A method for epitaxial filling of trench structured silicon carbide

Abstract

The present invention discloses an epitaxial filling method for a trench-structured silicon carbide, comprising the steps of: (1) conducting an epitaxial growth test on a first growth test wafer, adjusting the H2 flow rate, and obtaining its R G‑Bare 、R G‑TB 、R G‑MT ; (2) conducting an epitaxial growth test on a second growth test wafer and obtaining its R G‑Bare ; (3) conducting an epitaxial growth test on a third growth test wafer and obtaining its R G‑Bare ; (4) conducting an epitaxial growth test on a fourth growth test wafer and obtaining its R G‑TB = v2 and R G‑MT = v3, and 0.8v2 < v3 < 1.2v2; (5) conducting an epitaxial growth test on a fifth growth test wafer and obtaining its R G‑TB = v4 and R G‑MT = v5, and 0

Description

Technical Field

[0001] The present invention relates to the technical field of semiconductor materials, and in particular to an epitaxial filling method of trench-structured silicon carbide. Background Art

[0002] Superjunction technology uses a periodically alternating structure of P (pillar) and N (pillar) regions to replace the traditional single doped drift region as a voltage-resistant layer, so that the breakdown voltage of the superjunction device is mainly controlled by the superjunction thickness rather than the drift region doping. In turn, the device's specific on-resistance (R ON,SP ), the one-dimensional theoretical limit of semiconductor unipolar devices is increased from R ON,SP ∝BV 2.5 The relationship is reduced to R ON,SP ∝BV 1.32 , achieving both a larger breakdown voltage and a smaller specific on-resistance.

[0003] The method of realizing the alternating arrangement structure of P column areas and N column areas in silicon carbide can be multiple epitaxial growth combined with ultra-high energy ion implantation on planar silicon carbide, multiple epitaxial growth combined with ultra-high energy ion implantation, or epitaxial filling of epitaxial materials with opposite doping types on the silicon carbide trench structure formed by the etching process. The trench filling method is more suitable for commercial mass production, and its process methods generally have two kinds: the first method is similar to the method disclosed in Chinese patent CN114242644A. No etching gas is added during the epitaxial filling process. The epitaxial material grown on the mesa of the trench structure is etched after each epitaxy to prevent the trench from closing and forming a gap. The epitaxy and etching processes need to be repeated many times to complete the filling. Because the epitaxial material on the mesa is etched at the same time as the epitaxial material filled at the bottom of the trench, the disadvantage of this method is that the filling rate is slow, which is not conducive to the capacity improvement and cost control of mass production; the second method is to add etching gas during the epitaxial filling process, and take advantage of the fact that the growth of the bottom of the trench is less affected by etching than the growth of the mesa. Under the premise that the trench will not close during the epitaxial process, the filling rate in the trench is increased as much as possible. The excess epitaxial growth on the mesa can be removed through subsequent grinding and polishing processes. Therefore, this method is more conducive to capacity improvement and cost control in mass production.

[0004] Nevertheless, the second trench filling process mentioned above still needs to overcome the following difficulties to achieve industrialization: in the epitaxial growth process of silicon carbide filling the trench structure, the changes in each process parameter of trichlorosilane, ethylene, hydrogen, hydrogen chloride flow rate and reaction chamber pressure have a significant impact on the growth rate R of the trench structure mesa. G-MT and the bottom growth rate R G-TBThe change trends caused by this are not consistent, so when optimizing the optimal process parameters for trench silicon carbide epitaxy with specific doping type, doping concentration and aspect ratio, it is often necessary to fix other process parameters to specifically optimize a single process parameter. In this way, the various processes are repeatedly adjusted, and sometimes a certain process parameter needs to be repeatedly optimized, which leads to the use of a large number of trench structure silicon carbide wafers (trench structure silicon carbide wafers need to be realized through processes such as lithography and plasma etching, which are high in cost and low in production capacity), and each time the trench filling rate and filling rate are measured, they are determined by the SEM cross-section of the destructive cracks, resulting in huge expenditures on raw material costs, processing costs and testing costs. Summary of the Invention

[0005] In view of the above problems, the purpose of the present invention is to provide a solution for optimizing the process parameters of the continuous filling trench structure silicon carbide epitaxial method. The solution is applicable to the chemical vapor deposition filling growth silicon carbide process in the trichlorosilane, ethylene, and hydrogen chlorination reaction system to achieve the purpose of high trench bottom growth rate and void-free filling. In addition, the amount of trench structure silicon carbide wafers used and the number of cross-sectional SEM inspections used are minimized during the optimization process, thereby achieving the effect of saving process development costs.

[0006] In order to achieve the above object, the present invention provides a method for epitaxial filling of trench structured silicon carbide, comprising the steps of:

[0007] (1) A silicon carbide growth test piece having both a trench structure and a planar structure epitaxial layer is prepared, referred to as a first growth test piece. The first growth test piece is subjected to a chemical vapor deposition (CVD) epitaxial growth test in a reaction system of trichlorosilane (SiHCl3, TCS), ethylene (C2H4), hydrogen (H2) and hydrogen chloride (HCl), maintaining a carbon-silicon ratio C / Si=1, a fixed flow rate TCS=t1, C2H4=c1, HCl=l1, a reaction chamber pressure P=p1, and a growth temperature T=T1, and adjusting the flow rate of H2 to obtain H2=h1, so as to simultaneously satisfy the growth rate R of the planar surface of the first growth test piece. G-Bare Greater than 0 and the growth rate R of the trench terrace G-MT Less than 0, the growth rate R of the bottom of the groove is obtained G-TB =v1, and the growth rate R of the plane surface is obtained G-Bare and the growth rate R of the trench terrace G-MT The difference is v0;

[0008] (2) A silicon carbide growth test piece containing only a planar structure epitaxial layer is provided, referred to as a second growth test piece, and a chemical vapor deposition epitaxial growth test is performed on the second growth test piece, maintaining a carbon-silicon ratio C / Si=1, a fixed flow rate H2=h1, HCl=l1, a pressure P=p1, and a growth temperature T=T1, and adjusting the flow rates of TCS and C2H4 to obtain TCS=t2 and C2H4=c2 to satisfy the epitaxial growth rate R of the second growth test piece at this time. G-Bare In the 2v0 to 4v0 range;

[0009] (3) A silicon carbide growth test piece containing only a planar structure epitaxial layer is provided, referred to as a third growth test piece, and a chemical vapor deposition epitaxial growth test is performed on the third growth test piece. The flow rate TCS = t2, C2H4 = c2, H2 = h1, the pressure P = p1, and the growth temperature T = T1 are fixed, and the flow rate of HCl is adjusted to obtain HCl = l2 to meet the epitaxial growth rate R of the third growth test piece at this time. G-Bare In the range of v0 / 2 to v0;

[0010] (4) A test piece identical to the formal silicon carbide piece with a groove structure to be filled is provided, which is called the fourth growth test piece. The fourth growth test piece is subjected to a chemical vapor deposition epitaxial growth test, with fixed flow rates C2H4=c2, H2=h1, HCl=l2, pressure P=p1, and growth temperature T=T1. The flow rate of TCS is adjusted to obtain TCS=t3, so as to satisfy the groove bottom growth rate R of the fourth growth test piece at this time. G-TB =v2, trench terrace growth rate R G-MT =v3, and 0.8v2 <v3<1.2v2;

[0011] (5) A test piece identical to the formal silicon carbide piece with a groove structure to be filled is provided, which is called the fifth growth test piece. The fifth growth test piece is subjected to a chemical vapor deposition epitaxial growth test, with fixed flow rates TCS=t3, C2H4=c2, H2=h1, HCl=l2 and growth temperature T=T1. The flow rate of pressure P is adjusted within a certain range to obtain P=p2, so as to satisfy the growth rate R of the bottom of the groove of the fifth growth test piece at this time. G-TB =v4, Groove terrace growth rate R G-MT =v5, and 0 <v5<v4 / 3;

[0012] (6) Using the trench structured silicon carbide wafer, using the process parameters determined in steps (1)-(5): flow rate TCS=t3, C2H4=c2, H2=h1, HCl=l2 and pressure P=p2, growth temperature T=T1, chemical vapor deposition epitaxial growth filling is performed, and the epitaxial filling time S≥L / v4, where L is the depth of the trench structure, that is, the epitaxial filling of the trench structure is completed;

[0013] (7) After the epitaxial filling growth is completed, polish and remove the excess silicon carbide epitaxial layer with a thickness of H, where H ≥ v5 × S.

[0014] In some embodiments, the method for maintaining the carbon-silicon ratio C / Si = 1 in step (1) is as follows:

[0015] Keep the temperature of the TCS liquid source at 19°C. By controlling the flow rate of the H2 carrier gas entering and exiting, maintain the saturated vapor pressure of TCS at 1500 mbar, so that the flow rate ratio of TCS and C2H4 introduced into the reaction chamber satisfies the formula (c1 × 2) / (t1 × 0.758) = 1.

[0016] In some embodiments, the method for maintaining the carbon-silicon ratio C / Si = 1 in step (2) is as follows:

[0017] Keep the temperature of the TCS liquid source at 19°C. By controlling the flow rate of the H2 carrier gas entering and exiting, maintain the saturated vapor pressure of TCS at 1500 mbar, so that the flow rate ratio of TCS and C2H4 introduced into the reaction chamber satisfies the formula (c2 × 2) / (t2 × 0.758) = 1.

[0018] In some embodiments, the TCS flow rates t1, t2, and t3 satisfy the relationship t1 < t2 < t3, and 40 sccm < t1 < 500 sccm, t2 < 500 sccm, t3 < 500 sccm.

[0019] In some embodiments, the pressures p1 and p2 satisfy the relationship p1 < p2, and 70 mbar < p1 < 700 mbar, p2 < 700 mbar.

[0020] In some embodiments, the HCl flow rates l1 and l2 satisfy the relationship l1 < l2, and 1 slm < l1 < 5 slm, l2 < 5 slm.

[0021] In some embodiments, the growth temperature T = T1 remains unchanged in steps (1)-(6), and its selection range is 1500°C < T1 < 1700°C.

[0022] In some embodiments, in step (1), 40 slm < h1 < 160 slm.

[0023] The beneficial effects of the present invention are as follows:

[0024] (1) Propose an efficient and reasonable scheme for adjusting the process parameters of trench-structured silicon carbide epitaxial growth, and optimize and adjust five process parameters including hydrogen, reaction gas, hydrogen chloride, carbon-silicon ratio, and pressure in sequence, avoiding repeated conditions and cyclic optimization;

[0025] (2) Optimizing and regulating the growth process to avoid severe over-etching of the mesa and the formation of voids by trench closure is conducive to obtaining process conditions that are compatible with both filling rate and filling quality;

[0026] (3) A planar silicon carbide test piece was used to blindly adjust the reaction gas and hydrogen chloride parameters, which can effectively reduce the amount of trench structure substrates used and the number of destructive tests used, which is conducive to controlling mass production costs; BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic structural diagram of the silicon carbide epitaxial layer having both a trench structure and a planar structure according to the present invention.

[0028] Figure 2 Shows that with decreasing H2 flow rate, R G-Bare 、R G-MT and R G-TB Rate change graph.

[0029] Figure 3 It shows that with increasing the reaction gas flow rate (fixed C / Si=1), R G-Bare Rate change graph.

[0030] Figure 4 It shows that with increasing HCl flow rate, R G-Bare Rate change graph.

[0031] Figure 5 Shows that with decreasing C / Si (fixed ethylene flow rate), R G-MT and R G-TB Rate change graph.

[0032] Figure 6 Shows that with increasing reaction chamber pressure, R G-MT and R G-TB Rate change graph.

[0033] Figure 7 Optical spectrum showing the trench structure before silicon carbide epitaxial filling.

[0034] Figure 8 Optical spectrum showing the trench structure after silicon carbide epitaxial filling.

[0035] Figure 9 The SEM cross-section shows the trench structure silicon carbide epitaxial filling after 3.5 hours. DETAILED DESCRIPTION

[0036] The following will clearly and completely describe the concept and technical effects of the present invention in conjunction with the embodiments to fully understand the purpose, features and effects of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, other embodiments obtained by those skilled in the art without creative work are all within the scope of protection of the present invention.

[0037] Some embodiments provide a method for epitaxial filling of trench-structured silicon carbide, comprising the steps of:

[0038] (1) A silicon carbide growth test piece having both a trench structure and a planar structure epitaxial layer is prepared, referred to as a first growth test piece. The first growth test piece is subjected to a chemical vapor deposition (CVD) epitaxial growth test in a reaction system of trichlorosilane (SiHCl3, TCS), ethylene (C2H4), hydrogen (H2) and hydrogen chloride (HCl), maintaining a carbon-silicon ratio C / Si=1, a fixed flow rate TCS=t1, C2H4=c1, HCl=l1, a reaction chamber pressure P=p1, and a growth temperature T=T1, and adjusting the flow rate of H2 to obtain H2=h1, so as to simultaneously satisfy the growth rate R of the planar surface of the first growth test piece. G-Bare Greater than 0 and the growth rate R of the trench terrace G-MT Less than 0, the growth rate R of the bottom of the groove is obtained G-TB =v1, and the growth rate R of the plane surface is obtained G-Bare and the growth rate R of the trench terrace G-MT The difference is v0;

[0039] (2) A silicon carbide growth test piece containing only a planar structure epitaxial layer is provided, referred to as a second growth test piece, and a chemical vapor deposition epitaxial growth test is performed on the second growth test piece, maintaining a carbon-silicon ratio C / Si=1, a fixed flow rate H2=h1, HCl=l1, a pressure P=p1, and a growth temperature T=T1, and adjusting the flow rates of TCS and C2H4 to obtain TCS=t2 and C2H4=c2 to satisfy the epitaxial growth rate R of the second growth test piece at this time. G-Bare In the 2v0 to 4v0 range;

[0040] (3) A silicon carbide growth test piece containing only a planar structure epitaxial layer is provided, referred to as a third growth test piece, and a chemical vapor deposition epitaxial growth test is performed on the third growth test piece. The flow rate TCS = t2, C2H4 = c2, H2 = h1, the pressure P = p1, and the growth temperature T = T1 are fixed, and the flow rate of HCl is adjusted to obtain HCl = l2 to meet the epitaxial growth rate R of the third growth test piece at this time. G-Bare In the range of v0 / 2 to v0;

[0041] (4) A test piece identical to the formal silicon carbide piece with a groove structure to be filled is provided, which is called the fourth growth test piece. The fourth growth test piece is subjected to a chemical vapor deposition epitaxial growth test, with fixed flow rates C2H4=c2, H2=h1, HCl=l2, pressure P=p1, and growth temperature T=T1. The flow rate of TCS is adjusted to obtain TCS=t3, so as to satisfy the groove bottom growth rate R of the fourth growth test piece at this time. G-TB =v2, trench terrace growth rate R G-MT =v3, and 0.8v2 <v3<1.2v2;

[0042] (5) A test piece identical to the formal silicon carbide piece with a groove structure to be filled is provided, which is called the fifth growth test piece. The fifth growth test piece is subjected to a chemical vapor deposition epitaxial growth test, with fixed flow rates TCS=t3, C2H4=c2, H2=h1, HCl=l2 and growth temperature T=T1. The flow rate of pressure P is adjusted within a certain range to obtain P=p2, so as to satisfy the growth rate R of the bottom of the groove of the fifth growth test piece at this time. G-TB =v4, Groove terrace growth rate R G-MT =v5, and 0 <v5<v4 / 3;

[0043] (6) Using the trench structured silicon carbide wafer, using the process parameters determined in steps (1)-(5): flow rate TCS=t3, C2H4=c2, H2=h1, HCl=l2 and pressure P=p2, growth temperature T=T1, chemical vapor deposition epitaxial growth filling is performed, and the epitaxial filling time S≥L / v4, where L is the depth of the trench structure, that is, the epitaxial filling of the trench structure is completed;

[0044] (7) After the epitaxial filling growth is completed, the excess silicon carbide epitaxial layer with a thickness of H is removed by grinding and polishing, where H≥v5×S.

[0045] Please refer to Figure 1 , providing a silicon carbide growth test piece with an epitaxial layer having both a trench structure and a planar structure, wherein the depth, width, doping type, doping concentration of the trench structure and the doping type and doping concentration of the planar structure are the same as those of the silicon carbide formal piece with a trench structure to be filled, and at the same time, the planar structure surface of the silicon carbide growth test piece is consistent with the height of its trench structure table.

[0046] It should be noted that in step (1), when optimizing the hydrogen flow parameters, a silicon carbide test piece with both a groove structure and a planar structure epitaxial layer was used for debugging. By gradually reducing the hydrogen flow rate, R G-Bare and R G-MT While decreasing, R G-TB On the contrary, it will be improved to a certain extent. Make the groove structure table in a slightly excessive corrosion state (R G-MT <0 and RG-Barer > 0), obtaining R G-Bare and R G-MT The rate difference v0 between them is used to estimate the blind adjustment of the reaction gas flow rate and hydrogen chloride flow rate parameters later.

[0047] In step (2), a silicon carbide growth test piece with only a planar structure is used to blindly adjust the reaction gas parameters. Keeping the carbon-silicon ratio C / Si = 1, while gradually increasing the flow rates of trichlorosilane and ethylene, making the rate of R G-Bare greater than 2v0 and less than 4v0. R G-Bare 、R G-MT and R G-TB increase simultaneously as the reaction gas flow rate increases, but the growth degree of R G-Bare 、R G-MT will be significantly greater than that of R G-TB . According to experience, when R G-Bare is greater than 2v0, the trench structure mesa can get out of the over-etching state, and R G-Bare being less than 4v0 is to prevent the mesa rate from being too fast and causing the trench to close. [[ID=2,7]]

[0048] In step (3), a silicon carbide growth test piece with only a planar structure is used to blindly adjust the hydrogen chloride parameters. By gradually increasing the hydrogen chloride flow rate, making R G-Bare drop to greater than v0 / 2 and less than v0. R G-Bare 、R G-MT and R G-TB decrease simultaneously as the hydrogen chloride flow rate increases, but the decrease amplitude of R G-Bare 、R G-MT will be significantly greater than that of R G-TB . The purpose of R G-Bare being less than v0 is to make the trench structure mesa enter the slightly over-etched state or low growth rate state again, and R G-Bare being greater than v0 / 2 is to avoid severe over-etching of the mesa and causing damage to the trench structure.

[0049] In step (4), a silicon carbide growth test piece with a trench structure is used to adjust the carbon-silicon ratio. Fixing the ethylene flow rate and increasing the trichlorosilane flow rate to reduce the carbon-silicon ratio, making 0.8v2 < v3 < 1.2v2. R G-MT and R G-TB increase simultaneously as the carbon-silicon ratio decreases, but the increase amplitude of R G-MT will be significantly greater than that of R G-TB , making 0.8v2 < v3 < 1.2v2 can ensure that while increasing the growth rate R G-TB of the trench bottom, preventing the mesa rate R G-MT from being too large and causing the trench to close.

[0050] In step (5), the carbon-silicon ratio is adjusted by using a grooved silicon carbide growth test piece and increasing the pressure of the reaction chamber so that 0 < R G-MT <R G-TB / 3. By increasing the pressure of the reaction chamber, R G-MT While decreasing, R G-TB On the contrary, it will be improved to a certain extent, making 0<R G-MT <R G-TB / 3, under the premise that there is no excessive etching on the terrace, the bottom of the groove obtains a higher growth rate.

[0051] In step (6), according to the previously determined parameters such as trichlorosilane, ethylene, hydrogen, hydrogen chloride flow rate and reaction chamber pressure, a trench structured silicon carbide wafer is used for CVD epitaxial growth filling, and the epitaxial filling time is controlled to completely fill the trench.

[0052] In step (7), the excess thickness of the silicon carbide epitaxial layer grown on the mesa is removed by polishing.

[0053] The following are specific embodiments, but they do not limit the scope of protection of the present invention.

[0054] Example 1

[0055] The silicon carbide substrate for the trench structure to be filled has a depth of about L = 19 μm, an opening width of about W = 3.5 μm, an n-type doping concentration of 2e 16 cm -3 The 4-inch trench structure silicon carbide epitaxial wafer is epitaxially filled using the following method. The epitaxial filling doping material is trimethylaluminum (P type) and the Al atomic doping concentration is 1e 16 cm -3 .

[0056] A method for epitaxial filling of trench structured silicon carbide, comprising the steps of:

[0057] (1) Inductively coupled plasma (ICP) etching was used to produce a silicon carbide growth test piece with both a trench structure and a planar structure epitaxial layer. The epitaxial layer was n-type doped with a doping concentration of 2e 16 cm -3 The depth of the groove structure is about 19μm, the opening width is about 3.5μm, and the height of its planar structure surface is consistent with that of its groove structure table, which is called the first growth test piece. Figure 1 As shown,

[0058] The first growth test piece was subjected to a chemical vapor deposition (CVD) epitaxial growth test in a reaction system of trichlorosilane (SiHCl3, TCS), ethylene (C2H4), hydrogen (H2) and hydrogen chloride (HCl). Under the premise of maintaining the carbon-silicon ratio C / Si=1, the flow rates TCS=40sccm, C2H4=15sccm, HCl=2slm, the reaction chamber pressure P=200mbar, and the growth temperature T=1650℃ were fixed. The H2 flow rate was reduced from 100slm, 80slm and 60slm in sequence, and a multi-furnace epitaxial growth test was performed. After each growth, the epitaxial wafer was split and the growth rate at each position was determined using cross-sectional SEM images. The results are shown in Figure 2. Figure 2 shown.

[0059] from Figure 2 It can be seen that when H2=60slm, the growth rate of the plane surface (R G-Bare =1.4μm / h) is greater than 0, and the trench terrace growth rate (R G-MT =-0.8μm / h) is less than 0, at this time the groove bottom growth rate R G-TB = v1, v1 = 0.8 μm / h, recording the growth rate difference R between the flat surface and the grooved terrace G-Bare -R G-MT =v0=2.2μm / h.

[0060] (2) Using n-type doping, the doping concentration is 2e 16 cm -3 , a CVD epitaxial growth test was conducted on a silicon carbide growth test piece with only a planar structure epitaxial layer (called the second growth test piece). With fixed flow rate H2 = 60slm, HCl = 2slm and pressure P = 200mbar, growth temperature T = 1650℃, and under the premise of maintaining the carbon-silicon ratio C / Si = 1, the flow rate of TCS (the corresponding C2H4 flow rate is 15sccm, 22.5sccm and 30sccm respectively) and the flow rate of ethylene were increased from 40sccm, 59sccm and 79sccm in sequence, and a multi-furnace epitaxial growth test was conducted. After each growth was completed, a Fourier transform infrared spectrometer (FTIR) was used to detect the growth thickness to convert the planar surface growth rate R G-Bare , the results are as follows Figure 3 shown.

[0061] from Figure 3 It can be seen that when TCS = 79 sccm and C2H4 = 30 sccm, R G-Bare =5.5μm / h, within the range of (2v0,4v0)=(4.4μm / h,8.8μm / h).

[0062] (3) Using n-type doping, the doping concentration is 2e 16 cm-3 , a CVD epitaxial growth test was conducted on a silicon carbide growth test piece with only a planar structure epitaxial layer (called the third growth test piece). Fixed flow rate TCS = 79sccm, C2H4 = 30sccm, H2 = 60slm and pressure P = 200mbar, growth temperature T = 1650℃, and increasing the flow rate of HCl from 2.0slm, 2.1slm, 2.2slm, 2.3slm and 2.35slm in sequence, and a multi-furnace epitaxial growth test was conducted. After each growth was completed, the thickness was detected by FTIR to convert R G-Bare , the results are as follows Figure 4 shown.

[0063] from Figure 4 It can be seen that when HCl=2.35slm, R G-Bare =2.1μm / h, and its value is in the range of (v0 / 2,v0)=(1.1μm / h,2.2μm / h).

[0064] (4) A CVD epitaxial growth test was conducted using a test piece identical to the formal silicon carbide piece with a groove structure to be filled (referred to as the fourth growth test piece). The flow rates C2H4 = 30 sccm, H2 = 60 slm, HCl = 2.35 slm, pressure P = 200 mbar, and growth temperature T = 1650°C were fixed. The flow rate of TCS was increased from 79 sccm, 88 sccm, 93 sccm, and 99 sccm in sequence, so that the C / Si ratio was reduced to 1, 0.9, 0.85, and 0.80 in sequence. Multiple-furnace epitaxial growth tests were conducted. After each growth, the cross-section SEM image was detected after the piece was split to determine the R G-TB and R G-MT , the results are as follows Figure 5 shown.

[0065] from Figure 5 It can be seen that when TCS = 99 sccm (C / Si = 0.8), R G-TB =v2=3.0μm / h, R G-MT =v3=3.5μm / h, meeting 0.8v2 <v3<1.2v2。

[0066] (5) A CVD epitaxial growth test was conducted using the same test piece (fifth growth test piece) as the formal silicon carbide piece with the groove structure to be filled. The flow rate TCS = 99 sccm, C2H4 = 30 sccm, H2 = 60 slm, HCl = 2.35 slm and the growth temperature T = 1650 ° C were fixed. The pressure P was increased from 200 mbar, 300 mbar and 400 mbar in sequence. Multiple furnace epitaxial growth tests were conducted. After each growth, the cross-section SEM image was detected after the piece was split to determine the R G-TB and R G-MT , the results are as follows Figure 6shown.

[0067] from Figure 6 It can be seen that when P = 400mbar, R G-TB =v4=5.3μm / h, R G-MT =v5=1.7μm / h, satisfying v5 is in the range of (0,v4 / 3).

[0068] (6) Using the trench structured silicon carbide wafer, the process parameters determined in steps (1) to (5) are as follows: flow rate TCS = 99 sccm, C2H4 = 30 sccm, H2 = 60 slm, HCl = 2.35 slm, pressure P = 400 mbar, growth temperature T = 1650 ° C, and formal CVD epitaxial filling production is carried out. The epitaxial filling time S = L / v4 = 19 μm / (5.3 μm / h) ≈ 3.6 h is controlled to complete the epitaxial filling of the trench structure. The results are as follows: Figure 7-9 As shown, from Figure 7-9 It can be seen that the present invention achieves gap-free filling.

[0069] (7) Grind and polish away the excess silicon carbide epitaxial layer with a thickness H of v5×S=1.7μm / h×3.6h=6.12μm.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the scope of protection of the present invention. Although the present invention is described in detail with reference to the preferred embodiments, it is not limited to those listed in the embodiments. Those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A method for epitaxial filling of trench structured silicon carbide, characterized in that: Including steps: (1) A silicon carbide growth test piece having both a trench structure and a planar structure epitaxial layer is prepared, which is called the first growth test piece. The first growth test piece is subjected to a chemical vapor deposition epitaxial growth test in a TCS, C2H4, H2 and HCl reaction system, maintaining the carbon-silicon ratio C / Si=1, fixing the TCS flow rate=t1, C2H4 flow rate=c1, HCl flow rate=l1 and the reaction chamber pressure P=p1, growth temperature T=T1, and adjusting the H2 flow rate to obtain the H2 flow rate=h1, so as to simultaneously meet the growth rate R of the planar surface of the first growth test piece. G-Bare Greater than 0 and the growth rate R of the trench terrace G-MT Less than 0, the growth rate R of the bottom of the groove is obtained G-TB = v1, giving the growth rate R on the planar surface G-Bare and the growth rate R of the trench terrace G-MT The difference is v0; (2) A silicon carbide growth test piece containing only a planar structure epitaxial layer is provided, which is called the second growth test piece. The second growth test piece is subjected to a chemical vapor deposition epitaxial growth test, maintaining the carbon-silicon ratio C / Si=1, fixing the H2 flow rate=h1, HCl flow rate=l1 and pressure P=p1, and growth temperature T=T1, and adjusting the flow rates of TCS and C2H4 to obtain TCS flow rate=t2 and C2H4 flow rate=c2, so as to meet the epitaxial growth rate R of the second growth test piece at this time. G-Bare In the range of 2 v0 to 4 v0; (3) A silicon carbide growth test piece containing only a planar structure epitaxial layer is provided, referred to as a third growth test piece, and a chemical vapor deposition epitaxial growth test is performed on the third growth test piece. The TCS flow rate = t2, C2H4 flow rate = c2, H2 flow rate = h1, pressure P = p1, and growth temperature T = T1 are fixed, and the HCl flow rate is adjusted to obtain the HCl flow rate = l2 to meet the epitaxial growth rate R of the third growth test piece at this time. G-Bare In the range of v0 / 2 to v0; (4) Provide a test piece identical to the formal silicon carbide piece with a groove structure to be filled, called the fourth growth test piece, and perform a chemical vapor deposition epitaxial growth test on the fourth growth test piece. Fix the C2H4 flow rate = c2, H2 flow rate = h1, HCl flow rate = l2 and pressure P = p1, growth temperature T = T1, and adjust the TCS flow rate to obtain TCS flow rate = t3 to meet the groove bottom growth rate R of the fourth growth test piece at this time. G-TB =v2, trench terrace growth rate R G-MT =v3, and 0.8v2 <v3<1.2v2; (5) Provide a test piece identical to the formal silicon carbide piece with a groove structure to be filled, called the fifth growth test piece, and perform a chemical vapor deposition epitaxial growth test on the fifth growth test piece. Fix TCS flow rate = t3, C2H4 flow rate = c2, H2 flow rate = h1, HCl flow rate = l2 and growth temperature T = T1, and adjust the pressure P within a certain range to obtain P = p2 to meet the growth rate R of the bottom of the groove of the fifth growth test piece at this time. G-TB =v4, Groove terrace growth rate R G-MT =v5, and 0 <v5<v4 / 3; (6) Using the trench structured silicon carbide wafer, using the process parameters determined in steps (1)-(5): TCS flow rate = t3, C2H4 flow rate = c2, H2 flow rate = h1, HCl flow rate = l2 and pressure P = p2, growth temperature T = T1, chemical vapor deposition epitaxial growth filling is performed, and the epitaxial filling time S ≥ L / v4, where L is the depth of the trench structure, that is, the epitaxial filling of the trench structure is completed; (7) After the epitaxial filling growth is completed, the excess silicon carbide epitaxial layer with a thickness of H is removed by grinding and polishing, where H≥v5×S.

2. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: The method for maintaining the carbon-silicon ratio C / Si=1 in step (1) is as follows: The temperature of the TCS liquid source was maintained at 19°C, and the saturated vapor pressure of TCS was maintained at 1500 mbar by controlling the flow rate of the H2 carrier gas entering and exiting the reaction chamber, so that the flow rate ratio of TCS to C2H4 entering the reaction chamber satisfied the formula (c1×2) / (t1×0.758)=1.

3. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: The method for maintaining the carbon-silicon ratio C / Si=1 in step (2) is as follows: The temperature of the TCS liquid source was maintained at 19°C, and the saturated vapor pressure of TCS was maintained at 1500 mbar by controlling the flow rate of the H2 carrier gas entering and exiting the reaction chamber, so that the flow rate ratio of TCS to C2H4 entering the reaction chamber satisfied the formula (c2×2) / (t2×0.758)=1.

4. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: TCS flow rates t1, t2, and t3 satisfy the relationship t1 < t2 < t3, and 40 sccm <t1<500 sccm,t2<500 sccm,t3<500 sccm。 5. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: The pressures p1 and p2 satisfy the relationship p1 < p2, and 70 mbar < p1 < 700 mbar, p2 < 700 mbar.

6. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: The HCl flow rates l1 and l2 satisfy the relationship l1< l2, and 1 slm< l1<5 slm, l2<5 slm.

7. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: The growth temperature T = T1 remains unchanged in steps (1) to (6), and its selection range is 1500 ℃ < T1 < 1700 ℃.

8. The epitaxial filling method of trench structured silicon carbide according to claim 1, characterized in that: In step (1), 40 slm< h1< 160 slm.

Images