JP2003282576A - Heat treatment setting method for silicon substrate, method for heat treatment of silicon substrate and manufacturing method for silicon substrate - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a finding method for cooling conditions, capable more effectively developing an IG capacity in accordance with the initial amount of contamination of heavy metal and the density of oxygen deposit of a silicone substrate. <P>SOLUTION: The heat treatment condition of the silicon substrate is set by a method, wherein the relation between a heat treatment temperature T and a heat treatment time t upon continuous cooling is calculated by equations (1): 1/τ=3 D[(C<SB>0</SB>-C<SB>eq</SB>)/(C<SB>p</SB>-C<SB>eq</SB>)]<SP>1/3</SP>[4πn/3]<SP>2/3</SP>, (2): t=τ1n[(CE-C<SB>eq</SB>)/(C<SB>0</SB>-C<SB>eq</SB>)], (3): T=T<SB>0</SB>-R<SB>c</SB>t from the initial heavy metal contamination density C<SB>0</SB>, the density of oxygen deposit n and the density CE of a desired impurity, then, a C-C-T chart, showing the calculated relation between the heat treatment temperature T and the heat treatment time t, is created and a tangential line, passing a point of a temperature T<SB>0</SB>on the temperature axis of the created C-C-T chart, whereat the initial heavy metal contamination density C<SB>0</SB>becomes the degree of solid solution C<SB>eq</SB>of the heavy metal, while contacting with a curve in the C-C-T chart with a maximum inclination, is drawn to determine optimum cooling speed by the inclination of the tangential line. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is to set an effective cooling rate when removing impurities by so-called internal gettering, in which impurities in a silicon wafer are gettered to defects in a silicon substrate. The present invention relates to a technique for manufacturing a device with a silicon substrate whose surface is clean.

[0002]

2. Description of the Related Art Conventionally, the Czochralski method has been used as a semiconductor wafer for producing semiconductor devices such as IC and LSI.
Although a silicon single crystal grown by the (CZ method) or the floating zone melting method (FZ method) is used, if any heavy metal contamination occurs during the heat treatment for device fabrication, the completed device operation will be adversely affected. Reach Therefore, a method has been developed in which, when heavy metal impurities are mixed in the wafer, the heavy metal impurities are removed from the surface, which is the device operation region, and confined inside or on the back surface of the wafer.

This is called a gettering technique. This gettering method is distinguished by the position of the wafer in which the heavy metal impurities are confined, and the method of confining inside the wafer is determined by the internal gettering method.
g: IG), the method of confining it on the back side is called External Gettering (EG).

A typical example of the former is a method of forming oxygen precipitates inside the wafer and capturing heavy metal impurities therein. In the latter, a sand blast method for forming a mechanical strain layer on the back surface or a poly silicon for depositing a polycrystalline silicon film. Back seal method (Pol
y-Si Back Seal: PBS) is available. Conventionally, an excellent silicon single crystal wafer has been produced which has the property of removing heavy metal impurities from the device operating region by adding these gettering methods individually or in combination and adding them to the silicon single crystal wafer.

By the way, when impurities are removed by IG, C
The oxygen precipitates formed in the Z silicon substrate are used as impurity precipitation nuclei, and the impurities are diffused into the oxygen precipitates during the device fabrication heat treatment to be precipitated. When the rate at which the impurities are actually captured is defined as the gettering rate, it is known that the gettering rate has temperature dependence.

That is, at high temperatures, the diffusion of impurities in the silicon substrate is fast, but since the solid solubility of the impurities is also high, the amount that can be trapped in oxygen precipitates is small, and the trapping rate of impurities is relatively low. Since the solid solubility is low, the driving force for impurity precipitation is high, but since the diffusion is slow, the gettering speed is low.

Therefore, although optimum gettering should be performed at an appropriate temperature, conventionally, there is no method for setting an appropriate heat treatment condition. Therefore, the heat treatment condition should be an empirically obtained condition. Or, it was determined by the convenience of device fabrication in the device process.
Therefore, it often takes unnecessary time to obtain a silicon substrate having a desired impurity concentration (a silicon substrate having a desired impurity concentration by gettering).

The invention made by the inventor of the present application to solve such a problem (see Japanese Patent Application Laid-Open No. 11-283986) is an epoch-making method capable of setting the optimum heat treatment temperature and time. From the initial heavy metal contamination concentration in the silicon substrate, the oxygen precipitate density, and the desired impurity concentration, FIG.
T-T-T diagram (Time-Temperature-Transformat
By creating a drawing called ion diagram)
The optimum heat treatment conditions can be set easily. This figure shows the relationship between the heat treatment temperature and the time for reducing the concentration of heavy metal impurities to the target concentration, and the optimum heat treatment condition can be determined at the shortest part of the heat treatment time called the nose. Also, from this figure, it is clear at a glance how much time is required at a certain temperature even when the heat treatment cannot be performed under the optimum conditions due to other circumstances.

However, in the actual device manufacturing process, the low temperature heat treatment step of isothermal holding for reducing heavy metal contamination is disadvantageous in terms of cost and is not performed so often. Therefore, in reality, the effect of removing heavy metals is obtained in the continuous cooling step when the wafer is taken out from the high temperature heat treatment. However, the T-T-T diagram is originally
It shows how the concentration of heavy metals decreases due to isothermal holding.Although it can be applied during cooling as a qualitative rough estimate,
Strict handling is not possible. Therefore, when the cooling process is regarded as a gettering process, there is no strict sense of the optimal cooling time setting method.

[0010]

As described above, the conventional method is a qualitative condition setting method from the viewpoint of removing impurities in the continuous cooling step, and may not always be the optimum condition. It was The present invention has been made in view of the above problems, and provides a method for finding a cooling condition that can more effectively exert the IG capability, depending on the initial heavy metal contamination amount of the silicon substrate and the oxygen precipitate density. To aim.

[0011]

According to the present invention for solving the above-mentioned problems, in a method of setting a heat treatment condition of a silicon substrate when performing internal gettering on the silicon substrate, an initial heavy metal contamination concentration in the silicon substrate is set. C ₀ ,
From the oxygen precipitate density n and the desired impurity concentration C _E , the relationship between the heat treatment temperature T and the heat treatment time t in the case of continuous cooling is calculated from the following equations (1), (2) and (3), A C-C-T diagram showing the relationship between the calculated heat-treatment temperature T and heat-treatment time t was created, and the initial heavy metal contamination concentration C ₀ was silicon on the temperature axis of the created C-C-T diagram. A tangent line is drawn through the point of the temperature T ₀ at which the solid solubility of the heavy metal is C _eq and the curve of the CCT diagram has the maximum slope, and the optimum cooling rate is determined by the slope of the tangent line. This is a method for setting the heat treatment conditions for the silicon substrate. 1 / τ = 3D [(C ₀ -C _eq ) / (C _p -C _eq )] ^1/3 [4πn / 3] ² /3… (1) t = -τln [(C _E -C _eq ) / _{(C 0 -C eq)] ...} (2) T = T 0 -R c t ... (3) ( wherein, C _0: initial heavy metal contamination concentration (atoms / cm ^3), C _E: desired impurity concentration ( Atom / cm ³ ), C _eq : solid solubility of heavy metal in silicon (atom / cm ³ ), k: Boltzmann constant, T: heat treatment temperature (K), t: heat treatment time
(s), 1 / τ: gettering rate (s ^-1 ), D: diffusion coefficient of heavy metals (cm ² / s), C _p : heavy metal concentration in precipitates, n: oxygen precipitate density (internal microdefects) Density, pieces / cm ³ ) T ₀ : Temperature at which the initial heavy metal contamination concentration becomes solid solubility, R _c : Cooling rate. )

As described above, the continuous heavy metal contamination concentration in the silicon substrate, the oxygen precipitate density, and the desired impurity concentration are continuously calculated from the above equations (1), (2) and (3). From the relationship between the heat treatment temperature T and the heat treatment time t during cooling, C-CT (Continuas-Cooling-Transformation D
iagram) diagram is created, and the point of temperature T _{0 at which} the initial heavy metal contamination concentration C ₀ becomes the solid solubility C _eq of heavy metal in silicon is plotted on the temperature axis of the created C-C-T diagram. Street C-C-T
Draw a tangent line that is tangent to the curve of the diagram so as to have the maximum slope, and determine the optimum cooling rate by the slope of the tangent line to set the heat treatment conditions, so that there is little error with the actual heat treatment, Gettering occurs during cooling with extremely high precision, and it is possible to set cooling conditions that can reduce the concentration of heavy metal to a desired level.

In this case, it is preferable to calculate the relationship between the heat treatment temperature and the heat treatment time in the case of continuous cooling by the equations (1), (2) and (3) by the difference method (claim 2). . In this way, by performing the analysis of the above equation using the difference method, the analysis can be performed relatively easily.
-A C-T diagram can be created.

Further, the present invention is a method for heat treating a silicon substrate, characterized in that the silicon substrate is heat-treated under the heat treatment conditions set by the method of the present invention (claim 3). If the heat treatment is performed by this method, the heat treatment can be performed under cooling conditions suitable for removing the impurities, and thus the impurities can be efficiently removed from the silicon substrate.

Further, the present invention is a method for manufacturing a silicon substrate, which comprises at least the heat treatment step of the present invention (claim 4). Such a method of manufacturing a silicon substrate is a method of reliably manufacturing a silicon substrate having a desired impurity concentration by gettering. Therefore, it is possible to surely improve the device yield.

Further, the present invention is a program for causing a computer to set the heat treatment conditions of a silicon substrate, the program causing the computer to set the heat treatment conditions by the method for setting the heat treatment conditions of the present invention. A program for causing a computer to set heat treatment conditions for a silicon substrate (claim 5).

As described above, the program for executing the present invention causes the computer to calculate the heat treatment condition by the method of the present invention. Therefore, if the computer is made to set the heat treatment condition by this, the program can be performed very accurately. Gettering occurs during cooling, and the heat treatment conditions of the wafer that can reduce the concentration of the heavy metal to a desired value can be set easily and easily.

Further, the present invention is a recording medium on which the program of the present invention is recorded (claim 6).
As described above, if the program of the present invention is recorded in the recording medium, it can be input to each computer and used at a necessary place when necessary, which is extremely convenient.

The present invention will be described in more detail below.
The inventors of the present invention have hitherto been able to perform appropriate heat treatment conditions by subjecting silicon substrates having various initial heavy metal contamination concentrations and oxygen precipitate densities to a large number of silicon substrates under various heat treatment conditions for a long time. However, as a result of various studies as to whether or not it is possible to more easily find out by numerical calculation, it was found that the above-mentioned invention (JP-A-11-2
83986). However, this method is effective only for isothermal holding, and it is possible to qualitatively grasp the tendency of continuous cooling, but to grasp the state of impurity reduction during continuous cooling in actual strict cases. I couldn't.

Therefore, the present inventor has developed, in order to perform strict handling at the time of cooling, a method that takes into account the temperature change which changes momentarily during cooling, in addition to the above-mentioned method. Therefore, when continuous cooling is used as an impurity reduction process (gettering process), the initial heavy metal contamination concentration, the oxygen precipitate density, and the desired impurity concentration can be determined simply by numerical calculation in a short time and with high accuracy. It became possible to know the more effective cooling rate.

Particularly, in the present invention, the above formulas (1) and (2) are used.
A C-C-T diagram is created from the formula and the formula (3), and the cooling rate is determined on the created C-C-T diagram, thereby significantly improving the accuracy of heat treatment condition setting. It is characterized by points.

Then, the heat treatment of the silicon substrate is performed under the heat treatment conditions set by the above method,
When a silicon substrate is manufactured, it is possible to easily and surely obtain a silicon substrate having a desired impurity concentration because it is possible to set a heat treatment condition with high precision, which is not adventitious and empirically determined as in the past. it can.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION The embodiments of the present invention will be described below by exemplifying the contents of a model for concrete calculation, but the invention is not limited thereto. Fig. 2 shows the amount of Fe when the density of oxygen precipitates at the heat treatment temperature of 280 ° C is 8 × 10 ⁹ atoms / cm ³ and the initial contamination amount is 8 × 10 ¹² atoms / cm ³ to 2 × 10 ¹⁴ atoms / cm ³ . The gettering rate, that is, the state of solid solution Fe concentration attenuation due to the precipitation of Fe in oxygen precipitates is shown. Thus, the concentration of superatomic solid solution Fe atoms in the silicon substrate decreases in accordance with the exponential function of time (Tobe et al .; Japan Society of Applied Physics, Silicon Technology No. 5, 9th November 1998, p. 44-49
reference). The state is shown by the following equation (4). C = C ₀ + (C ₀ -C _eq ) exp (-t / τ) 1 / τ = 4πD nr (4) (where C: solid solution Fe concentration (atoms / cm ³ ), C ₀ : initial pollution Concentration (atoms / cm ³ ), C _eq : Solid solubility of Fe in silicon (= 4.3 × 10 ²² exp (-2.1 eV /
kT), atoms / cm ³ ) (see M.Aoki et al; J.Appl.Phys. 72 (3) (1992) 895-898), k: Boltzmann constant, T: absolute temperature (K), t: heat treatment Time (s), 1 / τ: Gettering speed (s ^-1 ), D: Diffusion coefficient of Fe (= 1.3 × 10 ^-3 exp (-0.68 eV / kT), cm ² /
s) (ERWeber; Appl.Phys.A30 (1983) 1.) n: Oxygen precipitate density (internal microdefect (BMD) density, number / c
m ³ )).

Hereinafter, this 1 / τ will be referred to as a gettering speed. It can be seen from FIG. 2 that the progress of gettering can be accurately expressed by the equation (4). However, in the actual impurity precipitation process, the radius of impurities such as Fe is not constant and increases with the precipitation of impurities. Therefore, deriving the gettering speed from equation (4) may not match the actual case. Therefore, the inventor of the present application has proposed the following equation (1) in consideration of the change in the impurity radius in the above-mentioned Japanese Patent Laid-Open No. 11-283986. This (1)
According to the formula, the gettering speed that is actually present can be derived by numerical calculation. 1 / τ = 3D [(C ₀ -C _eq ) / (C _p -C _eq )] ^1/3 [4πn / 3] ^2/3 (1) (where C _p : Fe precipitate (FeSi ₂ )
e concentration (= 25.6 × 10 ²¹ atoms / cm ³ )).

Next, using the above equation (1), TT-
A method for obtaining a T-line diagram will be shown. For example, although the gettering speed 1 / τ at each temperature can be calculated by the equation (1), a desired impurity concentration which may remain in the wafer after the gettering is finished is C _E. In this case, assuming that C in the equation (4) is equal to C _E , the time at that time is calculated from the equation (4). The time t is shown by the following equation (2). t = -τln [(C _E -C _eq ) / (C ₀ -C _eq )]… (2)

In this equation (2), τ and C _eq are functions of temperature. Therefore, the time t at which the residual Fe concentration represented by the equation (2) becomes C _E is also a function of temperature. For a certain temperature (1)
Substituting τ obtained by using the equation into the equation (2) to obtain the time t, and plotting the time t against the temperature is T−
FIG. 4 is a TT diagram, an example of which is FIG. 3 already shown. In this example, the oxygen precipitate density is 1 × 10 ⁹ atoms / cm ³ , the initial contamination concentration is 1 × 10 ¹³ atoms / cm ³ , and the desired impurity concentration C _E atoms / cm ³ .
The case of cm ³ is calculated. In FIG. 3, it can be seen that the shortest temperature and time at which the residual Fe concentration reaches the target target concentration C _E is the point a that forms the nose (nose) of the curve A in FIG. That is, the optimum heat treatment for gettering is 610 ℃, 1
It is an isothermal heat treatment for 10 minutes, and a silicon wafer having a desired residual Fe concentration can be obtained even if it is optionally cooled thereafter.

When it is desired to progress the gettering to a desired concentration by cooling at a certain speed instead of the above-described isothermal heat treatment, at a temperature at which the initial contamination concentration at time 0 in FIG. 3 corresponds to the solid solubility. Draw a straight line from point b to point a. The slope of this straight line is the optimum cooling rate, and in the example of Fig. 3, it is -2.4 ° C.
/ Minute.

Alternatively, as shown by the curve B in FIG.
It is also shown that gettering is desired to proceed to C _E = 1 × 10 ¹² atoms / cm ^{3 in the} equation (2). In this way, a T-T-T diagram for each condition can be created, and the optimum isothermal heat treatment temperature for gettering and the minimum time required for it can be obtained, or the optimum cooling temperature can be obtained.

However, as can be seen by looking at the procedure for creating the T-T-T diagram described above, the T-T-T diagram is basically a relationship diagram of heat treatment temperature and time assuming isothermal holding. Therefore, when this is applied to the cooling process, it can only serve as a constant guide. Therefore, in the present invention, in order to strictly handle cooling, among the factors in the formula (1), C _eq, which is a factor depending on temperature, is solid solubility of heavy metal in silicon and D is diffusion coefficient of heavy metal. Was not a constant value, and the temperature that changed every moment during cooling was recaptured as a function of time. Ie current temperature
T can be expressed by the following equation (3). _{T = T 0 -R c t ...} (3)

Here, T ₀ is the temperature at which the initial pollution concentration and the solid solubility coincide with each other, and corresponds to the cooling start temperature. R _c is the cooling rate and t is the time. Considering this, the time on the left-hand side of equation (2) does not become a simple function of temperature, but becomes a so-called hidden equation, which makes it impossible to solve analytically.

Therefore, a numerical solution is needed. As an embodiment of the present invention, it is preferable to use a difference method that is relatively simple and easy to handle. By this difference method,
FIG. 1 is a C-C-T diagram showing the temperature and time for the Fe concentration to decay to the target concentration during continuous cooling. Although the figure is similar to the T-T-T diagram of FIG. 3, it can be seen that the reduction achievement curve up to the target concentration recedes toward the long time side. Since this is continuous cooling, at the start of cooling,
This is because the degree of supersaturation of Fe is not sufficient, and the gettering behavior does not proceed unless the cooling progresses to a low temperature to some extent. That is, the time shifts in the necessary direction as compared with the case of holding the same temperature. This figure shows the situation well.

From this C-C-T diagram, in order to obtain the optimum cooling rate, the temperature at which the initial contamination concentration on the temperature axis coincides with the solid solubility and the maximum toward the curve of the C-C-T diagram. A tangent line is drawn so that it becomes the inclination of, and the inclination becomes the optimal cooling rate. Figure 1
In the example, the initial contamination was 10 ¹³ atoms / cm ³ , and the BMD density was 10 ⁹ atoms / cm ³
Under the conditions, if the target reduction concentration is 10 ¹² atoms / cm ³ , -6 ℃ /
If the reduction target concentration is 10 ¹¹ atoms / cm ³ , the maximum cooling rate is -1 ° C / min. If the cooling rate is slower than the optimum cooling rate thus determined, the Fe concentration can be attenuated to the reduction target value even if an arbitrary cooling rate is used.

The C-C-T diagram shows the reduction time to the target concentration of gettering with respect to temperature during continuous cooling. Therefore, the case of isothermal holding cannot be considered using this C-C-T diagram. In the case of isothermal holding, the TT diagram shown above is more rigorous.
However, as described above, in the actual device manufacturing process, low temperature heat treatment for isothermal holding is not often performed,
In most cases, the method using the C-C-T diagram of the present invention is advantageous.

The heat treatment condition called the optimum cooling rate in the present invention means that it is the optimum condition and the shortest condition for removing Fe up to a predetermined concentration by IG.
When actually manufacturing a silicon substrate, it may be carried out under conditions other than the above optimum heat treatment conditions due to the convenience of the device process and the like.

That is, when setting the cooling rate, for example, the condition of the tangent line having the maximum inclination is not necessarily limited, and if the cooling is performed from the tangential line having the maximum inclination, the present invention can be sufficiently performed. The effect of
The heat treatment conditions can be appropriately changed or modified according to the conditions under which the present invention is carried out.

[0036]

EXAMPLES Hereinafter, the present invention will be specifically described with reference to Examples and Comparative Examples, but the present invention is not limited thereto. (Example) By the method of setting the heat treatment conditions of the silicon substrate according to the present invention, the heat treatment temperature and the heat treatment in the case of continuous cooling were determined from the initial heavy metal contamination concentration in the silicon substrate, the oxygen precipitate density, and the desired impurity concentration. A C-C-T diagram is created by calculating the time relationship, and the C-C-T diagram is created.
The optimum cooling rate was determined from the CT chart. Next, in order to verify the result obtained by the above calculation, an experiment in which a silicon substrate was actually heat-treated was conducted and compared with the above calculation prediction.

First, in this example, the initial heavy metal contamination concentration and the oxygen precipitate density of the silicon substrate which was actually subjected to the heat treatment were measured. The measurement result shows that the initial heavy metal contamination concentration is 1 × 1
The concentration was 0 ¹³ atoms / cm ³ , and the density of oxygen precipitates was 1 × 10 ⁹ atoms / cm ³ . Accordingly, also in the above heat treatment conditions, the heat treatment conditions were set such that the initial heavy metal contamination concentration C ₀ = 1 × 10 ¹³ atoms / cm ³ and the oxygen precipitate density n = 1 × 10 ⁹ atoms / cm ³ . The desired impurity concentration C _E was C _E = 10 ¹¹ atoms / cm ^3, and the optimum cooling rate in this case was set.

The relationship between the heat treatment temperature and the heat treatment time in the case of continuous cooling is calculated from the equations (1), (2) and (3), and a C-C-T diagram as shown in FIG. 1 is prepared. did. The calculation of the relationship between the heat treatment temperature and the heat treatment time in this case was performed using the difference method. Then, as shown in FIG. 1, the temperature of the created C-C-T diagram passes through the point where the initial heavy metal contamination concentration becomes the solid solubility of heavy metal in silicon (824 ° C. in this case). , Desired impurity concentration C _E
A tangent line was drawn to obtain the maximum slope with the curve for = 10 ¹¹ atoms / cm ³ . As a result, the optimal cooling rate -1 ℃ /
Got a minute.

Next, in order to verify the results obtained by the above calculation, the cooling rate -1 obtained by the calculation was calculated from the temperature 824 ° C. at which the initial contamination concentration became the solid solubility on the above-mentioned silicon substrate.
An experiment was conducted to remove heavy metal impurities by performing a heat treatment of cooling at ℃ / min. As a result, the impurity concentration of the silicon substrate after cooling was C _E = 0.9 × 10 ¹¹ atoms / cm ³ , which was a target value. Therefore, it can be seen that the correctness of the calculation according to the present invention has been proved.

(Comparative Example) The T-T-T diagram was obtained by calculating the relationship between the heat treatment temperature and the heat treatment time from the initial heavy metal contamination concentration in the silicon substrate, the oxygen precipitate density, and the desired impurity concentration. Was prepared and the optimum cooling rate was determined from the T-T-T diagram. Next, in order to verify the result obtained by the above calculation, an experiment in which a silicon substrate was actually heat-treated was conducted and compared with the above calculation prediction.

As in the example, the initial heavy metal contamination concentration C ₀ = 1
The optimum cooling rate in this case was set at × 10 ¹³ atoms / cm ³ , oxygen precipitate density n = 1 × 10 ⁹ atoms / cm ³ , and desired impurity concentration C _E = 10 ¹¹ atoms / cm ³ . However, in this comparative example,
From the above values, a T-T-T diagram as shown in FIG. 3 was created from the equations (1) and (2). And the created T
The temperature at which the initial heavy metal contamination concentration becomes the solid solubility of heavy metal in silicon on the temperature axis of the -TT diagram (in this case, 824
(B) of the curve A and the point a, which is the nose point of the curve A for the desired impurity concentration C _E = 10 ¹¹ atoms / cm ³ , were drawn. The optimum cooling rate of -2.4 ° C / min was obtained from the slope of this line.

Next, in order to verify the results obtained by the above calculation, from a temperature of 824 ° C. at which the initial contamination concentration becomes a solid solubility on a silicon substrate having the same initial heavy metal contamination concentration and oxygen precipitate density as in the example. , Cooling rate calculated -2.4 ℃ /
An experiment was conducted to remove heavy metal impurities by performing heat treatment for cooling by minutes. As a result, the impurity concentration of the silicon substrate after cooling was C _E = 4.9 × 10 ¹¹ atoms / cm ³ , which was not the target value.

The present invention is not limited to the above embodiment. The above-described embodiments are merely examples, and the present invention has substantially the same configuration as the technical idea described in the scope of claims of the present invention, and has any similar effects to the present invention. It is included in the technical scope of.

For example, the heat treatment conditions referred to in the present invention refer to all the heat treatments added in the process of processing a silicon single crystal ingot into a wafer, the device manufacturing process, etc., and accordingly, the heat treatment conditions of the silicon substrate to be set. The present invention is not limited to the case after the specific device step, and in any case after the wafer processing, during the device step, after the device step, etc., if the heat treatment is applied to the silicon substrate, the present invention Can be applied to set heat treatment conditions.

In the above, Fe is used as a heavy metal impurity.
However, the present invention is not limited to this, and Cu, Ni, Co
It is naturally applicable and effective for removing other heavy metal contamination.

[0046]

As described above, according to the present invention,
In the method of setting the heat treatment conditions for the silicon substrate when performing internal gettering on the silicon substrate, the initial heavy metal contamination concentration in the silicon substrate, the oxygen precipitate density, and the desired impurity concentration, in an extremely short time, The cooling rate of heat treatment can be calculated easily and accurately,
Thereby, proper heat treatment conditions can be set. In particular, the present invention can accurately set the cooling conditions when the continuous cooling step in the actual device manufacturing process is regarded as the gettering process, and is easier to apply to the actual manufacturing step than the conventional method. .

Therefore, it is not necessary to actually perform heat treatment for a long time to find out the conditions by using a large number of wafers having different initial heavy metal contamination concentrations and oxygen precipitate densities. Heat treatment conditions can be determined. And since it is an ad hoc method, which is not empirically determined as in the past, and the heat treatment conditions are accurate, it is possible to improve the device yield when the substrate is actually flown into the manufacturing process of the silicon substrate. Can be achieved.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a C-C-T diagram.

FIG. 2 is a diagram showing a relationship between a solid solution Fe concentration and a heat treatment time.

FIG. 3 is a diagram showing an example of a TT-T diagram.

Claims (6)

[Claims] 1. A method for setting a heat treatment condition for a silicon substrate when internal gettering is performed on the silicon substrate, wherein an initial heavy metal contamination concentration C in the silicon substrate is set.
₀ , the oxygen precipitate density n, and the desired impurity concentration C _E ,
The relationship between the heat treatment temperature T and the heat treatment time t in the case of continuous cooling is calculated from the following equations (1), (2), and (3), and the relationship between the calculated heat treatment temperature T and the heat treatment time t is C- C-
A T diagram is created, and the initial heavy metal contamination concentration C ₀ passes through the temperature T ₀ point at which the initial heavy metal contamination concentration C ₀ becomes the solid solubility C _eq of heavy metal in silicon on the temperature axis of the created C-C T diagram. A method for setting a heat treatment condition for a silicon substrate, which comprises drawing a tangent line tangent to the curve of the C-C-T diagram so as to have a maximum slope, and determining the optimum cooling rate based on the slope of the tangent line. 1 / τ = 3D [(C ₀ -C _eq ) / (C _p -C _eq )] ^1/3 [4πn / 3] ² /3… (1) t = -τln [(C _E -C _eq ) / _{(C 0 -C eq)] ...} (2) T = T 0 -R c t ... (3) ( wherein, C _0: initial heavy metal contamination concentration (atoms / cm ^3), C _E: desired impurity concentration ( Atom / cm ³ ), C _eq : solid solubility of heavy metal in silicon (atom / cm ³ ), k: Boltzmann constant, T: heat treatment temperature (K), t: heat treatment time
(s), 1 / τ: gettering rate (s ^-1 ), D: diffusion coefficient of heavy metal (cm ² / s), C _p : concentration of heavy metal in precipitates, n: density of oxygen precipitates (internal microdefects) Density, pieces / cm ³ ) T ₀ : Temperature at which the initial heavy metal contamination concentration becomes solid solubility, R _c : Cooling rate. ) 2. The method for setting the heat treatment conditions for a silicon substrate according to claim 1, wherein the relationship between the heat treatment temperature and the heat treatment time in the case of continuous cooling according to the equations (1), (2) and (3). The method of setting the heat treatment conditions of the silicon substrate, characterized in that is calculated by the difference method. 3. A method for heat treating a silicon substrate, which comprises subjecting the silicon substrate to heat treatment under the heat treatment conditions set by the method according to claim 1. 4. A method of manufacturing a silicon substrate, comprising at least the heat treatment step according to claim 3. 5. A program for causing a computer to set heat treatment conditions for a silicon substrate, the program causing a computer to set the heat treatment conditions according to the method of claim 1 or 2. A program for causing a computer to set heat treatment conditions for a silicon substrate. 6. A recording medium on which the program according to claim 5 is recorded.