Method for determining the distribution of charge balance state in a superjunction device

By measuring and correcting the on-resistance in the superjunction device and calculating the charge equilibrium state, the problem of difficulty in judging the in-plane charge equilibrium of superjunction devices in the prior art is solved, and accurate matching analysis and process optimization of the superjunction structure are achieved.

CN114023666BActive Publication Date: 2024-11-12SHANGHAI HUAHONG GRACE SEMICON MFG CORP
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202111210619.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-10-18
Publication Date
2024-11-12
Estimated Expiration
2041-10-18

AI Technical Summary

Technical Problem

The prior art is difficult to accurately judge the in-plane charge equilibrium state distribution of superjunction devices, resulting in poor matching of reverse breakdown voltages, affecting device performance.

Method used

By measuring the on-resistance of N-type and P-type columns at different depths, and using the resistance correction coefficient to correct, the charge equilibrium state of the superjunction unit is calculated and the in-plane charge matching situation is quantitatively analyzed.

Benefits of technology

It realizes accurate judgment of the charge matching condition of the superjunction structure in the plane, quickly identify process differences, select excellent process conditions, and improve device performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114023666B_ABST
    Figure CN114023666B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for determining the distribution of the charge balance state in a super junction device, comprising: step 1, obtaining the first and second resistance correction coefficients of N-type columns and P-type columns of different depths; step 2, measuring the first and second on-resistances corresponding to the N-type columns and P-type columns of the super junction unit at different positions on the wafer; step 3, calculating the charge balance state of the super junction unit at each position, comprising: multiplying the first on-resistance by the first resistance correction coefficient to obtain the first corrected on-resistance, multiplying the second on-resistance by the second resistance correction coefficient to obtain the second corrected on-resistance; taking the inverse of the first and second corrected on-resistances respectively and normalizing them; dividing the normalized second corrected conductance by the normalized first corrected conductance to obtain the charge balance state of the super junction unit. The present invention can quantitatively analyze the charge matching in the plane of the super junction structure.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a method for manufacturing a semiconductor integrated circuit, and in particular to a method for determining the distribution of an in-plane charge balance state of a super junction device. Background Art

[0002] Super junction products are a type of MOSFET structure that utilizes the PN charge balance internal surface field reduction (Resurf) technology to improve the reverse breakdown (BV) of the device while maintaining a small on-resistance. The pillar structure of the PN interval is the biggest feature of the super junction. Currently, there are two main methods for making the pillar structure of the PN interval, one is to obtain it through multiple epitaxy and ion implantation, and the other is to make it through deep trench etching and EPI filling.

[0003] When manufacturing super junction devices through deep trench etching and filling process, the reverse breakdown voltage of the device is very sensitive to the total doping amount matching of the P-type region and the N-type region, so it is very critical to accurately control the total doping amount of the two regions. However, in the actual process, due to the in-plane differences in the opening size and angle of the trench (TH) caused by lithography and etching, it is always difficult for the P-type region and the N-type region to achieve the best match in the plane at the same time, resulting in poor in-plane distribution of reverse breakdown voltage in wafer testing (CP).

[0004] An important task in the development of super junction structure is to improve the PN charge balance in-plane matching relationship. There are two main means: one is to improve lithography, improve hard mask layer (HM) etching, improve TH etching and other etching processes, and try to ensure that the etching morphology and volume in-plane distribution are uniform; the other is to improve the total concentration distribution of epitaxial (EPI) filling to match the existing groove volume morphology in-plane distribution state. In this way, the PN matching state can be synchronized as much as possible, ensuring that the performance of devices such as BV and EAS meets the standards. Among them, the EAS capability will deteriorate rapidly when it is undermatched.

[0005] How to effectively evaluate the improvement plan is very important for selecting the development direction and process plan. Generally, the final BV is selected as the basis for judgment. However, BV as a criterion has many interference factors and limitations. BV is only useful near the top of the deviation, and is affected by design factors. Even if the same set of processes has different design judgment results, there are differences. There is also the uncertainty factor of BV, which is relatively not very direct as a process criterion. In addition, BV and matching status are in a parabolic relationship. The same BV value may be over-matched or under-matched, which interferes with the judgment.

[0006] like Figure 1A, which is a schematic diagram of the structure of an existing relatively large super junction structure; an N-type epitaxial layer 102 is formed on a semiconductor substrate 101, and a P-type column 103a formed by trench filling is formed in the N-type epitaxial layer 102.

[0007] like Figure 1B As shown, Figure 1A A schematic structural diagram of a super junction structure of a smaller volume formed on the same semiconductor substrate; Figure 1B The P-type column in the semiconductor substrate 101 is marked with a mark 103b. The P-type columns 103a and 103b are both formed on the semiconductor substrate 101, but they are located at different positions of the semiconductor substrate 101, which makes the volumes of the two different, and the PN mismatch parameters of the super junction units corresponding to the two are also different.

[0008] like Figure 2 As shown, it is a curve 104 of the breakdown voltage of the existing super junction structure changing with PN mismatch. It can be seen that the curve 104 is a parabola. Figure 1A The corresponding super junction unit is overmatched, that is, the P-type doping amount of the P-type column 103 a is greater than the N-type doping amount of the N-type column 102 , and the breakdown voltage is located at the position of the dotted circle 105 a on the right side of the vertex. Figure 1B The corresponding super junction unit is undermatched, that is, the P-type doping amount of the P-type column 103b will be less than the N-type doping amount of the N-type column 102, and the breakdown voltage will be located at the position of the dotted circle 105b on the left side of the vertex. Therefore, it can be seen from the curve 104 that the PN matching of the super junction unit cannot be determined by the breakdown voltage, because the same breakdown voltage will correspond to both overmatching and undermatching. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide a method for determining the distribution of the in-plane charge balance state of a super junction device, which can quantitatively analyze the in-plane charge matching of the super junction structure.

[0010] In order to solve the above technical problems, the present invention provides a method for determining the distribution of the in-plane charge balance state of a super junction device, comprising the following steps:

[0011] Step 1: Obtain a first resistance correction coefficient of an N-type column of different depths and a second resistance correction coefficient of a P-type column of different depths in a super junction structure.

[0012] The P-type columns of the super junction structure are composed of a P-type epitaxial layer filled in a trench, the trench is formed in an N-type epitaxial layer, and the N-type columns are composed of the N-type epitaxial layer located between the P-type columns.

[0013] The N-type epitaxial layer is formed on a semiconductor substrate. A plurality of trenches are provided on the semiconductor substrate. The trenches are formed simultaneously by the same process. Due to process deviation, the depths of the trenches on the semiconductor substrate are uneven.

[0014] A super junction unit is formed by one of the N-type columns and one of the adjacent P-type columns, and the depth of the N-type column and the depth of the P-type column of the super junction unit are both the depth of the trench at the super junction unit.

[0015] The first resistance correction coefficient corrects the first on-resistance of the N-type column to a first corrected on-resistance, which is the on-resistance when the depth of the N-type column is adjusted to a first reference depth while maintaining the charge distribution of the N-type column.

[0016] The second resistance correction coefficient corrects the second on-resistance of the P-type column to a second corrected on-resistance, which is the on-resistance when the depth of the P-type column is adjusted to the first reference depth while maintaining the charge distribution of the P-type column.

[0017] Step 2: measuring the first on-resistance of the N-type column and the second on-resistance of the P-type column of the super junction unit at different positions on the wafer.

[0018] Step 3: Calculate the charge balance state of the super junction unit at each position and thereby form an in-plane charge balance state distribution of the super junction device. The step of calculating the charge balance state of the super junction unit at each position includes:

[0019] The first corrected on-resistance is obtained by multiplying the first on-resistance by the first resistance correction coefficient corresponding to the depth of the super junction unit, and the second corrected on-resistance is obtained by multiplying the second on-resistance by the second resistance correction coefficient corresponding to the depth of the super junction unit.

[0020] The first modified on-resistance is taken as the inverse to form a first modified conductance and the first modified conductance is normalized. The second modified on-resistance is taken as the inverse to form a second modified conductance and the first modified conductance is normalized.

[0021] The charge balance state of the super junction unit is obtained by dividing the normalized second corrected conductance by the normalized first corrected conductance.

[0022] A further improvement is that step one further includes dividing the first resistance correction coefficient corresponding to each depth by the second resistance correction coefficient to obtain a third resistance correction coefficient.

[0023] The calculation step of the charge balance state of the super junction unit at each position in step 3 is modified as follows:

[0024] The first on-resistance is taken as the inverse to form a first conductance and is normalized, and the second on-resistance is taken as the inverse to form a second conductance and is normalized.

[0025] The normalized second conductance is divided by the normalized first conductance and then multiplied by the third resistance correction coefficient to obtain a charge balance state of the super junction unit.

[0026] A further improvement is that the semiconductor substrate comprises a silicon substrate.

[0027] A further improvement is that the semiconductor substrate is in a wafer structure.

[0028] A further improvement is that the distribution structure of the depth of the groove on the semiconductor substrate includes:

[0029] A uniform first depth is formed in a middle region of the semiconductor substrate.

[0030] The depth of the trench gradually increases in a direction from the edge of the middle region of the semiconductor substrate to the edge of the wafer.

[0031] A further improvement is that the first reference depth is set to the first depth.

[0032] A further improvement is that the normalization in step three is normalized to the middle area.

[0033] A further improvement is that the step of obtaining the first resistance correction coefficient includes:

[0034] The deeper the depth of the N-type column, the greater the first on-resistance is, and a first changed resistance of an extension portion of the N-type column having a thickness whose first depth difference between the depth of the N-type column and the first depth is calculated.

[0035] The first resistance correction coefficient is obtained by subtracting the first change resistance from the first on-resistance of the N-type column and dividing the result by the first on-resistance.

[0036] A further improvement is that the depth of the groove increases linearly and gradually in a direction from the edge of the middle area of ​​the semiconductor substrate to the edge of the wafer, and the step of obtaining the first resistance correction coefficient of the N-type column at each depth includes:

[0037] The first resistance correction coefficient of the N-type column with the greatest depth is calculated, where the N-type column with the greatest depth is located closest to the edge of the wafer, and a first maximum correction coefficient deviation value is obtained by subtracting the first resistance correction coefficient of the N-type column with the greatest depth from 1.

[0038] The first depth difference of the N-type column with the largest depth is set as the first maximum depth difference.

[0039] The first correction coefficient deviation value of each N-type column is calculated by using the ratio of the first maximum correction coefficient deviation value to the first maximum depth difference value being equal to the relationship between the first correction coefficient deviation value and the first depth difference value of each N-type column.

[0040] The first resistance correction coefficient of each N-type column is obtained by subtracting the first correction coefficient deviation value of each N-type column from 1.

[0041] A further improvement is that the step of obtaining the second resistance correction coefficient includes:

[0042] The deeper the depth of the P-type column, the smaller the second on-resistance is, and a second change resistance generated by an extended portion of the P-type column having a thickness equal to a first depth difference between the depth of the P-type column and the first depth is calculated;

[0043] The second resistance correction coefficient is obtained by adding the second change resistance to the second on-resistance of the P-type column and dividing the result by the second on-resistance.

[0044] A further improvement is that the depth of the groove increases linearly from the edge of the middle area of ​​the semiconductor substrate to the edge of the wafer, and the step of obtaining the second resistance correction coefficient of the P-type column at each depth includes:

[0045] Calculating the second resistance correction coefficient of the P-type column with the greatest depth, where the P-type column with the greatest depth is located closest to the edge of the wafer, and subtracting 1 from the second resistance correction coefficient of the P-type column with the greatest depth to obtain a second maximum correction coefficient deviation value;

[0046] Let the first depth difference of the P-type column with the largest depth be the first maximum depth difference;

[0047] The second correction coefficient deviation value of each P-type column is calculated by using the ratio of the second maximum correction coefficient deviation value to the first maximum depth difference value to be equal to the relationship between the second correction coefficient deviation value and the first depth difference of each P-type column;

[0048] The second resistance correction coefficient of each P-type column is obtained by adding 1 to the second correction coefficient deviation value of each P-type column.

[0049] A further improvement is that the first on-resistance of each of the N-type columns is obtained by testing the super junction device formed on the super junction unit corresponding to the N-type column.

[0050] A further improvement is that the super junction device includes a super junction MOSFET.

[0051] A further improvement is that a test structure for the second on-resistance of the P-type column at each depth is provided on the semiconductor substrate, and each of the test structures comprises:

[0052] A plurality of P-type pseudo pillars are provided, wherein the N-type pseudo pillars are arranged between each of the P-type pseudo pillars, and the depth of each of the P-type pseudo pillars is equal to the depth of the P-type pillar to be tested.

[0053] Contact holes are formed at both ends of the middle P-type pseudo column, and the middle P-type pseudo column is tested through the two contact holes to obtain the second on-resistance of the P-type columns with the same depth.

[0054] A further improvement is that each of the test structures is arranged on a cutting path.

[0055] A further improvement is that the length of the P-type dummy gate in each of the test structures is greater than 5 times the depth.

[0056] A further improvement is that the number of the P-type dummy gates in each of the test structures is 5.

[0057] The present invention utilizes the characteristics that the first on-resistance of an N-type column is related to the total charge of the N-type column and the second on-resistance of a P-type column is related to the total charge of the P-type column, and also utilizes the characteristic that the charge amount involved in matching in N-type columns and P-type columns of different depths is related to the depth, and obtains in advance the first resistance correction coefficient of N-type columns of different depths and the second resistance correction coefficient of P-type columns of different depths. In this way, by utilizing the characteristic that the first on-resistance and the second on-resistance can be directly obtained by measurement, the superjunction units at various depths in the plane of the superjunction device can be measured to obtain the corresponding first on-resistance and second on-resistance. By using the first resistance correction coefficient and the second resistance correction coefficient to correct the first on-resistance and the second on-resistance, the charge amount involved in charge matching in the total charge of the N-type column and the P-type column can be obtained, so that finally the charge matching conditions of the superjunction units at different depths can be obtained, and finally the charge matching conditions in the plane of the superjunction structure can be quantitatively analyzed.

[0058] By using the quantitative analysis results of the in-plane charge matching of the super junction structure, it is possible to quickly and accurately determine the quality differences of super junction structures formed by different processes, and thus select a better process condition to form the super junction structure.

[0059] Compared with the existing method of using breakdown voltage to judge the in-plane charge matching of the super junction structure, the present invention does not need to rely on the complete accuracy of the pull-bias condition. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0061] Figure 1A It is a schematic diagram of the structure of the existing larger super junction structure;

[0062] Figure 1B Yes and Figure 1A A schematic structural diagram of a super junction structure of a smaller volume formed on the same semiconductor substrate;

[0063] Figure 2 This is the curve of the breakdown voltage of the existing super junction structure changing with PN mismatch;

[0064] Figure 3 It is a flow chart of a method for determining the distribution of the in-plane charge balance state of a super junction device according to an embodiment of the present invention;

[0065] Figure 4A is a top view of super junction units in different regions on the same semiconductor substrate in an embodiment of the present invention;

[0066] Figure 4B yes Figure 4A The corresponding curve of the mismatch coefficient of the super junction unit changing with the position;

[0067] Figure 5 1 is a summary diagram of various factors affecting the PN mismatch coefficient of the super junction unit in an embodiment of the present invention;

[0068] Figure 6 It is a schematic diagram of associating the total N-type charge (QSN) of the N-type column and the total P-type charge (QSP) of the P-type column of the super junction unit with the first on-resistance (Ron) and the second on-resistance (RS-Pillar) respectively in an embodiment of the present invention;

[0069] Fig. 7A and Figure 7B It is a schematic diagram of the structure of the N-type column when only the concentration of the N-type column changes in determining the relationship between QSN and Ron;

[0070] Fig. 8A and Figure 8B Is to determine the relationship between QSN and Ron Fig. 7A and Figure 7B A schematic diagram of the structure of an N-type column when a change in the critical dimension (CD) of the N-type column is added on the basis of FIG.

[0071] Fig.9A and Fig. 9B Is to determine the relationship between QSN and Ron Fig. 7A and Figure 7B A schematic diagram of the structure of the N-type column when the side angle of the N-type column is changed on the basis of FIG.

[0072] Fig. 10A and Fig. 10B Is to determine the relationship between QSN and Ron Fig. 8A and Figure 8B A schematic diagram of the structure of the N-type column when the depth variation of the N-type column is added on the basis of FIG.

[0073] Fig.11 is a cross-sectional view of a P-type column of a super junction unit in the middle region of an embodiment of the present invention and a corresponding parameter chart;

[0074] Fig.12 is a curve of the proportion of a 1 micron thick N-type epitaxial layer in the first on-resistance at different depths of a super junction unit in an embodiment of the present invention;

[0075] Fig.13 is a distribution curve of the depth of the super junction unit at different positions of the wafer in an embodiment of the present invention;

[0076] Fig.14 is a distribution curve of the first resistance correction coefficient at different positions of the wafer in an embodiment of the present invention;

[0077] Fig.15 is a test data summary diagram of the second on-resistance in an embodiment of the present invention;

[0078] Fig.16 is a distribution curve of the second resistance correction coefficient at different positions of the wafer in an embodiment of the present invention;

[0079] Fig.17A yes Fig.14 and Fig.16 A graph of the curves when put together;

[0080] Fig. 17B is a distribution curve of the third resistance correction coefficient at different positions of the wafer in the embodiment of the present invention;

[0081] Fig.18 4 is a top view of a test structure of a second on-resistance in an embodiment of the present invention. DETAILED DESCRIPTION

[0082] like Figure 3 , which is a flow chart of a method for determining the distribution of the in-plane charge balance state of a super junction device according to an embodiment of the present invention; the method for determining the distribution of the in-plane charge balance state of a super junction device according to an embodiment of the present invention comprises the following steps:

[0083] Step 1: Obtain a first resistance correction coefficient of the N-type column 201 of different depths and a second resistance correction coefficient of the P-type column 202 of different depths in the super junction structure.

[0084] like Figure 4AAs shown, the P-type pillars 202 of the super junction structure are composed of a P-type epitaxial layer filled in a trench, the trench is formed in an N-type epitaxial layer, and the N-type pillars 201 are composed of the N-type epitaxial layer located between the P-type pillars 202 .

[0085] The N-type epitaxial layer is formed on a semiconductor substrate. A plurality of trenches are provided on the semiconductor substrate. The trenches are formed simultaneously by the same process. Due to process deviation, the depths of the trenches on the semiconductor substrate are uneven.

[0086] A super junction unit is formed by one N-type column 201 and one adjacent P-type column 202 , and the depth of the N-type column 201 and the depth of the P-type column 202 of the super junction unit are both the depth of the trench at the super junction unit.

[0087] The first resistance correction coefficient corrects the first on-resistance of the N-type column 201 to a first corrected on-resistance, which is the on-resistance when the depth of the N-type column 201 is adjusted to a first reference depth while maintaining the charge distribution of the N-type column 201 .

[0088] The second resistance correction coefficient corrects the second on-resistance of the P-type column 202 to a second corrected on-resistance, which is the on-resistance when the depth of the P-type column 202 is adjusted to the first reference depth while maintaining the charge distribution of the P-type column 202 .

[0089] Step 2: measuring the first on-resistance of the N-type column 201 and the second on-resistance of the P-type column 202 of the super junction unit at different positions on the wafer.

[0090] Step 3: Calculate the charge balance state of the super junction unit at each position and thereby form an in-plane charge balance state distribution of the super junction device. The step of calculating the charge balance state of the super junction unit at each position includes:

[0091] The first corrected on-resistance is obtained by multiplying the first on-resistance by the first resistance correction coefficient corresponding to the depth of the super junction unit, and the second corrected on-resistance is obtained by multiplying the second on-resistance by the second resistance correction coefficient corresponding to the depth of the super junction unit.

[0092] The first modified on-resistance is taken as the inverse to form a first modified conductance and the first modified conductance is normalized. The second modified on-resistance is taken as the inverse to form a second modified conductance and the first modified conductance is normalized.

[0093] The charge balance state of the super junction unit is obtained by dividing the normalized second corrected conductance by the normalized first corrected conductance.

[0094] Preferably, step one further comprises dividing the first resistance correction coefficient corresponding to each depth by the second resistance correction coefficient to obtain a third resistance correction coefficient.

[0095] The calculation step of the charge balance state of the super junction unit at each position in step 3 is modified as follows:

[0096] The first on-resistance is taken as the inverse to form a first conductance and is normalized, and the second on-resistance is taken as the inverse to form a second conductance and is normalized.

[0097] The normalized second conductance is divided by the normalized first conductance and then multiplied by the third resistance correction coefficient to obtain the charge balance state of the super junction unit. After adopting the third resistance correction coefficient, the calculation process of step three can be further simplified, but the result is consistent.

[0098] The semiconductor substrate includes a silicon substrate.

[0099] The semiconductor substrate is in a wafer structure.

[0100] The distribution structure of the depth of the trench on the semiconductor substrate includes:

[0101] A uniform first depth is formed in a middle region of the semiconductor substrate.

[0102] The depth of the trench gradually increases in a direction from the edge of the middle region of the semiconductor substrate to the edge of the wafer. Fig.13 The curve 304 in FIG. 1 shows the distribution structure of the depth of the groove on the semiconductor substrate. Since the semiconductor substrate is a wafer, Fig.13 The depth of the groove on one diameter of the wafer is used to represent the distribution of the groove on the entire wafer.

[0103] like Fig.11 As shown, it is a cross-sectional view of the P-type column of the super junction unit in the middle area of ​​an embodiment of the present invention and a corresponding parameter chart; the parameter table 301 is displayed on the right side of the P-type column 202, and the parameters include Depth, Trench width and Trench Space, and the corresponding values ​​are listed for each parameter.

[0104] Preferably, the first reference depth is set to the first depth. The normalization in step 3 is normalized to the middle area. In other embodiments, other depths are used as the first reference depth, and finally normalization is performed to the area corresponding to the first reference depth, which will not be described in detail here.

[0105] The step of obtaining the first resistance correction coefficient comprises:

[0106] The deeper the depth of the N-type column 201, the greater the first on-resistance. Fig. 10A and Fig. 10B As shown, the positive and negative electrodes added to the first on-resistance of the N-type column 201, namely the "+" electrode and the "-" electrode, are respectively located at the bottom and the top of the N-type column 201, so a voltage is applied longitudinally to the N-type column 201. Therefore, when the depth, namely the height, of the N-type column 201 increases, the first on-resistance will increase.

[0107] A first changed resistance of the extended portion of the N-type column 201 having a thickness whose first depth difference between the depth of the N-type column 201 and the first depth is calculated.

[0108] The first resistance correction coefficient is obtained by subtracting the first change resistance from the first on-resistance of the N-type column 201 and dividing the result by the first on-resistance.

[0109] by Fig. 10B For example, due to Fig. 10B The depth d of the N-type column 201 is E Less than Fig. 10A The depth d of the N-type column 201 is C , d C That is the first depth. Since it needs to be normalized to the first depth for comparison, only Fig. 10B The depth of the N-type column 201 is d C The resistance of the part within the range is taken as the modified on-resistance, namely the first modified on-resistance; at this time, when the N-type column 201 from the depth d is calculated C to d E When the first resistance of the extended part is changed, the first resistance correction coefficient can be finally calculated. After the first resistance correction coefficient is obtained, the first on-resistance only needs to be measured and then multiplied by the corresponding first resistance correction coefficient to obtain the first corrected on-resistance.

[0110] like Fig.12 As shown, it is a curve of the proportion of the 1 micron thick N-type epitaxial layer in the first on-resistance at different depths of the super junction unit in an embodiment of the present invention; Fig.12 The horizontal axis corresponds to the depth of the N-type epitaxial layer, namely, NEPI Depth, and the vertical axis corresponds to the proportion of NEPI per 1 micron in Ron. Curve 302 corresponds to the curve of conventional uniformly doped NEPI, and curve 303 corresponds to the linearly doped NEPI curve. By doping the corresponding NEPI, the corresponding first changed resistance can be obtained by extending the curve 302 or 303 accordingly.

[0111] In the embodiment of the present invention, the distribution curve of the groove is Fig.13 From the curve 304, it can be seen that the depth of the groove increases linearly from the edge of the middle area of ​​the semiconductor substrate to the edge of the wafer. By using the distribution characteristics of the groove, the calculation step of the first resistance correction coefficient can be further simplified, that is, the step of obtaining the first resistance correction coefficient of the N-type column 201 at each depth includes:

[0112] The first resistance correction coefficient of the N-type column 201 with the greatest depth is calculated. The N-type column 201 with the greatest depth is located closest to the edge of the wafer. The first maximum correction coefficient deviation value is obtained by subtracting the first resistance correction coefficient of the N-type column 201 with the greatest depth from 1.

[0113] The first depth difference of the N-type column 201 with the largest depth is set as the first maximum depth difference.

[0114] The first correction coefficient deviation value of each N-type column 201 is calculated by using the ratio of the first maximum correction coefficient deviation value to the first maximum depth difference value, which is equal to the relationship between the first correction coefficient deviation value and the first depth difference value of each N-type column 201 .

[0115] The first resistance correction coefficient of each N-type column 201 is obtained by subtracting the first correction coefficient deviation value of each N-type column 201 from 1. The distribution curve of the first resistance correction coefficient obtained is as follows: Fig.14 As shown in the curve 305, Fig.14 In the figure, since the horizontal axis is the wafer position, the groove depth corresponding to different positions of the wafer on the same diameter is Fig.13 The curve 304 is determined by Fig.14 The curve 305 in FIG. 3 actually corresponds to the first resistance correction coefficient of the trenches with different depths, and Fig.14 The curve 305 can directly link the first on-resistance obtained by testing at different positions with the corresponding first resistance correction coefficient during the measurement process. Fig.14 A numerical table 305a of the first resistance correction coefficient corresponding to the curve 305 is also given. In the numerical table 305a, the Ron depth coefficient is the first resistance correction coefficient, the values ​​corresponding to X represent the position on the wafer, and D_Factor represents the depth coefficient.

[0116] In the embodiment of the present invention, the second on-resistance of the P-type column 202 is obtained by forming contact holes at both ends of the P-type column 202 for testing, so the deeper the depth of the P-type column 202, the smaller the second on-resistance. The step of obtaining the second resistance correction coefficient includes:

[0117] A second changed resistance generated by an extended portion of the P-type column 202 having a thickness having a first depth difference between the depth of the P-type column 202 and the first depth is calculated.

[0118] The second resistance correction coefficient is obtained by adding the second change resistance to the second on-resistance of the P-type column 202 and dividing the result by the second on-resistance.

[0119] Preferably, the depth of the groove increases linearly from the edge of the middle area of ​​the semiconductor substrate to the edge of the wafer, and the step of obtaining the second resistance correction coefficient of the P-type column 202 at each depth includes:

[0120] The second resistance correction coefficient of the P-type column 202 with the greatest depth is calculated. The P-type column 202 with the greatest depth is located closest to the edge of the wafer. The second resistance correction coefficient of the P-type column 202 with the greatest depth is subtracted from 1 to obtain the second maximum correction coefficient deviation value. Fig.15 To explain: Fig.15 As shown, the first depth is 42 microns, and the depth of the P-type column 202 with the largest depth is 44 microns. Multiple tests are performed to obtain multiple second resistances of the P-type column 202 corresponding to the first depth of 42 microns and the average value is taken. Multiple tests are performed to obtain multiple second resistances of the P-type column 202 corresponding to the depth of 44 microns and the average value is taken. By comparing the second resistance of the P-type column 202 corresponding to 42 microns and the second resistance of the P-type column 202 corresponding to 44 microns, the proportion of resistances within the first depth range in the P-type column 202 with the largest depth can be obtained. The proportion is the second resistance correction coefficient of the P-type column 202 with the largest depth. If the calculation result is:

[0121] 23895 / 24330=98.22%. Therefore, if the depth increases by 2 microns, the impact is about 1.8%, that is, the second maximum correction coefficient deviation is about 1.8%. By similar deduction, if the depth increases by 2.5 microns, the second maximum correction coefficient deviation is estimated to be approximately 2.5%.

[0122] The first depth difference of the P-type column 202 with the largest depth is assumed to be the first maximum depth difference.

[0123] The second correction coefficient deviation value of each P-type column 202 is calculated by using the ratio of the second maximum correction coefficient deviation value to the first maximum depth difference value, which is equal to the relationship between the second correction coefficient deviation value and the first depth difference value of each P-type column 202 .

[0124] The second resistance correction coefficient of each P-type column 202 is obtained by adding 1 to the second correction coefficient deviation value of each P-type column 202. The distribution curve of the second resistance correction coefficient finally obtained is as follows: Fig.16 As shown in curve 307, Fig.16 A numerical table 307a of the second resistance correction coefficient corresponding to the curve 307 is also given. In the numerical table 307a, the Rs_Pillar depth coefficient is the second resistance correction coefficient, each value corresponding to X represents the position on the wafer, and D_Factor represents the depth coefficient.

[0125] like Fig.17A As shown, Fig.14 and Fig.16 The curves 305 and 307 are put together because they are on the same diameter of the same wafer, so the curves 305 and 307 can use the same horizontal coordinate.

[0126] Combination Fig.17A The third resistance correction coefficient can also be easily obtained, such as Fig. 17B As shown, it is a distribution curve 308 of the third resistance correction coefficient at different positions of the wafer in an embodiment of the present invention; Fig. 17B A numerical table 308a of the comprehensive depth factor of Ron VS Rs_Pillar is also provided, wherein N_D_Factor in the numerical table 308a represents the first resistance correction factor, P_D_Factor represents the second resistance correction factor, and NP_D_Factor represents the third resistance correction factor. At each position, the value of NP_D_Factor can be obtained by dividing the value of N_D_Factor by the value of P_D_Factor.

[0127] In the embodiment of the present invention, the first on-resistance of each of the N-type pillars 201 is obtained by testing the super junction device formed on the super junction unit corresponding to the N-type pillar 201 .

[0128] The super junction device includes a super junction MOSFET.

[0129] like Fig.18 , which is a top view of a second on-resistance test structure 401 in an embodiment of the present invention; the second on-resistance test structure 401 of the P-type pillar 202 of various depths is arranged on the semiconductor substrate, and each of the test structures 401 includes:

[0130] A plurality of P-type dummy pillars 402 are provided, each of which has an N-type dummy pillar between them. The depth of each of the P-type dummy pillars 402 is equal and equal to the depth of the P-type pillar 202 to be tested.

[0131] Contact holes 403 are formed at both ends of the middle P-type dummy column 402 , and the middle P-type dummy column 402 is tested through the two contact holes 403 to obtain the second on-resistance of the P-type column 202 with the same depth.

[0132] Preferably, each of the test structures 401 is disposed on a dicing street.

[0133] The length of the P-type dummy gate in each of the test structures 401 is greater than 5 times of the depth.

[0134] The number of the P-type dummy gates in each of the test structures 401 is five.

[0135] The embodiment of the present invention utilizes the characteristics that the first on-resistance of the N-type column 201 is related to the total charge of the N-type column 201 and the second on-resistance of the P-type column 202 is related to the total charge of the P-type column 202. At the same time, the embodiment utilizes the characteristic that the charge amount involved in matching in the N-type column 201 and the P-type column 202 of different depths is related to the depth, and the first resistance correction coefficient of the N-type column 201 of different depths and the second resistance correction coefficient of the P-type column 202 of different depths are obtained in advance. In this way, by utilizing the characteristic that the first on-resistance and the second on-resistance can be directly obtained by measurement, the super-junction units at various depths in the plane of the super-junction device can be measured to obtain the corresponding first on-resistance and second on-resistance. By using the first resistance correction coefficient and the second resistance correction coefficient to correct the first on-resistance and the second on-resistance, the charge amount involved in charge matching in the total charge of the N-type column 201 and the P-type column 202 can be obtained, so that the charge matching conditions of the super-junction units at different depths can be obtained, and finally the charge matching conditions in the plane of the super-junction structure can be quantitatively analyzed.

[0136] By using the quantitative analysis results of the in-plane charge matching of the super junction structure, it is possible to quickly and accurately determine the quality differences of super junction structures formed by different processes, and thus select a better process condition to form the super junction structure.

[0137] Compared with the existing method of using breakdown voltage to judge the in-plane charge matching of the super junction structure, the embodiment of the present invention does not need to rely on the complete accuracy of the pull-bias condition.

[0138] In order to further illustrate that the embodiment of the present invention can quantitatively obtain the in-plane charge matching of the super junction structure by measuring the first on-resistance and the second on-resistance and combining the pre-obtained distribution of the first resistance correction factor and the second resistance correction factor with depth, the following is further described:

[0139] like Figure 4A , which is a top view of super junction units in different regions on the same semiconductor substrate in an embodiment of the present invention; Figure 4A The super junction unit of three regions is shown, the N-type column 201 is also represented by N, the P-type column 202 is also represented by P, and the three positions are represented by Edge, Middle and Center respectively. Edge represents the edge area of ​​the wafer, Center represents the middle area of ​​the wafer, and Middle represents the area between Edge and Center.

[0140] The total charge of the N-type column 201, i.e., QSN, and the total charge of the P-type column 202, i.e., QSP, can be respectively expressed as Figure 4A The QSN and QSP at three different positions in the figure are represented. QSN is determined by the N-type epitaxial growth and trench formation process; QSP is determined by the trench formation process and the P-type epitaxial filling process. As the process adjustment changes, the distribution of QSN and QSP will change. As shown in Tables 1 and 2 below, they are the distributions of QSN and QSP under two process conditions.

[0141] Table 1

[0142] Location QSN QSP(EPI1) Edge 8 8 Middle 9 9 Center 10 8

[0143] Table 2

[0144] Location QSN QSP(EPI2) Edge 8 8 Middle 9 9 Center 10 9

[0145] EPI1 in Table 1 indicates that the filling process of the P-type epitaxial layer of the P-type column 201 is the first epitaxial process, and EPI2 in Table 2 indicates that the filling process of the P-type epitaxial layer of the P-type column 201 is the second epitaxial process. It can be seen that only the QSP at the Center position will change in Table 1 and Table 2, that is, a pull-off data is generated, but it is not necessarily the pull-off vertex, that is, Figure 2 The vertex of the middle curve 104.

[0146] Therefore, when the QSP data is a certain deviation, it is not necessarily the deviation apex. How to quantitatively evaluate the quality of the P / N in-plane match of the two processes? At this time, the QSN / QSP can be normalized to the center and the in-plane distribution parameters of the N charge and P charge can be calculated respectively. The following Tables 3 and 4 show the normalized numerical tables of the parameters corresponding to Tables 1 and 2 respectively.

[0147] Table 3

[0148]

[0149] Table 4

[0150]

[0151]

[0152] Tables 3 and 4 also give the ratio of QSN to QSP, i.e., QSN / QSP. It can also be taken as the inverse of QSN / QSP, i.e., QSP / QSN.

[0153] Normalization is achieved by dividing the values ​​at the edge and middle positions by the values ​​at the center position. Comparing Table 1 and Table 3, it can be seen that if QSN is directly divided by QSP or QSP is divided by QSN in Table 1, it cannot reflect the direct difference in PN matching at different positions; only after normalization in Table 3, can the difference in PN matching at different positions be intuitively compared from the values ​​of QSN / QSP or QSP / QSN. For QSN / QSP, less than 1 indicates overmatching, and greater than 1 indicates undermatching; the opposite is true for the value of QSP / QSN.

[0154] like Figure 4B As shown, Figure 4A The corresponding curve of the mismatch coefficient of the super junction unit changing with the position is obtained by changing the value of QSN / QSP with the position in Table 3 and Table 4. Figure 4B 1 represents the position Edge, 2 represents the position Middle, 3 represents the position Center, curve 203 corresponds to the QSN / QSP curve in Table 3, and curve 204 corresponds to the QSN / QSP curve in Table 4. It can be seen that the convergence of curve 204 is better, so the process conditions of EPI2 are better.

[0155] From the above, we can know that the PN matching of the super junction unit in the super junction structure can be finally obtained through QSN and QSP. Therefore, it is necessary to determine which factors affect QSN and QSP, so as to finally obtain a method to determine QSN and QSP.

[0156] like Figure 5 , which is a summary diagram of various factors affecting the PN mismatch coefficient of the super junction unit in an embodiment of the present invention; Figure 5Mark 205 corresponds to P / N Match Factors, i.e., factors affecting PN matching, and marks 206a to 206f specify the factors affecting PN matching into six situations, wherein mark 206a corresponds to NEPI concentration distribution, mark 206b corresponds to PEPI concentration distribution, mark 206c corresponds to angle, i.e., groove angle, mark 206d corresponds to depth, i.e., groove depth, mark 206e corresponds to critical dimension (CD), i.e., critical dimension of groove, and mark 206f corresponds to Profile. Among them, NEPI represents an N-type epitaxial layer, PEPI represents a P-type epitaxial layer filling the groove, and Profile represents the morphology of a P-type column.

[0157] Factors affecting the BV of super junction devices: Figure 5 On the basis of , it is necessary to add design factors, and it is only judged near the peak of the deviation. In addition, due to the objective fluctuation of the process and the existence of factors such as uncertainty, although it can be used for the final judgment of the process, there are too many influencing factors and it is not direct enough. Therefore, it is difficult to use BV to measure PN matching in the existing method.

[0158] In the embodiment of the present invention, it is proposed to obtain the in-plane distribution of QSN and QSP from Ron, i.e., the first on-resistance, and RS-Pillar, i.e., the second on-resistance. Figure 6 As shown, it is a schematic diagram of linking QSN and QSP of the super junction unit with Ron and RS-Pillar respectively in an embodiment of the present invention; mark 207 corresponds to QSN, mark 208 corresponds to QSP, mark 209 corresponds to Ron, mark 210 corresponds to RS-Pillar, and VS indicates that QSN and QSP are respectively related to Ron and RS-Pillar, that is, RonVS QSN, RS-PillarVS QSP. Therefore, the first thing to do is how to get QSN from Ron, and how to get QSP from RS-Pillar. In this calculation process, depth is a more troublesome factor, because its influence on resistance value and matching is inconsistent. Its influence on resistance is either series, that is, series in Ron, or parallel, that is, parallel in RS_Pillar, so the influence of depth on resistance is more obvious, while the influence on matching charge can be basically ignored. It is an interference item that must be corrected and excluded as much as possible.

[0159] Other parameters (CD, angle, concentration, profile, etc.) have similar effects on resistance and matching charge, and have similar effects on Ron and QSM_N. , The distribution parameters of Rs_pillar and QSM_P have the same degree of influence.

[0160] In the path perpendicular to the resistance test current, the influencing factors have the same impact on QS and resistance.

[0161] In the extension line of the current path, the influencing factors have different effects on resistance and QS.

[0162] QSM evaluates the matching part of QS.

[0163] The analysis is as follows:

[0164] like Fig. 7A and Figure 7B , which is a schematic diagram of the structure of the N-type column when only the concentration of the N-type column changes in determining the relationship between QSN and Ron; Fig. 7A The three-dimensional structure diagram of the N-type column 201 at the center position shows two test electrodes "+" and "-" and multiple dimensions, where W represents the length of the N-type column 201, a represents the width of the N-type column 201, d represents the height of the N-type column 201, S represents the cross-sectional area of ​​the N-type column 201, and CON represents the cross-sectional area of ​​the N-type column 201. C Indicate the doping concentration.

[0165] Figure 7B The three-dimensional structure diagram of the N-type column 201 at the edge position, CON E Indicates doping concentration, other markings show content and Figure 7B same.

[0166] Fig. 7A and Figure 7B The calculation formulas for the cross-section S, charge Qs (i.e., QSN), on-resistance Ron, on-conductance Gon, and on-resistance ratio Gon Ratio and charge ratio Qs Ratio are uniformly shown in Table 5 below.

[0167] Table 5

[0168]

[0169]

[0170] It can be seen from Table 5 that Gon Ratio can directly reflect Qs Ratio.

[0171] like Fig. 8A and Figure 8B As shown, the relationship between QSN and Ron is determined in Fig. 7A and Figure 7B A schematic diagram of the structure of the N-type column when the key dimension changes of the N-type column are added on the basis of Fig. 7A On the basis of Fig. 8A The key dimension of the N-type column, i.e. the width, is marked separately with a C indicates; in Figure 7B On the basis of Figure 8BThe key dimension of the N-type column, i.e. the width, is marked separately with a E express. Fig. 8A and Figure 8B The calculation formulas for the cross-section S, charge Qs (i.e., QSN), on-resistance Ron, on-conductance Gon, and on-resistance ratio Gon Ratio and charge ratio Qs Ratio are uniformly shown in Table 6 below.

[0172] Table 6

[0173]

[0174]

[0175] It can be seen from Table 6 that Gon Ratio can directly reflect Qs Ratio, so CD has the same effect on resistance RON and charge QS.

[0176] like Fig.9A and Fig. 9B As shown, the relationship between QSN and Ron is determined in Fig. 7A and Figure 7B The schematic diagram of the structure of the N-type column when the side angle of the N-type column changes on the basis of Fig. 7A On the basis of Fig.9A The top width of the N-type column is marked with a CT Indicates that the bottom width of the N-type column is marked with a CB indicates; in Figure 7B On the basis of Fig. 9B The top width of the N-type column is marked with a ET Indicates that the bottom width of the N-type column is marked with a EB express. Fig. 8A and Figure 8B The calculation formulas for the cross-section S, the charge Qs (i.e., QSN), the on-resistance Ron, the on-conductance Gon, and the on-resistance ratio Gon Ratio and the charge ratio Qs Ratio are uniformly shown in Table 7 below.

[0177] Table 7

[0178]

[0179] In Table VII, the angle can be decomposed into Top CD, i.e. top width, and Bottom CD, i.e. bottom width.

[0180] The in-plane angle difference is small, and the Top CD and Bottom CD Bias can be considered consistent. It is only used to correct the depth coefficient.

[0181] All normalizations are done toward the center. A is a coefficient and a constant. Resistance is converted to conductance for easier explanation.

[0182] It can be seen that the Angle, i.e. the angle, has the same effect on the resistance RON and the charge QS.

[0183] like Fig. 10A Shown and Fig. 10B The following figure shows the relationship between QSN and Ron. Fig. 8A and Figure 8B A schematic diagram of the structure of the N-type column when the depth of the N-type column is changed based on the above; Fig. 8A On the basis of Fig. 10A The depth of the N-type column is marked separately with d C indicates; in Figure 8B On the basis of Fig. 10B The depth of the N-type column is marked separately with d E express. Fig. 10A and Fig. 10B The calculation formulas for the cross-section S, charge Qs (i.e., QSN), on-resistance Ron, on-conductance Gon, and the on-resistance ratio Gon Ratio and charge ratio Qs Ratio of the two are uniformly shown in Table 8 below.

[0184] Table 8

[0185]

[0186] The concentration variation spec within the wafer surface is +-3%. Within this small variation range, it can be approximately considered that the resistance R is inversely proportional to the concentration CON. All normalizations are performed toward the center. A depth correction factor needs to be introduced.

[0187] Table 8 considers all cross-sectional charges Qs, but the purpose of this analysis is to determine the charge matching difference, so the Qs taken into account should only include the part that affects the matching. This situation will be analyzed in Table 9 below.

[0188] Table 9

[0189]

[0190]

[0191] Added cross section S in Table 9 M Depth normalization correction calculation, at this time, the cross section S at the Edge position M for a E *d C , that is, there is no need to consider the entire depth d E The amount of charge within the range only needs to consider the depth dC The amount of charge within the range, because only in this way can normalization to the Center be achieved.

[0192] In Table 9, all normalizations are performed toward the center. With the introduction of the depth correction factor, there are two cross-sectional areas in the calculation, and it is more convenient to directly decompose them in the formula.

[0193] Table 8 considers all the charge quantities Qs, but the purpose of this analysis is to determine the charge matching difference, so the Qs taken into consideration should only include the part that affects the matching. The corrected relationship is shown in Table 9.

[0194] From Table 9, we can see that *Gon Ratio multiplied by the coefficient (d E / d C ), which can directly reflect the depth normalization of Qs Ratio, as a measure of charge matching difference.

[0195] The present invention has been described in detail above through specific embodiments, but these do not constitute limitations of the present invention. Without departing from the principle of the present invention, those skilled in the art may also make many variations and improvements, which should also be considered as the protection scope of the present invention.

Claims

1. A method for determining the distribution of in-plane charge balance state of a super junction device, characterized in that: The steps include: Step 1, obtaining a first resistance correction coefficient of an N-type column of different depths and a second resistance correction coefficient of a P-type column of different depths in a super junction structure; The P-type pillars of the super junction structure are composed of a P-type epitaxial layer filled in a trench, the trench is formed in an N-type epitaxial layer, and the N-type pillars are composed of the N-type epitaxial layer located between the P-type pillars; The N-type epitaxial layer is formed on a semiconductor substrate, and a plurality of trenches are provided on the semiconductor substrate. The trenches are formed simultaneously by the same process, and the depths of the trenches on the semiconductor substrate are uneven due to process deviation. A super junction unit is formed by one of the N-type columns and one of the adjacent P-type columns, and the depth of the N-type column and the depth of the P-type column of the super junction unit are both the depth of the trench at the super junction unit; The first resistance correction coefficient corrects the first on-resistance of the N-type column to a first corrected on-resistance, where the first corrected on-resistance is the on-resistance when the depth of the N-type column is adjusted to a first reference depth while maintaining the charge distribution of the N-type column; The second resistance correction coefficient corrects the second on-resistance of the P-type column to a second corrected on-resistance, where the second corrected on-resistance is the on-resistance when the depth of the P-type column is adjusted to the first reference depth while maintaining the charge distribution of the P-type column; Step 2, measuring the first on-resistance of the N-type column and the second on-resistance of the P-type column of the super junction unit at different positions on the wafer; Step 3: Calculate the charge balance state of the super junction unit at each position and thereby form an in-plane charge balance state distribution of the super junction device. The step of calculating the charge balance state of the super junction unit at each position includes: The first on-resistance is multiplied by the first resistance correction coefficient corresponding to the depth of the super junction unit to obtain the first corrected on-resistance, and the second on-resistance is multiplied by the second resistance correction coefficient corresponding to the depth of the super junction unit to obtain the second corrected on-resistance; The inverse of the first modified on-resistance is taken to form a first modified conductance and the first modified conductance is normalized, and the inverse of the second modified on-resistance is taken to form a second modified conductance and the second modified conductance is normalized; The charge balance state of the super junction unit is obtained by dividing the normalized second corrected conductance by the normalized first corrected conductance.

2. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 1, characterized in that: Step one also includes dividing the first resistance correction coefficient corresponding to each depth by the second resistance correction coefficient to obtain a third resistance correction coefficient; The calculation step of the charge balance state of the super junction unit at each position in step 3 is modified as follows: The first on-resistance is taken as the reciprocal to form a first conductance and the first conductance is normalized, and the second on-resistance is taken as the reciprocal to form a second conductance and the first conductance is normalized; The normalized second conductance is divided by the normalized first conductance and then multiplied by the third resistance correction coefficient to obtain a charge balance state of the super junction unit.

3. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 1 or 2, characterized in that: The semiconductor substrate includes a silicon substrate.

4. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 3, characterized in that: The semiconductor substrate is in a wafer structure.

5. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 4, characterized in that: The distribution structure of the depth of the trench on the semiconductor substrate includes: A first uniform depth is provided in a middle region of the semiconductor substrate; The depth of the trench gradually increases in a direction from the edge of the middle region of the semiconductor substrate to the edge of the wafer.

6. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 5, characterized in that: The first reference depth is set as the first depth.

7. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 6, characterized in that: The normalization in step three is normalization to the middle area.

8. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 6, characterized in that: The step of obtaining the first resistance correction coefficient comprises: The deeper the depth of the N-type column, the greater the first on-resistance is, and a first changed resistance of an extension portion of the N-type column having a thickness whose first depth difference between the depth of the N-type column and the first depth is calculated; The first resistance correction coefficient is obtained by subtracting the first change resistance from the first on-resistance of the N-type column and dividing the result by the first on-resistance.

9. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 8, characterized in that: The depth of the groove increases linearly and gradually in a direction from the edge of the middle area of ​​the semiconductor substrate to the edge of the wafer, and the step of obtaining the first resistance correction coefficient of the N-type column at each depth includes: Calculating the first resistance correction coefficient of the N-type column with the greatest depth, where the N-type column with the greatest depth is located closest to the edge of the wafer, and subtracting the first resistance correction coefficient of the N-type column with the greatest depth from 1 to obtain a first maximum correction coefficient deviation value; Let the first depth difference of the N-type column with the largest depth be the first maximum depth difference; The first correction coefficient deviation value of each N-type column is calculated by using the ratio of the first maximum correction coefficient deviation value to the first maximum depth difference value to be equal to the relationship between the first correction coefficient deviation value and the first depth difference of each N-type column; The first resistance correction coefficient of each N-type column is obtained by subtracting the first correction coefficient deviation value of each N-type column from 1.

10. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 6, characterized in that: The step of obtaining the second resistance correction coefficient comprises: The deeper the depth of the P-type column, the smaller the second on-resistance is, and a second change resistance generated by an extended portion of the P-type column having a thickness equal to a first depth difference between the depth of the P-type column and the first depth is calculated; The second resistance correction coefficient is obtained by adding the second change resistance to the second on-resistance of the P-type column and dividing the result by the second on-resistance.

11. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 10, characterized in that: The depth of the groove increases linearly from the edge of the middle area of ​​the semiconductor substrate to the edge of the wafer, and the step of obtaining the second resistance correction coefficient of the P-type column at each depth includes: Calculating the second resistance correction coefficient of the P-type column with the greatest depth, where the P-type column with the greatest depth is located closest to the edge of the wafer, and subtracting 1 from the second resistance correction coefficient of the P-type column with the greatest depth to obtain a second maximum correction coefficient deviation value; Let the first depth difference of the P-type column with the largest depth be the first maximum depth difference; The second correction coefficient deviation value of each P-type column is calculated by using the ratio of the second maximum correction coefficient deviation value to the first maximum depth difference value to be equal to the relationship between the second correction coefficient deviation value and the first depth difference of each P-type column; The second resistance correction coefficient of each P-type column is obtained by adding 1 to the second correction coefficient deviation value of each P-type column.

12. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 1 or 2, characterized in that: The first on-resistance of each of the N-type columns is obtained by testing the super junction device formed on the super junction unit corresponding to the N-type column.

13. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 12, characterized in that: The super junction device includes a super junction MOSFET.

14. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 1 or 2, characterized in that: A test structure for the second on-resistance of the P-type column at each depth is provided on the semiconductor substrate, and each of the test structures comprises: A plurality of P-type pseudo columns, each of which has an N-type pseudo column between them, and each of which has an equal depth and is equal to the depth of the P-type column to be tested; Contact holes are formed at both ends of the middle P-type pseudo column, and the middle P-type pseudo column is tested through the two contact holes to obtain the second on-resistance of the P-type columns with the same depth.

15. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 14, characterized in that: Each of the test structures is arranged on a dicing street.

16. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 14, characterized in that: The length of the P-type dummy pillar in each of the test structures is greater than 5 times of its depth.

17. The method for determining the distribution of the in-plane charge balance state of a super junction device according to claim 14, characterized in that: The number of the P-type dummy columns in each of the test structures is 5.

Citation Information

Patent Citations

  • Trench type super junction and manufacturing method thereof

    CN108022924A

  • Method for analyzing PN matching state of super junction device

    CN109100623A

  • Manufacturing method of super junction

    CN111129117A

  • Super junction device and manufacturing method thereof

    CN112786677A

  • Semiconductor device with superjunction structure, and its manufacturing method

    JP2007251023A