/min,50≤d 1≤52mm;
    所述等径生长阶段的中期,取A=A 2,B=B 2计算得到G H2,取k=k 2计算得到G h2,根据G H2和G h2调节晶棒的提拉速度为v 2、液口距为d 2,其中,所述A 2、所述B 2、所述k 2、所述v 2和所述d 2满足:0.01≤A 2≤0.03,-0.2≤B 2≤-0.1,-0.25≤k 2≤-0.23,0.4≤v 2≤0.6mm/min,52≤d 2≤53mm;
    所述等径生长阶段的后期,取A=A 3,B=B 3计算得到G H3,取k=k 3计算得到G h3,根据G H3和G h3调节晶棒的提拉速度为v 3、液口距为d 3,其中,所述A 3、所述B 3、所述k 3、所述v 3和所述d 3满足:0.01≤A 3≤0.03,-0.2≤B 3≤-0.1,-0.16≤k 3≤0.14,0.6≤v 3≤0.8mm/min,54≤d 3≤55mm。
  13. 根据权利要求12所述的方法,其中,在固液界面至第二参考面的区域内:
    所述等径生长阶段的初期,控制边界层的中心位置轴向方向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向方向温度梯度变化量△G e为3.5~5.5K/cm;
    所述等径生长阶段的中期,控制边界层的中心位置轴向方向温度梯度变化量△G c为4~6K/cm,控制边界层的边缘位置轴向方向温度梯度变化量△G e不大于5.4K/cm;
    所述等径生长阶段的后期,控制边界层的中心位置轴向方向温度梯度变化量△G c为2~3K/cm,控制边界层的边缘位置轴向方向温度梯度变化量△G e为3.5~5.5K/cm;
    其中,所述边界层位于固液界面至第二参考面区域内。
  14. 根据权利要求12或13所述的方法,其中,在固液界面至第二参考面的区域内:
    所述等径生长阶段的初期,控制边界层的径向温度梯度G r不大于2K/cm;
    所述等径生长阶段的中期,控制边界层的径向温度梯度G r不大于4K/cm;
    所述等径生长阶段的后期,控制边界层的径向温度梯度G r不大于2K/cm。
  15. 根据权利要求1或11所述的方法,其中,根据方程T L=g*L 3+f*L 2+q*L+T 0计算得到T L,g、f和q为常数,T 0为三相点处的平均温度,单位为K,L为相对于固液界面高度,取值为0~300mm,T L为距离固液界面距离L处晶棒附近区域温度,单位为K,
    其中,所述晶棒附近区域是晶棒壁与水冷套内壁之间的中心区域。
  16. 根据权利要求15所述的方法,其中,取g=g 1,f=f 1和q=q 1计算得到T L1,根据T L1调节晶棒附近气体流量,其中,g 1取值为-8e -0.5~-7e -0.5,f 1取值为0.0411~0.0477,q 1取值为-10.309~-9.6535。
  17. 一种硅晶体,其中,所述硅晶体采用权利要求1-16中任一项所述的方法制备得到。
PCT/CN2022/140778 2021-12-27 2022-12-21 单晶体的制备方法及硅晶体 WO2023125206A1 (zh)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN202111619103.2 2021-12-27
CN202111619092.8A CN114481302B (zh) 2021-12-27 2021-12-27 单晶体的制备方法和硅晶体
CN202111619092.8 2021-12-27
CN202111619103.2A CN114438585A (zh) 2021-12-27 2021-12-27 单晶体的制备方法及硅晶体

Publications (1)

Publication Number Publication Date
WO2023125206A1 true WO2023125206A1 (zh) 2023-07-06

Family

ID=86997755

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/140778 WO2023125206A1 (zh) 2021-12-27 2022-12-21 单晶体的制备方法及硅晶体

Country Status (2)

Country Link
TW (1) TWI829486B (zh)
WO (1) WO2023125206A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1261928A (zh) * 1997-04-09 2000-08-02 Memc电子材料有限公司 低缺陷浓度的硅
CN112639175A (zh) * 2018-08-23 2021-04-09 胜高股份有限公司 单晶硅的培育方法
CN113549997A (zh) * 2021-06-25 2021-10-26 徐州鑫晶半导体科技有限公司 单晶生长的方法、装置及单晶体
CN114438585A (zh) * 2021-12-27 2022-05-06 徐州鑫晶半导体科技有限公司 单晶体的制备方法及硅晶体
CN114481302A (zh) * 2021-12-27 2022-05-13 徐州鑫晶半导体科技有限公司 单晶体的制备方法和硅晶体

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108914201B (zh) * 2018-08-29 2019-09-27 西安理工大学 一种直拉硅单晶生长过程工艺参数优化方法
CN110629283A (zh) * 2019-09-23 2019-12-31 上海新昇半导体科技有限公司 一种硅单晶的生长方法
CN112281210B (zh) * 2020-10-10 2022-05-17 徐州鑫晶半导体科技有限公司 晶体的生长装置及生长方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1261928A (zh) * 1997-04-09 2000-08-02 Memc电子材料有限公司 低缺陷浓度的硅
CN1854353A (zh) * 1997-04-09 2006-11-01 Memc电子材料有限公司 低缺陷密度、自间隙原子为主的硅
CN112639175A (zh) * 2018-08-23 2021-04-09 胜高股份有限公司 单晶硅的培育方法
CN113549997A (zh) * 2021-06-25 2021-10-26 徐州鑫晶半导体科技有限公司 单晶生长的方法、装置及单晶体
CN114438585A (zh) * 2021-12-27 2022-05-06 徐州鑫晶半导体科技有限公司 单晶体的制备方法及硅晶体
CN114481302A (zh) * 2021-12-27 2022-05-13 徐州鑫晶半导体科技有限公司 单晶体的制备方法和硅晶体

Also Published As

Publication number Publication date
TWI829486B (zh) 2024-01-11
TW202328512A (zh) 2023-07-16

Similar Documents

Publication Publication Date Title
KR100840751B1 (ko) 고품질 실리콘 단결정 잉곳 제조 방법, 성장 장치 및그로부터 제조된 잉곳 , 웨이퍼
CN111020691A (zh) 拉制晶棒的系统和控制方法
US8864906B2 (en) Method for producing silicon wafer
CN105189834B (zh) 冷却速率控制装置及包括其的铸锭生长装置
CN110904504B (zh) 一种拉晶炉及单晶硅棒的制备方法
KR102095597B1 (ko) 실리콘 단결정의 제조 방법
JP2008115050A (ja) エピタキシャルウェーハの製造方法
KR20140135981A (ko) 실리콘 단결정 웨이퍼의 제조방법
JP6927150B2 (ja) シリコン単結晶の製造方法
TWI664326B (zh) 熱遮蔽構件、單結晶提拉裝置及單結晶矽鑄錠的製造方法
KR101252404B1 (ko) 웨이퍼나 단결정 잉곳의 품질평가 방법 및 이를 이용한 단결정 잉곳의 품질 제어방법
KR20020019025A (ko) 실리콘 웨이퍼 및 실리콘 단결정의 제조방법
US20100127354A1 (en) Silicon single crystal and method for growing thereof, and silicon wafer and method for manufacturing thereof
CN114481302B (zh) 单晶体的制备方法和硅晶体
TWI593835B (zh) Silicon single crystal manufacturing method
WO2023125206A1 (zh) 单晶体的制备方法及硅晶体
US20240125005A1 (en) Method for crystal pulling
CN114438585A (zh) 单晶体的制备方法及硅晶体
JP3719088B2 (ja) 単結晶育成方法
JP5375636B2 (ja) シリコン単結晶の製造方法
JP2004161566A (ja) シリコンウェーハの製造方法およびその製造装置およびシリコンウェーハ
CN110965118B (zh) 一种导流筒装置和拉晶炉
JP5282762B2 (ja) シリコン単結晶の製造方法
JP6459944B2 (ja) シリコン単結晶の製造方法
TWI645080B (zh) 矽單結晶的製造方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22914460

Country of ref document: EP

Kind code of ref document: A1