JPH0630027B2 - Method and device for controlling temperature in heat treatment apparatus for semiconductor substrate - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a temperature control method for setting a temperature of a wafer to a predetermined target temperature in a heat treatment apparatus for heating a semiconductor substrate (hereinafter referred to as “wafer”). And a temperature control device.

(Prior Art and Problems Thereof) In a semiconductor device manufacturing process, heat treatment of a wafer is performed for various purposes such as uniform composition of an ion implantation layer, stabilization of a silicon film, and formation of ohmic contact. In such a heat treatment, it is necessary to rapidly and uniformly heat the wafer along a preset target temperature change curve and maintain the temperature accurately after the wafer temperature reaches the target heat treatment temperature. There is.

Therefore, in the heat treatment of wafers and the like, a heating furnace equipped with a radiant heat source such as a powerful halogen lamp is used, and the temperature inside the heating furnace is detected to obtain the deviation between this detected temperature and the target temperature value. Accordingly, feedback control is performed to correct the amount of power supplied to the radiant heat source.

FIG. 7 is a block diagram illustrating a heat treatment apparatus adopting such a conventional temperature control method. In the figure, a wafer 1 to be heat-treated is loaded into a heating furnace 3 having a plurality of radiant heat sources 2 arranged on both sides, and a wafer supporting mechanism 4 is provided.
Supported by.

On the other hand, in the target temperature setting device 10, a trapezoidal target temperature change curve F ₁ (T _O ) as shown in FIG. 8 is set as a function of time t. However, in FIG. 8, the temperature is an absolute temperature [K] and T _P is a steady processing temperature.
The target temperature signal T _O output from the target temperature setter 10 in FIG. 7 according to the target temperature change curve F ₁ (T _O ) is the temperature detector 5 (for example, thermocouple) in the heating furnace 3.
It is given to the subtractor 11 together with the actually measured temperature signal _Tr in the heating furnace 3 detected by.

The subtracter 11 obtains the deviation between the above-mentioned T _O and _Tr : ΔT = T _O −T _r (1) and outputs this deviation signal ΔT to the multipliers 13 to 15 in the PID control system 12. To do. Of these, the multiplier 13 multiplies the deviation ΔT by a constant K ₁ to give a deviation proportional signal S _P. Further, the multiplier 14 multiplies the deviation ΔT by a constant K _2, and the integrator 16 at the next stage performs time integration to give a deviation integration signal S _I. Furthermore, the multiplier 15 multiplies the deviation ΔT by a constant K _3, and the differentiator 17 at the next stage differentiates this with respect to time to give a deviation differentiation signal S _D. And
These signals S _P , S _I and S _D are added by the adder 18 and become the temperature control signal S _T.

The temperature control signal S _T is supplied to the power supply unit 19 for supplying the power P to the radiant heat source 2. In the power supply unit 19, the power controller 21 to input the temperature control signal S _T, the power P supplied from the power source 20 to the radiant heat source 2 is controlled to be proportional to the temperature control signal S _T. Therefore, in such a conventional device, the temperature control signal S _T itself functions as a power control signal.

In this way, the radiant energy from the radiant heat source 2 increases or decreases in accordance with the deviation ΔT, and the wafer 1 has the target temperature change curve F
₁ (T 2 _O ) will be subjected to temperature control.

However, in such a conventional temperature control, there is a problem that the temperature in the heating furnace cannot sufficiently follow the rapid rise of the target temperature. To see this problem more specifically, refer to FIG. This FIG. 9 is an enlarged view of the first half of FIG. 8, showing time-varying curves F ₂ (T _r ) and F of the measured temperature T _r and the supplied power P in the heating furnace 3.
It is the figure which drew ₃ (P) on it. As can be seen from this figure, when the target temperature rises rapidly (for example, when the temperature inside the heating furnace 3 is rapidly raised from 200 ° C. to 300 ° C. to 1000 ° C. or more in a few seconds), heating is performed. since the actual temperature T _r of the furnace 3 requires considerable time until the same becomes stable and steady process temperature T _P, it has become difficult to accurately achieve the desired heat treatment.

On the other hand, if the constants (K ₁ , K ₂ , K ₃ ) corresponding to the feedback gain are increased for the purpose of improving this,
Overshooting or hunting occurs near the time point t ₁ when the process goes from the heating process to the steady process temperature maintaining process, and thus such a method cannot sufficiently solve the above problem.

(Object of the Invention) The present invention is intended to overcome the above-mentioned problems in the prior art, and makes the temperature of the semiconductor substrate respond to the change of the target temperature value with high accuracy and high speed in the heat treatment apparatus for heating the semiconductor substrate. An object of the present invention is to provide a temperature control method and a temperature control device capable of performing the above.

(Means for Achieving the Purpose) In order to achieve the above-mentioned goal, the present invention generates radiant energy by the electric power supplied from the electric power supply means,
A temperature control method for setting the temperature of the semiconductor substrate to a predetermined target temperature in a heat treatment apparatus for heating a semiconductor substrate with the radiant energy, based on a relationship between the radiant energy of the radiant heating means and the temperature of the semiconductor substrate. , A conversion rule corresponding to the correspondence relationship between the power supplied to the radiant heating means and the temperature of the semiconductor substrate is obtained in advance, and a feedforward control system based on the target temperature is provided and obtained by the feedforward control system. The temperature control signal is converted into a power control signal according to the conversion rule, and the power supply amount from the power supply means to the radiant heating means is controlled according to the power control signal.

That is, conventionally, the temperature θ of the semiconductor substrate and the supplied power P
The temperature deviation ΔT without examining the relationship between
The temperature control signal obtained by applying the PID process to the power supply means is directly input to the power supply means as the power control signal, and the power supply means supplies the power proportional to this input to the radiant heat source (radiant heating means). However, since the thermal equilibrium temperature of the semiconductor substrate is generally not proportional to the supplied power, the conventional problem is that the semiconductor substrate (wafer) temperature is not proportional to the temperature control signal, and the above-mentioned problem occurs. To do.

In the present invention, the relationship between the supplied power P and the temperature of the semiconductor substrate θ is based on the relationship between the radiant energy I of the radiant heat source and the temperature θ of the semiconductor substrate in the thermal equilibrium state or the thermal non-equilibrium state: P = f (θ) (2) is obtained in advance as a functional expression or numerical data, and the temperature control signal is converted into a power control signal based on this function to perform temperature control. However, it is not essential to know the temperature θ of the semiconductor substrate directly, and in the case of using a closed heating furnace or the like, the temperature θ may be indirectly or approximately obtained from the temperature inside the furnace.

Further, in the present invention, the temperature control is realized by using feedforward control. This is because the response characteristics of the controlled object can be understood relatively well by knowing the relationship of Eq. (2) in advance, so feedforward control with less error due to transmission delay is suitable. However, the control characteristics are further improved by performing the feedback control together. Such an example will be described in the examples below.

(Example) A. Principle of Embodiment Before describing the specific configuration and operation of the embodiment of the present invention, the control principle thereof will be described by taking the above-described wafer heat treatment as an example. To do this, it is necessary to know Eq. (2) in a concrete form, so first find this formula.

FIG. 2 (A) is an internal schematic diagram of the heating furnace 3 described in FIG. 7. In FIG. 2 (A), the radiant heat source 2
The irradiation intensity per unit area of the wafer 1 to the wafer 1 is set to I, and the temperature of the wafer 1 is set to θ [K]. Also,
As shown in FIG. 2 (B), various quantities of the wafer 1 are
It is defined as follows.

S ... Surface area of one surface d ... Thickness c ... Specific heat ρ ... Density ε ... Emissivity Further, it is assumed that the radiant energy applied to the wafer 1 is all absorbed by the wafer 1. Then, the minute time Δt
The amount of heat given to the wafer 1 from the radiant heat source 2 is (εSIΔt) on both front and back sides, while the wafer 1
The amount of heat lost by radiation from the
The radiant energy per unit area of the black body of [K] is (2εSUΔt).

Therefore, the amount of heat corresponding to the subtraction: ΔQ = εS (I−2U) Δt (3) is stored in the wafer 1.

If the temperature of the wafer 1 is increased by Δθ due to the heat quantity ΔQ, the heat capacity of the wafer 1 is (ρcdS). ) Holds.

Therefore, in the limit of Δt → 0, Is established.

However, the radiant energy U of the black body is proportional to the fourth power of the temperature θ according to the well-known Stefan-Boltzmann law.

U = σθ ⁴ (6) However, σ is a constant, and using Boltzmann's constant K _B , high-speed c _l and Planck's constant h, σ = 2π ⁵ K _B ⁴ / and 15 c _l ² h ³ (7) ) Is given. Further, if the ratio of the radiant energy intensity I and the supplied power P in the radiant heat source 2, that is, the radiation efficiency is γ, then I = γP (8) holds.

Therefore, according to equations (5), (6), and (8), Is established. The expression (9) is an expression embodying the expression (2) in this embodiment.

According to this equation (9), the temperature θ of the wafer 1 is set to the target temperature T _O
In order to match with, the electric power P that is obtained as θ = T _{O on} the right side of the equation (9) must be given to the radiant heat source 2. Therefore, the control circuit is also constructed according to the equation (9).

Equation (9) is an equation obtained under some approximations, and in practice, for example, There may be an additional term g such as Such a term g is, for example, a radiant heat source filament (resistor).
It occurs due to the heat capacity of itself and the dissipation of heat to the outside of the heating furnace, but the handling of these will be described later.

B. First Embodiment FIG. 1A is a block diagram of a wafer heat treatment apparatus using a temperature control apparatus according to a first embodiment of the present invention. Then, in this embodiment, the feedforward control system FF ₁
And a feedback control system FB ₁ in combination. In FIG. 1A, the target temperature changer 10 is preset with the target temperature change curve shown in FIG. 8 as in the conventional device of FIG. The target temperature signal T _O from the target temperature setter 10 is used for the heating furnace 3
Together with the measured temperature signal T _r from the temperature detector 5 provided on the surface of the wafer 1 in the inner input to the subtracter 31, the deviation signal for where the feedback control _{△ T: △ T = T O} -T r ... It becomes (11). This part is similar to the conventional device.

The deviation signal ΔT obtained in this way is output from the multiplier 32 ...
It is input to 34 and multiplied by K _P , K _I and K _d , respectively. Of these, the multiplication output of the multiplier 32 is the PID
It corresponds to the proportional control part in control. Also, the multiplier 3
The multiplication output at 3 is integrated by the integrator 35 and becomes an integration control output. Then, these are added in the adder 36.

On the other hand, the multiplication of K _{d in} the multiplier 34 corresponds to the multiplication of the gain as the previous step of obtaining the differential control output in the related art, but the output of the multiplier 34 is added in the adder 36 as in the conventional device. Feed-forward control system FF ₁
Is given to the side. As can be seen from the equation (9), this is because the time derivative of temperature (dθ / dt) is included in the equation expressing the supply power P as a term different from the term of θ ^4. This is because it is desirable to handle to separate only the corresponding _(dT O / dt).

Further, the target temperature signal T _O from the target temperature setting device 10 is
Two multipliers 3 in the feedforward control system FF ₁
Also given to 7,38. Of these, in the multiplier 38,
The target temperature signal T _O is multiplied by the gain J _P. The signal thus obtained is given to the adder 39, and the adder 39 adds the feedback output S _F from the feedback control system FB ₁ . Therefore,
The output S ₁ of the adder 39 is a signal obtained by correcting the target temperature signal T _O, which is the basis of the feedforward control, with the feedback signal S _F according to the temperature deviation ΔT.

This signal S ₁ is added with a predetermined constant α given from the constant setter 41, and then is squared by the quadratic circuit 40. Then, the constant β from the other constant setter 42 is subtracted to form the signal S ₂ , and the signal S ₂ is input to the adder 44. The meanings of these constants α and β will be described later.

On the other hand, in the multiplier 37, the target temperature signal T _O is multiplied by J _d . Then, after the signal (K _d ΔT) from the multiplier 34 in the feedback control system FB ₁ is added, the signal is differentiated by the differentiator 43. Then, the differential signal S ₃ is added to the output S ₂ from the quadratic multiplier 40 by the adder 44 to become a signal S ₄ , which is given to the power supply unit 60 via a circuit described later.

By the way, the step of the fourth power calculation corresponds to the main feature of the present invention, whereby the signal S ₁ as a part of the "temperature control signal" obtained based on the target temperature value T _O is directly used. Instead of using the “power control signal”, it is converted into a more appropriate “power control signal” S ₂ in consideration of the correspondence relationship between the power supplied to the radiant heat source and the semiconductor substrate (wafer) temperature. .

Thus, the signal S ₂ obtained by the fourth power of the corrected value according to the target temperature T _O the deviation △ T, the signal S ₃ obtained by differentiating the time after applying the same correction to the target temperature T _O A circuit for supplying the electric power P represented by the equation (9) is realized by adding the signal S ₄ and giving it to the electric power supply unit 60 side. Thus, once by that it converted to an appropriate power control signal, the semiconductor substrate (wafer) temperature will be proportional to the target temperature signal T _O.

Here, the meanings of the above-mentioned constants α and β will be described. The target temperature T _O output from the temperature setter 10 is given as an absolute temperature, and an arbitrary temperature within the range of T _O ≧ 0 [K] can be set. Further, the quadratic process in the quadrupler 40 is based on the Stefan-Boltzmann law as described above, and this law is also valid from 0 [K]. Therefore, the above-mentioned fourth power processing is suitable for the case where the heating furnace 3 is initially heated from 0 [K]. Therefore, when the fourth power process is performed on the target temperature T _O and the signal is directly given to the power supply unit 60 side,
Electric power is supplied to the radiant heating source 2 under a situation where the inside of the heating furnace 3 is initially assumed to be 0 [K].

However, in a normal heat treatment of a wafer, the temperature inside the furnace is maintained at about 100 ° C. to 400 ° C. (hereinafter, referred to as T _n ) from the beginning. Therefore, the heating for increasing the temperature between 0 and T _n [K] is essentially unnecessary, and it is substantially necessary to supply the electric power for raising the temperature to T _n or higher. Is.

Therefore, in this embodiment, not only is the quadratic processing corrected the relationship between the target temperature value T _O and the supplied power P in accordance with Stefan-Boltzmann's law, but the origin of the quadratic processing in the quadrupler 40 is determined. By moving it, the above situation is addressed. That is, as described above, the adder 39
Is added to the output of the constant setter 41 to perform a quadratic process, and then the other constant setters 4
The constant β which is the output from 2 is subtracted and output.

Then, if there is no adder for these constants α and β,
There is a relation of S ₂ = (S ₁ ) ⁴ (12) between the inputs and outputs S ₁ and S ₂ of these circuit parts. However, when the addition and subtraction of α and β are performed, S ₂ = (S ₁ + Α) ⁴ −β (13)

Therefore, the origin of the quadratic processing is moved to (α, β) by the addition and subtraction, and these constants α,
By appropriately setting β, it is possible to correct that the temperature at the starting point of the heat treatment is not 0 [K].

Specifically, for example, the value of room temperature [K] is used as α, and the value of the power control signal S ₂ required to give that α is used as β. In this way, a control circuit adapted to the actual temperature control range can be realized.

The power control signal S ₄ thus obtained is stored in the data memory 46 and is also supplied to the radiant heat source correction circuit 50 via the switch 45. This radiant heat source correction circuit 50 is used for the radiant heat source 2 when the target temperature T _O changes.
This is a circuit for correcting the electric power P given to the radiant heat source 2 in consideration of the electric power required to heat the internal radiator (resistor) itself, the details of which are shown in FIG. 1B. A plurality of internal circuits of the radiant heat source correction circuit 50 and the subsequent circuits are provided corresponding to each of the plurality of unit radiant heat sources 2a, ..., 2n constituting the radiant heat source 2. This is because it is considered that the unit radiant heat source that actually faces the wafer 1 and another unit radiant heat source inside the heating furnace 3 have different thermal effects on the wafer 1.

Therefore, of the n unit correction circuits 50a, ... Included in the radiant heat source correction circuit 50, the Kth unit correction circuit 50K is included.
Will be described. Assuming that the heat capacity of the resistor of the unit radiant heat source 2K corresponding to this unit correction circuit 50K is C _d and its temperature is θ, the electric power required to give a temperature change to this resistor is Can be written.

Therefore, considering the above correction amount, the power P
Of the power P _K allocated to the unit radiant heat source 2K
Adding equation (14), If the electric power P ′ _K represented by the above is supplied to the unit radiant heat source 2K, the responsiveness can be further improved. This corresponds to an example of the additional term indicated by g in the equation (10). However, the constant L _K takes a different value for each unit radiant heat source, and the reason thereof is due to the facing relationship with the wafer 1 and the geometrical shape in the heating furnace 3 as described above.

By the way, as already mentioned, the energy radiated by an object at temperature θ is calculated by Stefan-Boltzmann's law ((6)
Expression). Therefore, if the radiant energy is proportional to the electric power P, the second term on the right side of the equation (15) can be obtained by solving the equation corresponding to the equation (6) for θ. Will be proportional to. Thus, if the coefficients determined so as to be experimentally optimum value _Z K, and _{Q d,} the unit correction circuit 50K, the _{input: S K = S 4 (Z} K / <Z>) ... ( 17) Should be calculated. However, in these equations, the coefficient Z _K is the power distribution coefficient (%). Also, <Z> is Z _a
Is an average value of Z _n and is a constant included for the purpose of normalization.

The unit correction circuit 50K of FIG. 1B is constructed according to such a principle. The input S ₄ is multiplied by Z _K by the multiplier 51K, and then 1 / <Z> is multiplied by the multiplier 52K. Then, the output S _K is given to the fourth root circuit 53K to obtain 4
Root And is differentiated by the differentiator 54K, and then the multiplier 55K
Is multiplied by the coefficient Q _d . This value is then multiplied by the multiplier 52
The power supply unit 6 of FIG. 1A is added with the output of K.
The power controller 61K in 0, is given as a power control signal _{A K.} Other unit correction circuits 50a, ..., 5 of FIG. 1B
The same applies to 0n.

In this way, the power controllers 61a-61n of FIG.
In the power control signal _A a is inputted, ..., in accordance with the An, the power _P a from the power source 62, ..., Pn (although, _{P a}
+ ... + P _n = P) is applied to the radiant heat sources 2a, ..., 2n to heat the wafer 1.

By the way, in this embodiment, since the contact type temperature detector 5 is provided on the wafer 1, it is not possible to perform the heat treatment of the product wafer while operating the control loop. Therefore, in this embodiment, a dummy wafer is used as the wafer 1, the temperature detector 5 as described above is provided, and desired heat treatment is performed on a trial basis. Thereby, optimum values are determined as various constants, and the power control signal value at that time is stored in the data memory 46.

After that, when heat-treating a product wafer not provided with the temperature detector 5, the switch 45 is switched to read the data of the power control signal value from the data memory 46, and this data is read by the power controllers 61a to 61n. To control the temperature.

However, when the non-contact temperature detector is used, the above problem does not occur, and the control loop can be operated even when the heat treatment of the product wafer is performed.

C. Second Embodiment FIG. 3 is a block diagram showing a second embodiment of the present invention. The first difference between this embodiment and the first embodiment is that
The target temperature J _p T _O, which has been multiplied by the constant J _{P in} the multiplier 38 in the feedforward control system FF ₂ , is used as the fourth power 4
That is, it is added to the PID feedback signal S _E from the feedback control system FB ₂ after the fourth power of 0. That is, when the feedforward control signal is generally x and the feedback signal is y, the first
In the second embodiment, the operation of the form (x + y) ⁴ is carried out, whereas in the second embodiment, the operation of the form (x ⁴ + y) is carried out. Of course, both give different control signals, but according to the present invention, x is a major factor rather than y because the response of the control system is enhanced.
The difference between (x + y) ⁴ and (x ⁴ + y) is relatively small. Therefore, the desired effect can be obtained also in the second embodiment.

Also, the differential control part (multiplier 3 of the PID control system
The output of 4) is not given to the feedforward control system FF ₂ side as in the first embodiment, but is differentiated by the differentiator 71 and then proportional / integral control is performed in the adder 72 in the feedback control system FB ₂ . It is adding to the output. Then, in the feedforward control system FF ₂ , the output of the multiplier 37 is directly differentiated by the differentiator 43. However,
All of these are finally added together.
As can be seen by comparing with FIG. A, there is no essential difference because only the order of addition and differentiation is exchanged.

D. Third Embodiment FIG. 4 shows a third embodiment. In this example, the PID
Differential control part of control system (multiplier 34 and differentiator 71)
The output of is added to the proportional / integral control output in the adder 72 of the feedback control system FB ₃ and the like, and the configuration is similar to that of the second embodiment. However, 4
The first embodiment in that the fourth power processing in the multiplier 40 is performed on the sum of the output of the multiplier 38 in the feedforward control system FF ₃ and the control output S _E of the feedback control system FB _3. It has a common configuration with.

Therefore, the third embodiment has an eclectic configuration of the first and second embodiments. The advantage of the second and third embodiments is that the conventional PID control unit can be used as it is.

E. Data Example FIGS. 5A and 5B are graphs showing measured values of temperature and electric power for the first and second examples, respectively, similar to FIG. 9 for the conventional device. As can be seen by comparing these figures with FIG. 9, in each of the above-described examples, the semiconductor substrate temperature curve F ₂ (T _r ) is more accurate than the target temperature change curve F ₁ (T _O ). It can be seen that it follows with a high-speed response. In particular, in FIG. 5A (first embodiment), the change in the target temperature is followed with almost no error. There is almost no hunting at the time t ₁ when the steady processing temperature T _p is reached or fluctuation of the temperature after t ₁ .

FIGS. 6A and 6B respectively show the _first temperature when the target temperature change curve F ₁ (T _O ) is a two-step curve.
3 is a similar actual measurement graph for the second example. The accuracy of control in the present invention can be understood from these figures as well.

F. Modifications The embodiments of the present invention have been described above, but the present invention is not limited to the above embodiments, and the following modifications are possible.

Although the Stefan-Boltzmann law is used as it is in the above embodiment to find the relationship between the supplied power and the semiconductor substrate temperature, this relationship can be variously modified depending on the situation. One of the reasons is that this Stefan-Boltzmann law is a law expressing the relation between temperature and total radiant energy. In other words, depending on the absorption spectrum distribution of the semiconductor substrate, absorption or radiation only in a specific wavelength band may be a problem.
If the conversion rule is modified in consideration of Planck's radiation law for obtaining radiant energy when focusing only on a specific wavelength, higher control performance can be obtained.

Other deformation factors include, for example, the influence of the atmosphere in the heating furnace and the dissipation of heat to the outside of the furnace. This invention not only controls the temperature of the semiconductor substrate in a closed heating furnace,
The present invention can also be applied to the case of heating a semiconductor substrate placed in a place open to the outside, but in this case, modification considering these factors is also possible.

In the above embodiment, the input values A _{a to} A _n of the power controllers 61 a to 61 _n and the output powers P _{a to} P _n are in a proportional relationship (linear characteristic). However,
As the configuration of the power supply unit, the relationship between the power control value A and the supplied power P may be such that the power P is the power control value A to the Mth power (M is an integer). Therefore, in this case, the Nth power is used instead of the fourth power. However, N is an integer that satisfies N × M = 4.

Further, when performing the fourth power processing, a look-up table type memory in which conversion data is given in advance may be used instead of the fourth power (arithmetic unit). Further, depending on the object to be controlled, it is possible to make modifications such as using a power process other than the fourth power or a polynomial conversion process.

In the above embodiment, a solid substance such as a wafer is simply heated. However, in the case of a process of heating and melting a solid, the presence of heat of fusion and the like are taken into consideration, and the power supplied at the melting point is considered. It is also possible to make modifications such as increasing.

The present invention can be applied not only to the radiant heat source of electromagnetic waves in the light region such as the halogen lamp, but also to the case where a semiconductor substrate is heated by microwaves, for example. In this case, the relationship between the power supplied to the microwave oscillator and the temperature of the semiconductor substrate is obtained in advance, and the control system is configured based on this.

In the above embodiment, the feedback control is also performed to improve the accuracy, but the feedforward control is essential in the present invention, and the feedback control does not necessarily have to be used. Further, in the above embodiment, the values of the coefficients other than the feedforward proportional coefficient J _P , that is, the values of J _d , K _P , K _I , and K _d can be set to zero if necessary.

INDUSTRIAL APPLICABILITY The present invention can be used not only for heat treatment of a semiconductor substrate, but also for maintaining the temperature of a semiconductor substrate used for another treatment (for example, chemical treatment) (for example, temperature control in a constant temperature container).

It goes without saying that the circuit of the embodiment may be realized by a high speed computer.

(Effects of the Invention) As described above, according to the present invention, based on the relationship between the radiant energy of the radiant heating means and the temperature of the semiconductor substrate, the power supplied to the radiant heating means and the temperature of the semiconductor substrate are The correspondence relationship is obtained in advance, and the temperature control signal obtained by the feedforward control system is converted into a power control signal according to the conversion rule according to the correspondence relationship to control the power supply amount to the radiant heating means. Therefore, it is possible to obtain a temperature control method and apparatus capable of making the temperature of the semiconductor substrate respond to the change of the target temperature with high accuracy and high speed.

[Brief description of drawings]

1A and 1B are block diagrams of a heat treatment apparatus for wafers using the first embodiment of the present invention, FIG. 2 is a schematic diagram of a heating furnace for explaining the principle of the present invention, FIG. 3 and FIG. FIG. 4 is a block diagram of a heat treatment apparatus for wafers using the second and third embodiments of the present invention, and FIGS. 5A, 5B, 6A and 6B show measured data in the embodiment of the present invention. FIG. 7, FIG. 7 is a block diagram showing an example of a conventional wafer heat treatment apparatus, FIG. 8 is a view showing an example of a target temperature change curve, and FIG. 9 is a graph showing characteristics of the apparatus of FIG. DESCRIPTION OF SYMBOLS 1 ... Wafer (semiconductor substrate), 2 ... Radiant heat source, 3 ... Heating furnace, 5 ... Temperature detector, 10 ... Target temperature setter, 40 ... Quadrupler, 60 ... Power supply unit, T _O ... Target temperature, T _r ... Semiconductor substrate temperature, P ... Supply power

Claims (10)

[Claims] 1. A temperature control method for setting a temperature of a semiconductor substrate to a predetermined target temperature in a heat treatment apparatus for generating radiant energy by electric power supplied from a power supply means and heating the semiconductor substrate by the radiant energy. Then, based on the relationship between the radiant energy of the radiant heating means and the temperature of the semiconductor substrate, a conversion rule according to the correspondence relationship between the power supplied to the radiant heating means and the temperature of the semiconductor substrate is obtained in advance. A feedforward control system based on the target temperature is provided, and a temperature control signal obtained by the feedforward control system is converted into a power control signal according to the conversion rule, and the power supply means is provided according to the power control signal. Of heat from a semiconductor substrate by controlling the amount of power supplied to the radiant heating means from Temperature control method in the apparatus. 2. The temperature control method in a heat treatment apparatus for a semiconductor substrate according to claim 1, wherein the conversion rule is determined based on Stefan-Boltzmann's law regarding the relationship between temperature and radiant energy. 3. The power supply means has a linear characteristic that gives a power supply amount proportional to the power control signal to the radiant heating means, and the conversion rule is a rule for converting the temperature control signal to the fourth power. 5. A temperature control method in a semiconductor substrate heat treatment apparatus according to claim 2. 4. The power conversion signal according to claim 4, wherein the fourth power conversion is performed after adding a first constant to the temperature control signal, and the second power constant is subtracted from the signal after the fourth power conversion. A method for controlling temperature in a semiconductor substrate heat treatment apparatus as set forth in claim 3. 5. A time derivative of a signal according to a target temperature is calculated,
Create a power control signal by adding the signal after the fourth power conversion,
A temperature control method for a heat treatment apparatus for a semiconductor substrate according to claim 3 or 4. 6. A feedback control system relating to the temperature of the semiconductor substrate is provided together with a feedforward control system,
The temperature control method in a heat treatment apparatus for a semiconductor substrate according to claim 1, wherein the conversion based on the conversion rule is performed on the sum of the outputs of the control of the feedback control system and the control of the feedforward control system. 7. A temperature control device for setting a temperature of a semiconductor substrate to a predetermined target temperature in a heat treatment device for generating radiant energy by electric power supplied from a power supply means and heating the semiconductor substrate by the radiant energy. A feedforward control system based on the target temperature is provided, and a temperature control signal obtained by the feedforward control system is converted according to the correspondence relationship between the power supplied to the radiant heating means and the temperature of the semiconductor substrate. A temperature in a heat treatment apparatus for a semiconductor substrate, characterized in that conversion means for converting into a power control signal is provided according to a rule, and the power supply amount from the power supply means to the radiant heating means is controlled according to the power control signal. Control device. 8. The semiconductor substrate according to claim 7, wherein the conversion means is a means for performing conversion according to a conversion rule determined based on Stefan-Boltzmann's law regarding the relationship between temperature and radiant energy. Control device in the heat treatment equipment of. 9. The power supply means is a means having a linear characteristic that gives a power supply amount proportional to the power control signal to the radiant heating means, and the conversion means converts the temperature control signal to the fourth power.
The temperature control device in the heat treatment device for a semiconductor substrate according to claim 8, which is a power conversion means. 10. A feedback control system relating to the temperature of the semiconductor substrate is provided together with a feedforward control system, and the conversion in the conversion means is performed on the sum of the respective outputs of the feedback control system and the feedforward control system. A temperature control device in a semiconductor substrate heat treatment apparatus according to claim 7.