TWI631438B - Method of controlling flatness of epitaxial wafer - Google Patents

Abstract

The present invention discloses a method for controlling the flatness of an epitaxial wafer, which comprises obtaining a flatness defect rate graph of an epitaxial wafer produced by a given epitaxial reactor, and obtaining a relationship graph, the relationship graph representing epitaxy The relationship between the growth conditions of the layer and the Z-axis Double Derivatives (ZDDs) of the epitaxial wafers used in the relationship, by the first epitaxial reactor in a given epitaxial reactor And obtaining a difference ZDD of the at least one sample epitaxial wafer generated under the first growth condition, and adjusting the first growth condition based on the relationship graph.

Description

Method for controlling flatness of epitaxial wafer

The present invention relates to a method of controlling the flatness of an epitaxial wafer.

On a germanium wafer having a low specific resistance doped with impurities such as boron, the germanium epitaxial layer is doped with a relatively small amount of impurities having a relatively high specific resistance, and the germanium epitaxial layer is grown to thereby form germanium epitaxial. Wafers, typically 矽 epitaxial wafers have high collection capacity, low latch-up characteristics, and high resistance to sliding defects at high temperatures.

The quality items required for such epitaxial wafers include flatness, particle contamination levels, etc., and the terms used for the epitaxial layer include thickness uniformity, specific resistance, metal contamination, stacking defects, sliding shift, and the like.

The thickness of the wafer can be measured and the flatness of the edge of the wafer can be measured using the measured wafer thickness.

Accordingly, the present invention provides a method of controlling the flatness of an epitaxial wafer that substantially obviates one or more of the problems caused by the limitations and disadvantages of the prior art.

It is an object of the present invention to provide a method for controlling the flatness of an epitaxial wafer in which process control is performed such that the defect rate is lowered regardless of the shape of the substrate or the flatness level of the substrate.

Other advantages, objects, and features of the invention will be set forth in part in the description which follows, It is understood or can be derived from the practice of the invention. The objectives and other advantages of the invention will be realized and attained by the <RTIgt;

In order to achieve these and other features of the present invention, the present invention is embodied and broadly described. A method of controlling the flatness of an epitaxial wafer includes obtaining an epitaxial wafer produced by a given epitaxial reactor. The flatness defect rate pattern is obtained by obtaining a relationship graph indicating the relationship between the growth conditions of the epitaxial layer and the Z-axis Double Derivatives (ZDDs) of the epitaxial wafers used for the relationship. The difference ZDD of the at least one sampled epitaxial wafer produced under the first growth condition is obtained by a first epitaxial reactor in the given epitaxial reactor, and the first growth condition is adjusted based on the relationship pattern.

The method further includes calculating a correction value based on the difference ZDD and the flatness defect rate pattern of the at least one sampled epitaxial wafer, and adjusting the first growth condition based on the correction value and the relationship pattern.

Obtaining the flatness defect rate graph includes obtaining the edge sector site front side reference Q (Site least squares plane Derivations; ESFQDs) and the Z-axis double Derivatives ;ZDDs), obtaining the difference ZDDs of the epitaxial wafers formed based on the substrates and the edge sector Site Front side reference Q (Site least squares planes Ranges; ESFQRs), based on the Lei The ESFQRs of the wafer determine the flatness defect of the epitaxial wafer, and the difference ZDDs of the ESFQDs and the epitaxial wafer based on the substrate, and obtain the flatness defect rate graph of the epitaxial wafer.

Obtaining the difference between the ZDDs and the ESFQRs of the epitaxial wafers includes obtaining the ZDDs and ESFQRs of the epitaxial wafers, and using the difference between the ZDDs of the substrate and the ZDDs of the epitaxial wafers to obtain the difference ZDDs of the epitaxial wafers.

Calculating the correction value includes setting a target segment based on a defect rate of the flatness defect rate graph, and calculating a correction value based on the difference ZDD of the at least one sampled epitaxial wafer and the target segment.

When determining the flatness defect of the epitaxial wafer, the flatness defect of the epitaxial wafer is determined based on the comparison result of the ESFQRs of the epitaxial wafer with a predetermined reference value. The predetermined reference value is from 100 nm to 120 nm.

In the flatness defect rate graph of the epitaxial wafer, the X axis represents the ESFQDs of the substrate, the Y axis represents the difference ZDDs of the epitaxial wafer, and the difference between the ESFQDs of the substrate and the epitaxial wafer ZDDs, epitaxial crystal The defect rate of the circle is divided into a plurality of regions.

When the relationship pattern is obtained, the difference ZDDs of the epitaxial wafers for relationship generated under different growth conditions are obtained, and the difference ZDDs of the epitaxial wafers obtained by using these growth conditions and the obtained relationship are obtained. Relationship graphics.

The growth conditions and first growth conditions for obtaining the relationship pattern include the flow rate of hydrogen, the flow rate of the TCS, or the growth temperature.

When the target segment is set, the target segment is a segment in which the flatness defect rate in the entire range of the ESFQDs of the substrate is less than a predetermined reference value. The predetermined reference value is 0.1% to 15%.

The correction value is the difference between the predetermined target value belonging to the target segment and the difference ZDD of the at least one sample epitaxial wafer.

The predetermined target value is the lower limit, upper limit, or median of the target zone.

When the first growth condition is adjusted, the first growth condition is adjusted to a second growth condition corresponding to a value obtained by increasing the correction value to the difference ZDD under the first growth condition of the first epitaxial reactor.

In another aspect of the present invention, a method for controlling the flatness of an epitaxial wafer includes: obtaining an edge sector Site Front side reference Q (Site least squares plane) of an edge of a first substrate. Derivations; ESFQDs) and the first Z-axis Double Derivatives (ZDDs), obtaining the edge-shaped location of the first epitaxial wafer based on the first substrate, the front-side reference location, the least square plane range (Edge sector Site Front) Side reference Q (Site least squares plane) Ranges; ESFQRs) and a second Z-axis double derivative (ZDDs), obtain a first difference ZDDs of the first epitaxial wafer, and determine the first based on the ESFQRs of the first epitaxial wafer The flatness defect of the epitaxial wafer and the flatness defect rate are obtained, and the relationship graph is obtained, and the relationship graph represents the relationship between the epitaxial layer growth condition of the second epitaxial wafer and the second ZDDs of the second epitaxial wafer. Obtaining a second difference ZDD of the at least one third epitaxial wafer generated under the first growth condition, and calculating a correction based on the second difference ZDD and the flatness defect rate graph ; Correction value based on the relationship between the graphics, the second adjusting the first growth conditions the growth conditions.

This method further includes producing an epitaxial wafer under the second growth conditions.

Obtaining a relationship graph includes generating a second epitaxial wafer by growing an epitaxial layer on the second substrate according to different values of the same growth condition item, obtaining a difference ZDDs of the generated second epitaxial wafer, and using the same growth condition The difference between the value of the item and the difference ZDDs of the second epitaxial wafer is obtained as a relational graph.

Obtaining a second difference amount ZDD included in the first growth condition, by growing an epitaxial layer on the plurality of third substrates, generating a plurality of third epitaxial wafers, and obtaining a difference ZDDs of the third epitaxial wafer, And setting an average value of the difference ZDDs of the obtained third epitaxial wafer to a second difference amount ZDD.

Calculating the correction value includes setting a target segment of the flatness defect pattern, wherein the flatness defect rate is less than a predetermined reference value; and calculating the correction value such that the second difference ZDD belongs to the target segment.

When the first growth condition is adjusted to the second growth condition, the first growth condition is changed by the growth conditions corresponding to the correction value.

It is to be understood that the foregoing general description of the invention and the claims

Preferred embodiments of the present invention will now be described in detail in conjunction with the drawings. In the following description of the embodiments, it will be understood that when each element (film), region, pad or structure is referred to as being "above" or "under" the substrate, another element (film), region or pad or pattern. It may be located "on" or "below" another element or indirectly formed with one or more intervening elements between the two. In addition, it is to be understood that an element will be described as "above" or "below". In addition, in the drawings, the same or similar elements are denoted by the same reference numerals even in different drawings.

1 is a flow chart of a method for controlling the flatness of an epitaxial wafer according to an embodiment of the present invention.

Referring to FIG. 1 , a method for controlling flatness of an epitaxial wafer according to an embodiment of the present invention includes obtaining a flatness defect rate pattern (step S110), and obtaining a relationship graph indicating a relationship between a growth condition and a difference ZDDs (step S120). And obtaining the difference ZDD2 (step S130), calculating the correction value based on the difference ZDD2 and the flatness defect rate pattern (step S140), adjusting the growth condition based on the correction value and the relationship pattern (step S150), and the adjusted growth condition The epitaxial wafer is fabricated (step S160).

The difference between the edge sector site front side reference Q (Site least squares plane Derivation; ESFQD) and the epitaxial wafer used for testing based on the edge of the substrate used for the test ZDDs (Z-axis Double Derivative; ZDD), a flatness defect rate pattern is obtained (step S110).

Fig. 2 is a flow chart showing in more detail the acquisition of the flatness defect rate pattern of the embodiment shown in Fig. 1 (step S110).

Referring to FIG. 2, obtaining a flatness defect rate pattern (step S110) includes acquiring ESFQDs and ZDDs (hereinafter referred to as ZDD1s) of the substrate for testing (step S210), acquiring a difference ZDD1s (step S220), and determining for The flatness defect of the tested epitaxial wafer (step S230), and the flatness defect rate pattern are acquired (step S240).

First, ESFQDs and ZDD1s are acquired (step S210).

The ESFQD is an abbreviation of the Edge Sector Site Front Side Reference Q (Site least squares plane Derivation) and has a positive (+) value or a negative (-) value.

7A is a schematic diagram of ESFQD and ESFQR, and FIG. 7B is a schematic cross-sectional view along line a-b of FIG. 7A.

Referring to FIGS. 7A and 7B, a location (eg, S1) obtained by radially dividing the surface (eg, the front side) of the wafer W1 by a specified angle θ (eg, θ=5 ̊) is measured from the reference plane. The thickness of the section 401. In addition, the maximum value and the minimum value of the thickness of the wafer W1 in the section 401 of one surface of the wafer W1 are measured. For example, the reference plane PLANE is an ideal plane of one surface of the wafer W1, such as a plane of z=0.

The segment 401 is a segment from a first point P1 to a second point P2 of a location (eg, S1), the first point P1 being a specified distance S2 from the edge E1 of the wafer W1 (eg, 1 mm to 2 mm) The point of the second point P2 is a point at a predetermined distance (for example, 10 mm) from the first point P1 toward the center c direction of the wafer W1.

If one surface of the wafer W1 is above the reference plane (eg, Z=0), the ESFQD has a positive (+) value (+EXFQD value), and if the surface of the wafer W1 is in a reference plane (eg, Z=0) Below, the ESFQD has a negative (-) value (-EXFQD value).

The ESFQD may have an average value of the wafer thickness of the segment shape 401, for example, a value obtained by integrating the thickness of the wafer W1 in the segment 401.

The ESFQR is an abbreviation of the Edge Sector Site Front Side Reference Q (Site least squares plane) range, and a method corresponding to the flatness of the wafer, which will be described below.

The ESFQR is defined as the difference (MAX-MIN) between the maximum (MAX) and minimum (MIN) of the measured thickness of the wafer W1 in the segment 401.

The ZDD is an abbreviation of the Z-axis Double Derivative, which is a parameter indicating the degree of roll-off of the surface of the edge of the wafer, thereby indicating the curvature of the surface of the edge of the wafer.

It is assumed that the surface of the wafer in the radial direction is the xy plane of the xyz coordinate system, and the z-axis direction is the thickness direction of the wafer. If the midway point between the two reference faces of the front side and the back side of the wafer has a Z-axis coordinate value of 0, the profiles of the thicknesses of the front side and the back side of the wafer are expressed as a Z function.

When calculating the average Z value of the radius R of the front side and the back side of the wafer in the range of 0 ̊ to 360 ,, the overall average radial profile of the wafer in the Z-axis direction is obtained, and the average radial profile obtained by the differential is obtained. Get ZDD twice.

That is, ZDD is defined by Equation 1 below.

[Equation 1]

The ESFQDs and ZDD1s of the substrate used for the test were obtained or calculated by the above method. In addition, the ESFQDs and ZDD1s of the substrate used for the test have known values when the substrate for testing is purchased.

Thereafter, the difference ZDDs (hereinafter referred to as "difference ZDD1s") and ESFQRs (hereinafter referred to as "ESFQR1s") of the epitaxial wafer for testing formed based on the substrate for testing are obtained (step S220).

3 is a flow chart for obtaining the difference ZDD1s and ESFQR1s in one embodiment of the present invention shown in FIG. 2 (step S220).

Referring to FIG. 3, an epitaxial wafer for testing is generated by growing an epitaxial layer on a substrate for testing (step S310).

Using M predetermined epitaxial reactors (N is a natural number > 1), M epitaxial wafers were produced by testing the epitaxial layers on the M substrates used for testing (M > 1) Natural number).

Each of the N epitaxial reactors grew an epitaxial layer (M is a natural number > 1) on a substrate randomly selected from the M substrates used for the test.

The substrate used for the test is divided into N groups corresponding to the N epitaxial reactors, and each of the N epitaxial reactors grows an epitaxial layer on a substrate belonging to one of the N groups.

For example, assume that the number of established epitaxial reactors is 10 and the number of substrates used for testing is 30,000, and each epitaxial reactor grows on 3,000 substrates randomly selected from 30,000 substrates used for testing. The layer is crystallized to produce 30,000 epitaxial wafers for testing.

Thereafter, ZDD (hereinafter referred to as "ZDD2") and ESFQR1 of each of the generated epitaxial wafers for testing are obtained (step S320).

In order to obtain ESFQR1, the methods shown in FIGS. 7A and 7B can be applied, and in order to obtain ZDD2, the above method can be applied to obtain ZDD1.

For example, 30,000 ZDD2s and 30,000 ESFQR1s were obtained for the 30,000 epitaxial wafers described above.

Thereafter, the difference ZDD (hereinafter referred to as "difference ZDD1 (Delta ZDD1)") of each of the epitaxial wafers for testing is obtained (step S330). Here, the difference ZDD is defined as the difference between the ZDD of the epitaxial wafer and the ZDD of the substrate of the epitaxial wafer.

Equation 2 below defines the difference ZDD1.

[Equation 2]

After the difference ZDD1s and ESFQR1s are obtained (step S220), the flatness defect of the epitaxial wafer for testing is determined based on the ESFQR1s of the epitaxial wafer for testing (step S230).

For example, the flatness defect of the epitaxial wafer used for the test is determined according to the comparison between the ESFQR1s of the epitaxial wafer for testing and a predetermined reference value. Here, the predetermined reference value is from 100 nm to 120 nm. For example, the predetermined reference value is 110 nm.

For example, if the ESFQR1 of the epitaxial wafer used for testing exceeds a predetermined reference value, the flatness of the epitaxial wafer used for the test is determined to be defective, and if the ESFQR1 of the epitaxial wafer used for the test is The predetermined reference value or less than the predetermined reference value, the flatness of the epitaxial wafer used for the test is determined to be defect-free.

Thereafter, a flatness defect rate pattern of the epitaxial wafer for testing is obtained based on the EDFQDs and the difference ZDD1s (step S240).

Fig. 4 is a diagram showing a flatness defect rate pattern obtained in step S240.

Referring to FIG. 4, the X-axis represents the ESFQDs of the substrate used for the test, which is divided by 0 at a predetermined interval (for example, 25 nm), and the unit of the X-axis is nanometer.

The Y axis represents the difference ZDDs of the epitaxial wafers used for the test, and is divided by 0 at a predetermined interval (for example, 5 or 10).

Based on the ESFQDs and the difference ZDD1s, the defect rate of the epitaxial wafer used for the test is divided into a plurality of regions, for example, regions 1 to 20.

The defect rate for each of the regions 1 to 20 is the ratio of the number of epitaxial wafers for testing determined to be defective to the number of epitaxial wafers for testing included in the corresponding regions.

For example, a defect rate of 0 in Zone 1 means that all of the epitaxial wafers included in Zone 1 for testing are defect free.

Thereafter, after the flatness defect rate pattern is obtained (step S110), the relationship between the growth conditions of the epitaxial layer and the difference ZDDs of the epitaxial wafers for the relationship or the relationship pattern thereof is obtained (step S120). According to another embodiment, step S110 and step S120 are exchanged. According to still another embodiment, step S110 and step S120 can be performed simultaneously.

For example, epitaxial wafers for relationships are created by growing epitaxial layers on substrates used for relationships in accordance with different values of the same growth condition terms of growth conditions. For example, the growth condition term includes the flow rate of trichlorosilane (TCS), the flow rate of hydrogen, the growth temperature, and the like.

Thereafter, the difference ZDDs of the epitaxial wafers for the relationship generated by the different values of the same growth condition term are obtained. Then, the relationship graph is obtained using different values of the growth condition terms and the difference ZDDs of the epitaxial wafers for the relationship. In addition, a relational graph (for example, a linear relationship equation) about a relationship is obtained using a relational graph. For example, the substrate used for the relationship is a substrate that is separate or different from the substrate used for testing.

5A is a graph showing the relationship between the difference ZDDs of the epitaxial wafers for the relationship and the hydrogen flow rate supplied to grow the epitaxial layer, and FIG. 5B is the difference indicating the epitaxial wafers for the relationship. A graph of the relationship between the amount of ZDDs and the growth temperature of the epitaxial layer, and FIG. 5C is a graph of the relationship between the difference ZDDs of the epitaxial wafers for the relationship and the flow rate of the TCS supplied to grow the epitaxial layer.

In FIGS. 5A to 5C, ◆ indicates an experimental result of the difference ZDDs of the epitaxial wafers for the relationship obtained when the growth conditions (the flow rate of the TCS, the flow rate of the hydrogen gas, and the growth temperature) are changed.

In FIGS. 5A to 5C, the difference ZDDs of the epitaxial wafers for relationship generated under a plurality of different growth conditions (for example, 5 growth conditions) are shown.

For example, under each different growth condition, the difference ZDDs of the epitaxial wafers used for the relationship is the average of a predetermined number (eg, 10) of epitaxial wafers for the relationship.

Linear equations y1, y2, and y3 that approximate the relationship between growth conditions and the difference ZDDs are obtained using a relational graph obtained based on the difference ZDDs of the epitaxial wafers for relationship generated under different growth conditions. In FIGS. 5A to 5C, R ² represents the degree of approximation between the experimental results and the linear equations y1, y2, and y3.

After obtaining the flatness defect rate pattern (step S110) and obtaining the relationship pattern (step S120), the current difference amount ZDD (hereinafter referred to as "difference amount ZDD2") expected to be used by the epitaxial reactor is obtained (step S130). In order to obtain the difference ZDD2, the above method of obtaining the difference ZDD1 can be applied.

By way of example, at least one sampled epitaxial crystal is produced by growing an epitaxial layer on at least one of the sampled substrates in the first set of epitaxial reactors using the currently desired epitaxial reactor under the first growth conditions. A circle, and a difference ZDD of at least one sampled epitaxial wafer is obtained as the difference ZDD2.

For example, under the first growth conditions, a plurality of sampled epitaxes are generated by growing an epitaxial layer on a plurality of sampling substrates (for example, five sampling substrates) using a first epitaxial reactor that is currently desired to be used. The difference ZDDs of the generated sample epitaxial wafers and the obtained average value of the difference ZDDs of the generated sample epitaxial wafers are used as the difference ZDD2.

Here, the term of the first growth condition coincides with the term of the growth condition at the time of acquiring the relationship graph. For example, the first growth condition and the growth conditions at which the relationship pattern is acquired include the flow rate of the TCS, the flow rate of hydrogen, or the growth temperature, as described above in connection with Figures 5A-5C. Although this embodiment representatively indicates that the first growth condition is including three growth condition items, the present disclosure is not limited thereto. For example, the first growth conditions of the first epitaxial reactor include other growth condition terms that have a linear relationship with the delta ZDD of the epitaxial wafer.

Thereafter, the correction value of the difference amount ZDD2 is calculated based on the difference amount ZDD2 shown in FIG. 4 and the flatness defect rate pattern (step S140).

Figure 6 is a flow chart for calculating a correction value for an embodiment of the present invention shown in Figure 1.

Referring to FIG. 6, the calculation of the correction value (step S140) includes setting the target section (step S410) and calculating the correction value (step S420).

The target section of the difference amount ZDDs is set based on the flatness defect rate graph shown in FIG. 4 (step S410).

For example, the target segment of the differential ZDDs is a region in which the flatness defect rate in the full range (for example, -75 to 25) of the ESFQDs of the substrate for testing in the flatness defect rate pattern is less than a predetermined reference value. segment. Here, the predetermined reference value is 0.1% to 15%.

For example, if in FIG. 4, the full range of the predetermined reference value with respect to the ESFQDs of the substrate is set to 0.1% to 15%, the target section of the difference ZDDs includes the areas 3, 8, 13, and 18, and the target section. The difference in ZDDs is from 0 nm to -5 nm. In the section where the difference ZDDs is from 0 nm to -5 nm, the flatness defect rate in the ESFQD range of the substrate is less than 4%.

According to another embodiment, the target segment of the difference ZDDs is set in consideration of the ESFQDs of the substrate. For example, if the range of ESFQDs of the substrate placed in the first epitaxial reactor is limited to 0 nm to -50 nm, the target segment of the difference ZDDs is set in accordance with a predetermined reference value. For example, if the predetermined reference value is set to 0.5%, the target segment of the difference ZDDs includes regions 8 and 13.

Thereafter, a correction value is calculated based on the difference ZDD2 of the current sample epitaxial wafer desired to be used for the first epitaxial reactor and the target segment (step S420). Thereafter, the difference value ZDD2 is corrected using the correction value so as to belong to the target section.

For example, the correction value is the difference between the predetermined target value belonging to the target zone and the difference ZDD2. Here, the predetermined target value belonging to the target zone is the lower limit, upper limit, or median in the target zone.

For example, if the target segment is a segment of 0 nm to -5 nm, the difference ZDD2 is 7, and the predetermined target value is the median in the target segment (for example, -2.5 nm), corrected The value is a value obtained by subtracting the difference ZDD (for example, 7) from a predetermined target value (for example, -2.5) (for example, -9.5).

Thereafter, the growth conditions of the first epitaxial reactor are adjusted based on the correction value and the relationship pattern (step S150). The current first growth condition of the first epitaxial reactor is varied by growth conditions corresponding to the correction values.

For example, under the first growth condition, the first growth condition is adjusted to a second growth condition, and the second growth condition corresponds to a value obtained by increasing the correction value to the difference ZDD.

For example, assume a correction value of -9.5 and the current flow rate of hydrogen in the first epitaxial reactor is 90 standard liters per minute (slm). Referring to the linear equation of Figure 5A (y1 = 0.5111x - 36.817), the difference in the current flow rate of hydrogen (90 slm), ZDD, is 9.182.

The value of -0.318 is obtained by adding the correction value (-9.5) to the difference ZDD (9.182), and in the linear equation of hydrogen, if the value of y1 is -0.318, the value of x is 72.352. Therefore, for calibration purposes, the flow rate of hydrogen in the first epitaxial reactor was adjusted from 90 slm to 72.352 slm.

For example, assume a correction value of -9.5 and a current temperature of the first epitaxial reactor of 1,130 °C. Referring to the linear equation of Figure 5B (y2 = 0.6308x - 700.71), the current temperature is 1, 130 ° C, the difference ZDD is 12.094, and for calibration purposes, the temperature of the first epitaxial reactor is adjusted from 1,130 ° C to about 1,123 ° C .

In addition, referring to FIG. 5C, the flow rate of the TCS of the first epitaxial reactor is adjusted for calibration purposes.

Finally, after the growth conditions are adjusted (step S150), the first epitaxial reactor generates an epitaxial wafer under the adjusted growth conditions (step S160). If the first epitaxial reactor produces epitaxial wafers under adjusted growth conditions, regardless of the ESFQD value of the substrate shown in FIG. 4, there is a high probability that the defect rate is within a controllable range, thereby performing a process Control to reduce the defect rate.

That is, the ZDD value varies depending on the flatness level of the substrate, but the ZDD value can be uniformly maintained regardless of the flatness level of the substrate, and in the embodiment, the control range of the difference ZDDs is set to have a low defect rate. The section, whereby the process control is performed regardless of the shape or flatness of the substrate, thereby reducing the defect rate.

As apparent from the above description, in the flatness control method of the epitaxial wafer according to an embodiment of the present invention, process control is performed regardless of the shape or flatness of the substrate, thereby reducing the defect rate.

Although the present invention has been disclosed above in the foregoing embodiments, it is not intended to limit the invention. It is within the scope of the invention to be modified and modified without departing from the spirit and scope of the invention. Please refer to the attached patent application for the scope of protection defined by the present invention.

W1‧‧‧ wafer

S1‧‧‧Location

c‧‧‧Central

θ‧‧‧Specified angle

P1‧‧‧ first point

P2‧‧‧ second point

Section 401‧‧‧

Edge of E1‧‧

S2‧‧‧Specified distance

MAX‧‧‧max

MIN‧‧‧min

PLANE‧‧‧ reference surface

1 is a flow chart of a method for controlling the flatness of an epitaxial wafer according to an embodiment of the present invention. Fig. 2 is a flow chart showing the defect rate pattern for obtaining the flatness shown in Fig. 1 in more detail. 3 is a flow chart for obtaining ZDD1s and ESFQR1s in an embodiment of the present invention shown in FIG. 2. Fig. 4 is a diagram showing a flatness defect rate pattern obtained in step S240. Figure 5A is a graph of the relationship between the differential ZDDs and the hydrogen flow rate supplied to grow the epitaxial layer. Fig. 5B is a graph showing the relationship between the difference ZDDs and the growth temperature of the epitaxial layer. Figure 5C is a graph of the relationship between the differential ZDDs and the TCS flow rate supplied to grow the epitaxial layer. Figure 6 is a flow chart showing the calculation of the correction value of the embodiment of the present invention shown in Figure 1. Figure 7A is a schematic diagram of ESFQD and ESFQR. Figure 7B is a schematic cross-sectional view taken along line a-b of Figure 7A.

Claims (16)

A method for controlling the flatness of an epitaxial wafer includes: obtaining a flatness defect rate pattern of an epitaxial wafer produced by a given epitaxial reactor; obtaining a relationship graph indicating an epitaxial layer The relationship between the growth conditions and the Z-axis Double Derivatives (ZDDs) of the epitaxial wafers used for the relationship; by a first epitaxial reactor in a given epitaxial reactor Obtaining, by the first growth condition, a difference ZDD of the generated at least one sampled epitaxial wafer; calculating a correction value based on the difference ZDD of the at least one sampled epitaxial wafer and the flatness defect rate graph; Adjusting the first growth condition based on the relationship pattern, wherein calculating the correction value comprises: setting a target segment based on the defect rate of the flatness defect rate pattern; determining the flatness defect of the epitaxial wafers based on the Determining the flatness defect of the epitaxial wafers by comparing the ESFQRs of the crystal wafer with a first predetermined reference value, the first predetermined reference value being 100 nm to 120 nm; and setting the target segment , the target section Rate of flatness defect over the entire range of the substrate in which a plurality of ESFQDs smaller than a second predetermined reference value is one segment, the second predetermined reference value is 0.1 to 15%. The method for controlling the flatness of an epitaxial wafer according to claim 1, wherein the first growth condition is adjusted based on the correction value and the relationship pattern. The method for controlling the flatness of an epitaxial wafer according to claim 2, wherein obtaining the flatness defect rate graph comprises: Obtaining the edge sector Site Front side reference Q (Site least squares plane Derivations; ESFQDs) and the Z-axis Double Derivatives (ZDDs) of the edge sectors of the substrates; Based on the epitaxial wafer formed by the substrate, the difference between the ZDDs and the edge sector site front side reference point (Site least squares plane) (Ranges; ESFQRs); based on the epitaxial crystal The rounded ESFQRs determine the flatness defects of the epitaxial wafers; and the flatness defect rate pattern of the epitaxial wafers is obtained based on the difference between the ESFQDs of the substrates and the epitaxial wafers. The method for controlling the flatness of an epitaxial wafer according to claim 3, wherein obtaining the difference ZDDs and ESFQRs of the epitaxial wafers comprises: obtaining ZDDs and ESFQRs of the epitaxial wafers; and using the same The difference between the ZDDs of the substrate and the ZDDs of the epitaxial wafers obtains the difference ZDDs of the epitaxial wafers. The method for controlling the flatness of an epitaxial wafer according to claim 2, wherein calculating the correction value comprises: calculating the correction value based on the difference ZDD of the at least one sample epitaxial wafer and the target segment. The method for controlling the flatness of an epitaxial wafer according to claim 3, wherein in the flatness defect rate pattern of the epitaxial wafers, the X axis represents ESFQDs of the substrates, and the Y axis represents the same The difference ZDDs of the epitaxial wafers, and the difference ZDDs between the ESFQDs of the substrates and the epitaxial wafers, the defect rates of the epitaxial wafers are divided into a plurality of regions. The method for controlling the flatness of an epitaxial wafer according to claim 1, wherein when the relationship pattern is obtained, a difference ZDDs of the epitaxial wafers for relationship generated under different growth conditions is obtained, and the The relationship between the growth conditions and the obtained difference ZDDs of the epitaxial wafers used for the relationship is obtained. The method for controlling the flatness of an epitaxial wafer according to claim 7, wherein the growth conditions and the first growth conditions for obtaining the relationship pattern include a flow rate of hydrogen gas, a flow rate of the TCS, or a growth temperature. The method for controlling the flatness of an epitaxial wafer according to claim 5, wherein the correction value is between a predetermined target value of one of the target segments and a difference ZDD between the at least one sampled epitaxial wafer. Difference. The method for controlling the flatness of an epitaxial wafer according to claim 9, wherein the predetermined target value is a lower limit, an upper limit, or a median of the target segment. The method for controlling the flatness of an epitaxial wafer according to claim 1, wherein, when the first growth condition is adjusted, the first growth condition is adjusted to a second growth condition, and the second growth condition corresponds to the first epitaxial A value obtained by increasing the correction value to the difference ZDD under the first growth condition of the reactor. A method for controlling the flatness of an epitaxial wafer includes: obtaining a frontal reference plane of a plurality of first substrates (Site least squares plane Derivations; ESFQDs) And Z-axis Double Derivatives (ZDDs); Obtaining the edge sector Site Front side reference Q (Site least squares plane) Ranges; ESFQRs and the second Z based on the plurality of first epitaxial wafers formed by the first substrate Axis double derivative (ZDDs); obtaining a first difference ZDDs of the first epitaxial wafers; determining flatness defects of the first epitaxial wafers based on ESFQRs of the first epitaxial wafers, and obtaining a flatness defect rate pattern; obtaining a relationship graph indicating a relationship between an epitaxial layer growth condition of the second epitaxial wafer and a second ZDDs of the second epitaxial wafer; obtaining the first a second difference ZDD of the at least one third epitaxial wafer generated under the growth condition; calculating a correction value based on the second difference ZDD and the flatness defect rate graph; and adjusting based on the correction value and the relationship graph The first growth condition is a second growth condition, wherein calculating the correction value comprises: setting a target segment based on a defect rate of the flatness defect rate pattern; and determining a flatness defect of the first epitaxial wafer, based on the Etc Determining a flatness defect of the first epitaxial wafer by comparing the ESFQRs of the epitaxial wafer with a first predetermined reference value, the first predetermined reference value being 100 nm to 120 nm; and setting the In the target segment, the target segment is a segment in which the flatness defect rate in the full range of the ESFQDs of the first substrates is less than a second predetermined reference value, and the second predetermined reference value is 0.1% to 15%. The method for controlling the flatness of an epitaxial wafer according to claim 12, further comprising generating an epitaxial wafer under the second growth condition. The method for controlling the flatness of an epitaxial wafer according to claim 12, wherein obtaining the relationship pattern comprises: growing an epitaxial layer on the second substrate according to different values of the same growth condition item; a second epitaxial wafer; obtaining a difference ZDDs of the generated second epitaxial wafers; and obtaining a difference graph by using a difference between the different values of the same growth condition item and the second epitaxial wafers . The method for controlling the flatness of an epitaxial wafer according to claim 12, wherein obtaining the second difference ZDD comprises: growing a epitaxial layer on the plurality of third substrates under the first growth condition to generate a plurality of third epitaxial wafers; and obtaining a difference ZDDs of the third epitaxial wafers, and setting an average of the difference ZDDs of the third epitaxial wafers obtained as a second difference ZDD . The method for controlling the flatness of an epitaxial wafer according to claim 12, wherein calculating the correction value comprises: setting a target segment of the flatness defect rate pattern, wherein the flatness defect rate is less than a predetermined reference value; The correction value is calculated such that the second difference ZDD belongs to the target segment.