JP2003086595A - Determining method of manufacturing condition - Google Patents

Abstract

PROBLEM TO BE SOLVED: To determine the manufacturing condition of a wafer having a desired COP free depth in a short time and at a low cost. SOLUTION: A heating temperature and a holding time in a heat treatment stage and a growing speed and temperature gradient in an ingot manufacturing stage are determined based on the group of the pieces of ingot manufacturing condition data obtaining a COP size formed on an ingot calculated based on the growing speed and the temperature gradient in the ingot manufacturing stage by Czochralski method and a group of the pieces of heat treatment condition data obtaining the COP free depth formed on a wafer surface part calculated based on the heat treatment temperature of the wafer and its holding time in the heat treatment stage of the wafer and the COP size formed at the wafer.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a cavity origin (Crystal Originated Particle;
OP) is a region where there is no desired depth has a desired depth, and more specifically, from the wafer surface in the region where there are no cavity defects formed in the wafer without performing actual equipment test. The present invention relates to a method for determining a wafer manufacturing condition for making a desired depth (hereinafter referred to as a COP-free depth) a desired depth.

[0002]

2. Description of the Related Art A wafer (generally referred to as a semiconductor silicon substrate) on which devices such as ICs and LSIs are formed is generally cut out from a single crystal grown by the Czochralski method (hereinafter referred to as CZ method). Are used. The wafer preferably has few defects to form the device. In particular, due to recent demands for high performance and high integration of devices, the density and size allowed for defects formed on a wafer have become increasingly severe.

Usually, COPs having a size of about 0.1 μm are formed at a density of 10 ⁴ to 10 ⁶ COPs / cm ³ on a wafer produced by cutting out an ingot grown by the CZ method. For example, it is known that a large COP existing in the device active region, that is, in the range of a few μm depth from the wafer surface deteriorates the gate oxide film characteristics of a highly integrated MOS device.

Therefore, various methods for reducing the COP density and COP size (hereinafter referred to as COP size) in the device active region have been studied. As an example of such a method, there is a method of controlling the growth rate when pulling up an ingot by the CZ method. The COP density and COP size depend on the diameter of the ingot and the growth conditions. Generally, the COP becomes low density / large size as the ingot diameter increases, and high density / small size as the growth speed increases. Turn into. Therefore, the growth rate is determined with respect to the diameter of the pulled ingot.

As another method, there is a method of growing an epitaxial layer having a thickness of several μm on the wafer surface by a vapor phase reaction. In this method, since the epitaxial layer is formed so that defects are not introduced, a wafer having almost no COP in the device active region can be obtained. Furthermore, there is also a method of subjecting the wafer to a heat treatment at about 1200 ° C. to eliminate the COP near the surface of the wafer to secure the device active region.

These methods are appropriately selected and used in consideration of a device manufacturing process to be performed later. However, today, due to the progress in high performance and high integration of devices, among these methods, the method of epitaxial growth or the method of extinguishing the COP near the surface of the wafer by heat treatment is mainly used. This trend is expected to increase as wafers with a diameter of 300 mm become the mainstream in the future.

[0007]

In order to secure a device active area, a COP near the surface of the wafer is subjected to a heat treatment.
When the method of eliminating heat is adopted, it is important to properly manage the heat treatment conditions. Heat treatment at high temperature, eg about 12
If it is performed at 00 ° C., slip dislocations may occur in the wafer, so the heat treatment temperature is preferably low. In particular, when a large-diameter wafer having a diameter of 300 mm is heat-treated at a high temperature, slip dislocation is likely to occur and the quality of the wafer itself is likely to deteriorate.

COP existing near the surface of the wafer
It is also necessary to control the heat treatment conditions according to the COP size in order to eliminate. Due to the heat treatment, the COP-free depth is
Depends on COP size and heat treatment conditions. Since the smaller the COP size, the more the COP disappears at a lower temperature, it is necessary to manage the ingot pulling conditions (growth rate, temperature gradient) such that the COP size becomes as small as possible when the ingot is pulled up.

In order to reduce the COP size before heat treatment, for example, the growth rate at the time of ingot pulling up is increased, and the temperature gradient in the crystal axis direction (pulling axis direction) is controlled (K. Nakamura et al., Proceedings of Kazusa Akademia
Forum, (1999), p.55) or nitrogen addition to ingot (M.Tamatsuka et al., Defects in Silicon III, (199
9), p.456) and other technologies have been developed.

As described above, although a tentative guideline for determining the heat treatment condition or the ingot pulling condition has been clarified, there is no method for actually determining these conditions uniquely.

Since the required COP free depth differs depending on the manufacturing process of the device formed on the wafer, it is very difficult to find heat treatment conditions and ingot pulling conditions for obtaining such COP free depth. It requires man-hours and cost. In recent years, since the life of the device itself is short, it is necessary to manufacture a wafer having a necessary COP-free depth in accordance with it, and in a short time and at low cost, the manufacturing conditions of a wafer having a desired COP-free depth can be set. It was requested to develop a decision method.

[0012]

[Means for Solving the Problems]
In order to uniquely determine the ingot pulling conditions (growth rate, temperature gradient) and heat treatment conditions (heat treatment temperature, holding time) for wafers with a P-free depth, actual machine tests were not performed and computer simulation (hereinafter, simulation We examined an algorithm that can determine these conditions only by omitting). This is because it is currently known that the simulation related to wafer fabrication is very consistent with the actual device test.

Therefore, various types of simulation are used.
While preliminarily obtaining each heat treatment condition that will be the COP free depth, we also seek each ingot growing condition that can obtain various COP sizes, and based on the combination of these conditions,
In order to manufacture a wafer having a desired COP-free depth, for example, by selecting an appropriate growth rate for an ingot from a plurality of growth rates and selecting other conditions corresponding to the growth rate, It was considered to decide each heat treatment condition and ingot growing condition of.

The present invention has been completed based on the above knowledge, and the gist thereof is the following method for determining the wafer manufacturing conditions.

The method for determining the manufacturing conditions of a wafer having a desired COP-free depth according to claim 1 is formed in an ingot calculated based on a growth rate and a temperature gradient at the time of pulling the ingot by the Czochralski method. CO
Obtain the ingot manufacturing condition data group that determined the P size, the heat treatment temperature and the holding time of the heat treatment performed on the ingot cut wafer, and the COP free depth formed on the wafer surface calculated based on the COP size. Based on the heat treatment condition data group, the heating temperature and the holding time of the heat treatment to be applied to the wafer, and the growth rate and the temperature gradient when pulling the ingot are determined.

The method for determining the manufacturing conditions for a wafer having a desired COP-free depth according to claim 2 is as follows: a growth rate and a temperature gradient at the time of pulling an ingot by the CZ method, and a heat treatment of a heat treatment applied to a wafer from which the ingot is cut out. The temperature and the holding time thereof are determined by the steps (1) to (5) below. (1) Heat treatment temperature and holding time of the wafer, and the step of obtaining the COP free depth calculated with the COP size formed in the ingot as a parameter (2) From the calculation result of (1) above, the desired COP free depth Step of selecting multiple combinations of heat treatment temperature, holding time and COP size that can obtain (3) For the COP size of each combination selected in (2) above, the growth rate and temperature gradient to obtain that COP size From the plurality of growth rates calculated in step (4) above (3), for the ingot to be produced, a step (5) above (4) of adopting an appropriate growth rate for the growth rate during pulling up of the ingot The temperature gradient, heat treatment temperature, and holding time that correspond to the growth rate adopted in 1. are set to the temperature gradient when pulling the ingot, the heat treatment temperature of heat treatment applied to the ingot-cut wafer, and the holding time. The step of use

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention relates to manufacturing conditions for manufacturing a wafer having a desired COP-free depth, more specifically, a growth rate f at the time of pulling an ingot in the CZ method.
It is an invention of a method for determining a wafer manufacturing condition that uniquely determines _p and a temperature gradient G _c , a heat treatment temperature T of heat treatment performed on a wafer from which an ingot is cut, and a holding time t at the heat treatment temperature. In describing the present invention, first, a series of steps for forming a wafer will be described. In addition, below, it demonstrates according to the invention described in the above-mentioned claim 2.

FIG. 1 is a diagram schematically showing a series of steps from ingot formation by the CZ method to heat treatment of a wafer. First, in the CZ method, a seed of silicon that becomes a nucleus of a single crystal is attached to a wire, and the silicon melted in a quartz crucible is gradually pulled up based on the seed to form an ingot (FIG. 1A). At this time, the characteristics of the ingot depend on the pulling rate, that is, the growth rate f _p in the present invention, and the temperature at the position of the ingot during the pulling, that is, the temperature gradient G _c in the present invention.

The right diagram in FIG. 1 (a) schematically shows the temperature at the position of the ingot, and by appropriately controlling the temperature gradient G _c , a good quality ingot can be obtained. In an actual machine, such a temperature gradient can be adjusted by disposing a shield plate on the side of the ingot.

The ingot obtained as described above is
After cutting off the head and tail, the wafer is cut into a wafer having a thickness of about 300 to 500 μm by a slicer through processes such as inspection (FIG. 1B).

Finally, heat treatment is applied to obtain a wafer containing no COP in the surface region (FIG. 1 (c)). In the heat treatment, by controlling the heat treatment temperature T and its holding time t, the thickness of the region not containing COP, that is, COP
Free depth can be adjusted. Next, the method for determining the wafer manufacturing conditions of the present invention will be described. In the present invention, the wafer manufacturing conditions are determined through the steps (1) to (5) as described above.

FIG. 2 is a diagram showing the flow of steps for determining the wafer manufacturing conditions in the present invention. 2 (1) to (5) correspond to the steps (1) to (5) defined in the present invention. In the following, the steps (1) to (5) will be described, and a case will be described as an example in which the manufacturing conditions of a wafer having an arbitrary COP free depth are determined.

First, in the step (1), the heat treatment temperature T and the holding time t of the wafer cut out from the ingot,
And COP size 2R formed in the ingot (R is COP
Calculate the COP-free depth calculated by using the average radius of) as a parameter. This corresponds to simulating the heat treatment step of FIG. 1 (c) by a computer, and the COP free depth can be calculated by assuming the following steps (a) and (b).

(A) Dissolution of inner wall oxide film Normally, an oxide film having a thickness of about 3 to 4 nm is formed on the inner wall of the COP, and the inner wall oxide film is dissolved. In (a), the time change of the oxide film thickness during the dissolution process of the inner wall oxide film can be described by the equation (1).

[0025]

[Equation 1] Here, l is the molecular volume of the oxide film thickness, Omega _SiO2 is SiO _{2, D 0} is oxygen diffusion constants, _{C 0} is the oxygen concentration, C
₀ (R) is the oxygen concentration at the COP interface. D ₀ is the formula (1
given by a)

[0026]

[Equation 2] (A) Injecting interstitial silicon into the COP or releasing vacancies Next, following the process of (a), proceed to the process of (a). The COP size 2R before heat treatment is given, and the time change of R is calculated at each position from the surface of the silicon substrate by the equation (2).

[0027]

[Equation 3] Here, Ω _Si is the atomic volume of silicon, C _I and C _V are the interstitial silicon and the concentration of vacancies during ingot growth, respectively.
C _I (R) and C _V (R) are the concentrations of interstitial silicon and vacancies at the COP interface, respectively. Further, D _V and D _I are diffusion constants of interstitial silicon and vacancies, respectively.
a) and equation (2b).

[0028]

[Equation 4] The COP free depth is the deepest depth at which R = 0 (zero) when calculated using such a calculation formula.

Next, in the step (2), a plurality of combinations of the heat treatment temperature, the holding time and the COP size capable of obtaining the desired COP free depth are selected from the calculation result of (1). From the calculation result of (1), the relationship between the heat treatment temperature T, the holding time t, the COP size 2R and the COP free depth becomes clear.

FIG. 3 is a diagram showing the relationship between the COP size 2R formed in the ingot, the heat treatment temperature T and the calculated COP free depth as an example of the calculation result of (1).
In addition, in order to simplify the calculation, the retention time was all set to 1 hour in (1). The figure shows the COP free depth on the horizontal axis and the COP size on the vertical axis.The isotherm drawn in the figure shows how many times a COP having an arbitrary COP size can be heat treated. You can see if you can get the COP free depth.

From the figure, the smaller the COP size or the higher the heat treatment temperature, the greater the COP free depth.
For example, when trying to obtain a COP-free wafer with a depth of 2 μm, the COP size is 1250 ° C. when the COP size is 300 nm, the 1200 ° C. when the COP size is 155 nm, and 11 when the COP size is 35 nm.
Heat treatment may be performed at 00 ° C. As described above, it is understood that a combination of a plurality of heat treatment temperatures, holding times and COP sizes can be considered when determining the manufacturing conditions of a wafer having an arbitrary COP free depth.

In the table of FIG. 2 (2), the heat treatment temperatures that can be selected are shown.
Virtual combination of degree, retention time and COP size
Manufacturing conditions are shown as. For example, using FIG.
When trying to obtain a COP-free wafer with a depth of 2 μm,
(T in the table of 2 (2)₁, T ₁, R₁), (T_Two, T
_Two, R_Two) And (T_Three, T_Three, R_Three) Are each (12
50, 1, 300), (1200, 1, 155) and (1100, 1, 3)
5)

In the following, using the table of FIG.
Which of the manufacturing conditions should be adopted?
Explain. The heat treatment temperature T in the table is T₁> T_Two
> T _ThreeSuppose that

Then, in the step (3), for the COP size 2R of each combination selected in (2), the growth rate f _p and the temperature gradient G _c for obtaining the COP size are calculated.
This corresponds to simulating the ingot manufacturing process shown in FIG. For the calculation of the growth rate f _p and the temperature gradient G _c , Voronkov's simulation (V.
V. Voronkov, J. Crystal Growth, 59, 625 (1982)) may be used.

According to Voronkov's simulation, the concentration C _I of interstitial silicon and the concentration C of vacancies during ingot growth
The time change of _V can be described by equations (3a) and (3b).

[0036]

[Equation 5] Here, the first term on the right side of equations (3a) and (3b) is
Growth rate f_pDrift due to slope diffusion of point defects, item 2
Is the concentration gradient diffusion of point defects, and the third term is the recombination reaction of point defects.
Corresponding to the calculation of k_FIs the reaction constant, C_I ^*, C_V ^*Haso
Represents the thermal equilibrium concentration of interstitial silicon and vacancies, respectively.
It In addition, D_I, D_V,C_I ^*And C_V ^*Is of temperature T
Is a function, and T is the crystal axis direction according to the comprehensive heat transfer analysis model.
Temperature gradient G_C= ΔT / ΔZ.

The COP formation and growth behavior during the growth of an ingot are represented by a COP size 2R and a size distribution function f (R, t) having time t as a function, and the time change of f (R, t) is represented. It can be obtained by solving the Fokker-Planck equations of the equations (4a) and (4b).

[0038]

[Equation 6] Here, A (R, t) and B in the formulas (4a) and (4b)
(R, t) satisfies the following relationship.

[0039]

[Equation 7] Here, in the equation (4c), k is the Boltzmann constant, and Δ
G = ΔG (R, C _I , C _V ) represents the amount of change in Gibbs free energy due to the formation of COP with radius R of COP, and this amount of change is obtained by equations (3a) and (3b). It can be calculated using the calculated values of the interstitial silicon and vacancy concentrations C _I and C _V. Then, expressions (3a) to (3b) and expressions (4a) to (4)
By sequentially solving c), f after the ingot is finally grown
(R, t) can be obtained. In the case where the nitrogen in the crystal is added, 0 D _V depending on the nitrogen concentration.
It should be calculated by multiplying by 5 to 0.1. (Masunori Akatsuka et al., 48th
28th Conference on Applied Physics, Proceedings 28p-s-5 (200
1) p.472).

By performing the above calculation, the growth rate f for obtaining the COP size for the COP size 2R
_{The p} and temperature gradient G _c can be calculated. Here, as shown in the table of FIG. 2 (3), COP size 2R is R ₁ , R ₂ , R
_In the case of ₃ , the growth rate f _p is f _p1 ,
f _{p2, f p2 is, G c1, G c 2,} G c2 respectively as the temperature gradient _{G c} is obtained.

Next, in the step (4), an appropriate growth rate is adopted as the growth rate at the time of pulling up the ingot, with respect to the ingot to be pulled up, from the plurality of growth rates calculated in (3).

Here, the "appropriate growth rate" is a rate determined depending on the diameter or quality of the ingot to be pulled up, and a rate determined based on the usage conditions of the manufacturing apparatus. From a production standpoint, it is sufficient to adopt the speed that allows the fastest growth with respect to the diameter of the ingot. The speed is, for example, 0.4 to 0.4 when the diameter of the ingot is 200 mm.
At 4.0 mm / min and 300 mm, it is about 0.2 to 3.0 mm / min.

On the other hand, when it is desired to avoid the quality of the ingot, for example, defects and distortion of the ingot due to rapid cooling, the "appropriate growth rate" is a pulling rate that satisfies these qualities. Furthermore, since the ingot cooling capacity of the manufacturing equipment depends on the cooling specifications of the equipment used, even if the fastest growth rate for the diameter of the ingot can be adopted, it is determined by the cooling capacity of the manufacturing equipment. There is also.

Here, in consideration of the diameter of the ingot, f _p1 is an appropriate growth rate, and in consideration of the quality of the ingot, f _p2 is an appropriate growth rate. It is used for the growing speed of time.

Finally, in the step (5), the temperature gradient corresponding to the growth rate adopted in (4), the heat treatment temperature and the holding time are the temperature gradient when the ingot is pulled up, and the heat treatment of the heat treatment applied to the wafer from which the ingot is cut out. Used for temperature and holding time.

In the step (4), the diameter of the ingot is taken into consideration.
If it is, the growth rate f of the manufacturing conditions _p1Ingot pull
Adopted for raising speed at the time of lifting. In response to this,
The temperature gradient, the heat treatment temperature and the holding time are determined by the CZ method.
Temperature gradient when pulling the ingot at
Heat treatment temperature of the heat treatment applied to the wafer from which the ingot was cut out
And the holding time at the heat treatment temperature.
On the other hand, in the process of (4), if the quality of the ingot is taken into consideration.
, Fostering speed f_p2It was adopted. In this case, the growth rate f
_p2There are two manufacturing conditions (manufacturing conditions)
Growth rate f _p2Temperature gradient, heat treatment temperature
And holding time cannot be determined. In this case
In consideration of other conditions,
You can select one from. For example, the lowest heat treatment temperature
Temperature gradient of manufacturing conditions (manufacturing conditions), heat treatment temperature
And holding time may be adopted.

Based on the above steps (1) to (5), the required calculation time when the growth rate, temperature gradient, heat treatment temperature and holding time are determined is about 15 minutes per condition.
Considering that it takes two weeks or more when the manufacturing conditions are determined by the actual machine test, it is much shorter time.

In fact, even if the growth rate, temperature gradient, heat treatment temperature and holding time determined in the present invention are introduced into an actual machine, there is no problem in wafer production and the desired CO
A wafer with P-free depth could be obtained.

[0049]

According to the present invention, in order to obtain a wafer having a desired COP-free depth, the growth rate and temperature gradient at the time of pulling up an ingot, the heat treatment temperature of heat treatment to be applied to the wafer from which the ingot is cut, and the holding time thereof are The wafer manufacturing conditions can be uniquely determined.

Since all of these manufacturing conditions can be determined by simulation, there is no cost and time loss required for the actual machine test, and it is possible to quickly respond to changes in manufacturing conditions accompanying product switching.

[Brief description of drawings]

FIG. 1 is a diagram schematically showing a series of steps from ingot formation by the CZ method to heat treatment of a wafer.

FIG. 2 is a diagram showing a flow of steps for determining wafer manufacturing conditions in the present invention.

FIG. 3 is a diagram showing a relationship between a COP size 2R formed in an ingot, a heat treatment temperature T, and a calculated COP free depth.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Masahiko Okui             4-53 Kitahama, Chuo-ku, Osaka City, Osaka Prefecture             Sumitomo Metal Industries, Ltd. F term (reference) 4G077 AA02 BA04 CF10 EH07 EH08                       PF51

Claims (2)

[Claims] 1. An ingot manufacturing condition data group for obtaining a COP size formed in an ingot calculated based on a growth rate and a temperature gradient in the ingot manufacturing process by the Czochralski method, and a wafer heat treatment in a wafer heat treatment process. Temperature and its holding time, and based on the heat treatment condition data group for calculating the COP free depth formed on the wafer surface calculated based on the COP size formed on the wafer, the heating temperature and holding time in the heat treatment step, And a method for determining a manufacturing condition of a wafer having a desired COP-free depth, which comprises determining a growth rate and a temperature gradient in an ingot manufacturing process. 2. The growth rate and temperature gradient at the time of pulling up the ingot by the Czochralski method, and the heat treatment temperature and the holding time of the heat treatment to be applied to the ingot cut wafer are determined by the following steps (1) to (5). , A method for determining a manufacturing condition of a wafer having a desired COP-free depth. (1) Heat treatment temperature and holding time of the wafer, and the step of obtaining the COP free depth calculated with the COP size formed in the ingot as a parameter (2) From the calculation result of (1) above, the desired COP free depth Step of selecting multiple combinations of heat treatment temperature, holding time and COP size that can obtain (3) For the COP size of each combination selected in (2) above, the growth rate and temperature gradient to obtain that COP size From the plurality of growth rates calculated in step (4) above (3), for the ingot to be produced, a step (5) above (4) of adopting an appropriate growth rate for the growth rate during pulling up of the ingot The temperature gradient, heat treatment temperature, and holding time that correspond to the growth rate adopted in 1. are set to the temperature gradient when pulling the ingot, the heat treatment temperature of heat treatment applied to the ingot-cut wafer, and the holding time. The step of use