JP2015092580A - Temperature controller for controlling sample temperature, sample stand for mounting sample, and plasma processing apparatus comprising them - Google Patents

Abstract

PROBLEM TO BE SOLVED: To stably perform control for making a temperature of a sample be desired, by accurately predicting a sample temperature.SOLUTION: A temperature controller for sample comprises: a sample stand 10 which includes a mounting surface for a sample and in which a flow channel for a heat transfer medium is formed; a plurality of heaters 13 provided inside the sample stand so as to be distributed to a plurality of areas in a sample stand surface; and a plurality of temperature sensors 15 which correspond to the plurality of areas and measure temperature of the sample stand. The temperature controller for sample: makes respective approximate expressions of correlations among temporal change of respective temperatures of the plurality of temperature sensors and temporal change of respective temperatures of the sample corresponding to the plurality of areas, temperatures obtained by subtracting a temperature of the heating medium from temperatures of the respective temperature sensors and sample, and supply power to the heater, in the case of measurement in a state where plasma processing is not performed; predicts, by using the respective approximate expressions, respective temperatures of the sample corresponding to the plurality of areas during plasma processing from respective temperatures of the plurality of temperature sensors during the plasma processing and respective supply power to the plurality of heaters; and controls the temperature of the sample 11 during the plasma processing by using the respective predicted temperatures of the sample 11.

Description

  The present invention relates to a temperature control device that controls the temperature of a sample during plasma processing, a sample stage on which a sample is placed, and a plasma processing apparatus including these.

  FIG. 18 is a diagram for explaining a conventional method for adjusting the sample temperature. As shown in FIG. 18, the sample is placed on a sample table 10 cooled by a refrigerant. A heater 13 is embedded in the sample stage, and the heating amount of the sample stage can be adjusted by adjusting the power supplied to the heater 13. Further, a temperature sensor 15 is embedded in the sample table, and the temperature of the sample table 10 can be measured (see Patent Document 1).

  FIG. 19 is a diagram for explaining a conventional temperature control apparatus. As shown in the figure, a method is used in which the electric power u1 supplied to the heater is feedback-controlled so that the measured value y1 of the temperature sensor becomes the target value r. Further, PI control is generally used as a feedback control method. In recent years, a structure in which the sample stage is divided into three regions of a center, an edge, and a middle and each has a heater and a temperature sensor has been put into practical use.

JP-T-2005-522051

  In the conventional method, the temperature of the sample stage can be controlled at high speed. However, (1) the heat conduction between the sample stage and the sample placed on the sample stage is not sufficient. For this reason, it takes a long time for the wafer temperature to reach a desired value. (2) As described above, when the sample stage is divided into several regions and the wafer temperature is controlled for each region, a gradient is given to the wafer temperature by giving a difference in the set value of the sensor temperature. In this case, the temperature gradient of the wafer may be reduced due to the heat conduction of the wafer itself. In such a case, the wafer temperature distribution and the sample stage temperature distribution do not match. Further, (3) during the actual etching process, the wafer is heated by heat input from the plasma. For this reason, the wafer temperature becomes higher than the set temperature.

  The present invention has been made in view of these problems, and provides a temperature control device capable of accurately predicting a sample temperature and stably controlling it to a desired temperature, a sample stage, and a plasma processing apparatus including these. To do.

In order to solve the above problems, the present invention mainly employs the following configuration.
A sample table having a mounting surface capable of adsorbing and holding a sample on the upper surface and having a heat transfer medium flow path formed therein, and a plurality of heaters provided in the sample table divided into a plurality of regions in the sample table surface And a plurality of temperature sensors for measuring the temperature of the sample stage corresponding to the plurality of regions,
Time variation of each temperature of the plurality of temperature sensors measured without plasma treatment, time variation of each temperature of the sample corresponding to the plurality of regions, and each temperature obtained by subtracting the temperature of the heat transfer medium The correlation between the temperature of the sensor and the sample and the power supplied to the heater is set to an approximate expression, and the temperature of each of the plurality of temperature sensors and the plurality of heaters during the plasma processing are set using the approximate expressions. The temperature of each of the samples corresponding to the plurality of regions during the plasma processing is predicted from each of the supplied powers, and the in-plane temperature of the sample during the plasma processing is calculated using the predicted temperatures of the samples. A control device for controlling is provided.

  Since the present invention has the above configuration, the sample temperature can be accurately predicted and the sample can be stably controlled to a desired temperature.

It is a figure explaining the 1st wafer temperature prediction method. It is a figure explaining the 2nd wafer temperature prediction method. It is a figure explaining the 3rd wafer temperature prediction method. It is a figure explaining the method of calculating | requiring the plasma heat input of real process conditions. It is a figure explaining the temperature control apparatus used by this invention. It is a figure explaining a microwave plasma etching apparatus. It is a figure explaining the detail of a sample stand. It is a figure which shows the change of the predicted value of the wafer temperature by the 1st prediction method. It is a figure which shows the change of the actual value of wafer temperature by the 1st prediction method. It is a figure which shows the change of the sensor temperature by a 1st prediction method. It is a figure which shows the change of the actual value of wafer temperature by the 1st prediction method. It is a figure which shows the change of the predicted value of the haha temperature by a 2nd prediction method. It is a figure which shows the change of the actual value of wafer temperature by the 2nd prediction method. It is a figure which shows the change of the predicted value of the wafer temperature by the 3rd prediction method. It is a figure which shows the change of the actual value of wafer temperature by the 3rd prediction method. It is a figure which shows the change of the predicted value in the case of plasma heat input by the 2nd prediction method. It is a figure which shows the change of the measured value in the case of plasma heat input by the 2nd prediction method. It is a figure explaining the adjustment method of the conventional sample temperature. It is a figure explaining the conventional temperature control apparatus.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 5 is a diagram for explaining a temperature control device used in the present invention. As shown in FIG. 5, the wafer temperature x2 that changes from time to time is predicted from the time change of the sensor temperature y1 and the heater power u1, and this predicted value (~ x2) (the superscript to x is given for convenience (~ x)). Is fed back to the heater power u1, and the wafer temperature x2 is controlled. As the wafer temperature prediction method, the following three types of methods are used.

  FIG. 1 is a diagram for explaining a first wafer temperature prediction method. In this method, in step S1-1, in a state in which a pseudo wafer having a temperature measurement mechanism is mounted on the sample stage in advance, the power of each heater to be described later is changed while the plasma is turned off, so that the wafer temperature and sensor temperature are changed. Measure the time change of the heater power. In step S1-2, the correlation among wafer temperature, sensor temperature, and heater power is approximated by a linear differential equation. In step S1-3, the predicted value of the wafer temperature is calculated from the sensor temperature and the heater power during actual sample processing by isomorphic observation (Luenberger's states obsever) using the linear differential equation obtained in step S1-2.

  FIG. 2 is a diagram for explaining a second wafer temperature prediction method. First, in step S2-1, the heater power is changed by the same method as the first method, and the time change of the wafer temperature, the sensor temperature, and the heater power is measured. In step S2-2, the heater power is changed in the same manner as in step S2-1 without mounting the sample on the sample table, and the sample table surface temperature is measured. In step S2-3, the correlation among the wafer temperature, sensor temperature, heater power, and sample stage surface temperature is approximated by a linear differential equation. In step S2-4, the predicted values of the sample stage surface temperature and the wafer temperature are calculated from the sensor temperature and the heater power during actual sample processing by isomorphic observation using the linear differential equation obtained in step S2-3.

  FIG. 3 is a diagram for explaining a third wafer temperature prediction method. About step S3-1 to 4, the predicted value of the sample stage surface temperature and the wafer temperature is calculated by the same method as the second method. These predicted values are defined as a sample stage surface temperature predicted value 1 and a wafer temperature predicted value 1, respectively. In step S3-5, with the pseudo wafer mounted on the sample stage, the plasma discharge is turned on to change the plasma heat input condition and the heater power, and the wafer temperature, sensor temperature, and heater power over time are measured. In step S3-6, the correlation between the wafer temperature, the sensor temperature, the heater power, the plasma heat input, the sample table surface temperature predicted value 1 and the wafer temperature predicted value 1 is approximated by a linear differential equation.

In step S3-7, plasma heat input under actual process conditions is obtained by a method described later. In step S3-8, the sample stage surface temperature predicted value 1, wafer temperature predicted value 1, actual wafer temperature predicted value 1 and actual temperature are calculated from the time variation of the sensor temperature and heater power and plasma heat input by isomorphic observation using the linear differential equation in step S3-6. The predicted value of the wafer temperature is calculated.

  Next, a method for obtaining plasma heat input under actual process conditions will be described. FIG. 4 is a diagram for explaining the outline of this method. In step S4-1, the wafer temperature and the sensor temperature are measured when the pseudo wafer is mounted on the sample stage, the heater power is turned off, and a steady state is reached under some plasma heat input conditions. In step S4-2, it is assumed that the relationship between the sensor temperature and the wafer temperature is linear, and a conversion matrix from the sensor temperature to the wafer temperature is created.

In step S4-3, the steady value of the sensor temperature under actual process conditions is measured. In step S4-4, the plasma heat input is calculated from the steady value of the sensor temperature and the steady value of the wafer temperature using the linear differential equation in step S3-6.

  FIG. 6 is a diagram for explaining a microwave plasma etching apparatus equipped with an automatic wafer temperature control function. In this apparatus, the microwave generated by the magnetron 5 is introduced into the decompression processing chamber 1 through the waveguide 6 and the quartz window 7 to generate plasma 8. At this time, the processing gas introduced from the gas inlet 9 is dissociated by the plasma 8 to generate radicals and positive ions. A butterfly throttle valve 3 is provided between the decompression processing chamber 1 and the pump, and the pressure in the processing chamber can be controlled by adjusting the opening of the throttle valve 3.

Further, the wafer 11 that is the material to be etched is placed on the sample table 10 to which a high frequency power source is connected. A negative bias voltage can be generated on the wafer by applying power from the high-frequency power source to the sample stage via the matching unit. The sample is etched by accelerating ions in the plasma with this negative bias voltage and irradiating the wafer (this high-frequency power is hereinafter referred to as bias power).

  Further, a barium fluoride window 12 is provided on the side surface of the processing chamber, and when there is no plasma, the surface temperature of the sample table can be measured with a radiation thermometer. In addition, this apparatus is provided with a circulating refrigerant cooling device, and the entire sample stage is cooled by circulating the refrigerant cooled to a constant temperature by the circulating refrigerant cooling apparatus between the circulating refrigerant cooling apparatus and the sample stage. It has a structure.

  FIG. 7 is a diagram for explaining the details of the sample stage 10. A dielectric film is sprayed on the surface of the sample stage 10, and a pair of positive and negative ESC electrodes 21a and 21b are provided in the film. By applying a DC voltage between the electrodes, the wafer 11 is sampled. The structure is made to adsorb to the base 10. Further, the dielectric film is provided with heaters 13a, 13b, and 13c divided into three regions of a center, a middle, and an edge for controlling the temperature distribution of the sample stage. The heater power supplies 14a, 14b, and 14c are used independently. Can be heated. Further, temperature sensors 15a, 15b, and 15c for measuring the temperatures of the center, middle, and edge sample stages are embedded in the dielectric film, respectively, and the control arithmetic unit 16 is based on the output signal of each temperature sensor. Is configured to control the output of each heater power supply. Further, in order to enhance the heat conduction between the wafer and the sample table, a constant pressure of He is filled between the wafer back surface and the sample table.

  A method for performing the first wafer temperature prediction control using the etching apparatus shown in FIG. 6 will be described according to the procedure of FIG.

  First, in step S1-1, the heater power u1 is increased stepwise, and the change in wafer temperature x1 and the change in sensor temperature y1 at this time are measured. Specifically, first, the refrigerant pressure is set to 30 ° C., and the He pressure is adjusted to 1 kPa in a state where the pseudo wafer with a temperature measurement function is mounted and adsorbed on the sample stage. When the wafer temperature reaches 30 ° C. after sufficient time, the outputs u1 of the heater power supplies 14a, 14b, and 14c are sequentially changed from 0 W to 1000 W, and the time of the wafer temperature x1 and the sensor temperature y1 at each of the center, middle, and edge portions Measure changes.

In step S1-2, the relationship between the wafer temperature x2, the sensor temperature y1, and the heater power u1 is approximated by equation (1). As a specific approximation method, the least square method is used, and the values of the elements of the constant matrices A11 to A22 and B11 to B21 are determined.

The subscripts c, m, and e mean the center, middle, and edge parts.

In step S1-3, based on the constant matrices A11 to A22 and B11 to B21, the wafer temperature in a desired process condition is predicted using isomorphous observation. Specifically, first, a variable z2 is defined, and the time change of z2 is calculated by equation (2).

A predicted value (˜x2) of the wafer temperature is determined from z2 by the following equation (3).

If calculation is performed with L set to an appropriate value, the predicted value (˜x2) substantially matches the wafer temperature when the heat input from the plasma is sufficiently small. Accordingly, if appropriate PI control is performed on (˜x2), high-speed and high-precision wafer temperature control is possible. In this embodiment, the unit matrix of Expression (4) is used as the value of L.

  Using this method, the heater power is controlled by sequentially changing the target value of the wafer temperature at the center / middle / edge part from 30/30/30 ° C. to 70/70/70 ° C. and then 70/60/50 ° C. did. FIG. 8 shows a change in the predicted value (˜x2) of the wafer temperature at this time, and FIG. 9 shows a change in the measured value x2 of the wafer temperature. It can be seen that the wafer temperature also changes following the change in the predicted value (˜x2), and the wafer temperature can be controlled to a desired value at high speed.

  Next, for comparison, a conventional method for PI control of the heater power with respect to the sensor temperature was also examined. First, using a pseudo wafer with a temperature measuring function, the wafer temperature x2 at the center / middle / edge portion is (a) 30/30/30 ° C., (b) 70/70/70 ° C. and (c) 70/60/50. The heater power u1 was adjusted so that the temperature was 0 ° C., and the sensor temperature y1 at the center / middle / edge portion at this time was measured. The measured values were (a) 30/30/30 ° C, (b) 59/58/55 ° C, and (c) 59/52/39 ° C, respectively. The heater power was controlled by setting this value as a target value of the sensor temperature at the center / middle / edge portion. FIG. 10 shows a change in sensor temperature y1 at this time, and FIG. 11 shows a change in measured value x2 of the wafer temperature.

  Although the sensor temperature y1 has reached the target value in 40 to 70 s after the start of processing, the wafer temperature x2 has not reached the target temperature indefinitely. That is, the time required for control is long. Further, in this method, it is necessary to adjust the heater power so as to obtain a desired wafer temperature using a pseudo wafer with a temperature measurement function each time the process condition is changed, and to check the sensor temperature at that time.

 As described above, according to this embodiment, it is understood that the wafer temperature can be controlled to a desired value at high speed and with high accuracy without actually measuring and adjusting the wafer temperature for each process condition. Further, even when there are machine differences in the sensor temperature measurement position and the heat transfer coefficient between the sample stage and the sample, high-speed temperature control without machine difference is possible.

 In this embodiment, the values of the constant matrices A11 to A22 and B11 to B21 are fixed to one. However, when the temperature of the refrigerant is greatly different or the flow rate of the refrigerant is different, the predicted value and the actual measurement are obtained. There is a case where the value shifts. In such a case, high-speed and high-precision control is possible by acquiring a matrix for each refrigerant condition and switching at the same time as changing the refrigerant.

A method for performing the second wafer temperature prediction control using the etching apparatus shown in FIG. 6 will be described according to the procedure of FIG.
First, in step S2-1, the time change of the wafer temperature x1 and the sensor temperature y1 of each part of the center, middle, and edge is measured as in step S1-1 of the first embodiment. In step S2-2, the heater power u1 is increased stepwise in exactly the same time sequence as in step S2-1, and the change in the sample stage temperature x3 is measured with the radiation thermometer of FIG. In step S2-3, the relationship between the heater power u1, the wafer temperature x1, the sensor temperature y1, and the sample stage temperature x3 measured in steps S2-1 and 2-2 is approximated by equation (5). As a specific approximation method, the values of the constant matrices A11 to A33 and B11 to B31 are determined using a least square method.

The subscripts c, m, and e mean the center, middle, and edge parts.

Next, in step S2-3, based on the constant matrices A11 to A33 and B11 to B31, the wafer temperature in a desired process condition is predicted using isomorphous observation. Specifically, first, the variables z2 and z3 are defined, and the time change of z2 and z3 is calculated by Equation (6).

The predicted value (˜x2) of the wafer temperature is determined from the z2 and z3 by the equation (7).

If calculation is performed by setting L1 and L2 to appropriate values, the predicted value (˜x2) substantially matches the wafer temperature when the heat input from the plasma is sufficiently small. Therefore, high-speed and high-accuracy wafer temperature control is possible if appropriate PI control is performed on the predicted value (˜x2). In the present embodiment, the unit matrices of the equations (8) and (9) are used as the values of L1 and L2.

  Using this method, the heater power is controlled by sequentially changing the target value of the wafer temperature at the center / middle / edge part from 30/30/30 ° C. to 70/70/70 ° C. and then 70/60/50 ° C. did. FIG. 12 shows the change in the predicted value (˜x2) of the wafer temperature at this time, and FIG. 13 shows the change in the measured value x2 of the wafer temperature.

  It can be seen that the wafer temperature also changes following the change in the predicted value (˜x2), and the wafer temperature can be controlled to a desired value at high speed. Further, it can be seen that the control can be performed with higher speed and higher accuracy than the first method shown in FIG.

  As described above, if the second wafer temperature prediction control method is used, it is possible to perform wafer temperature control with higher speed and higher accuracy than the first wafer temperature prediction control method and the conventional method. Further, according to this method, even when there is a difference in the sensor temperature measurement position and the heat transfer coefficient between the sample stage and the sample, high-speed temperature control without any difference is possible.

  In this embodiment, the values of the matrices A11 to A33 and B11 to B31 are fixed to one. However, when the temperature of the refrigerant is significantly different or the flow rate of the refrigerant is different, the predicted value and the actual measurement are obtained. There is a case where the value shifts. In such a case, high-speed and high-precision control is possible by acquiring a matrix for each refrigerant condition and switching at the same time as changing the refrigerant. In this example, the surface temperature of the sample stage was measured in a vacuum, but a sufficient effect can be obtained even if measured in the air.

  A third wafer temperature prediction control method will be described according to the procedure shown in FIG. 3 using the etching apparatus shown in FIG. In FIG. 6, in steps S3-1 to S3-4, the same processing as that in the second embodiment is executed to calculate a predicted value of wafer temperature (˜x2) and a predicted value of electrode surface temperature (˜x3). In step S3-5, temporal changes in wafer temperature x2 and sensor temperature y1 are measured under any three types of plasma conditions and different heater heating conditions. Specifically, in a state where a wafer with a temperature measuring function is mounted and adsorbed on a sample stage, the He pressure is adjusted to 1 kPa, and after a sufficient time has passed, the wafer temperature reaches 30 ° C. (1) Conditions where the plasma heat input at the center is large and the center / middle / edge set temperature is controlled at 40/30/30 ° C. (2) Heat input at the middle is large The wafer is processed under three conditions: a condition controlled at 30/40/30 ° C., and (3) a condition where the heat input at the edge is large and a condition controlled at 30/30/40 ° C., and the wafer temperature x2 And the change of sensor temperature y1 is measured.

Next, in step S3-6, the wafer temperature x2, the sensor temperature y1, the predicted value of the wafer temperature (˜x2) calculated by the equation (7), the predicted value of the electrode temperature (˜x3), and the plasma heat input u2 The relationship is approximated by equation (10). Specifically, the constant matrices A41 to A44 and B42 are determined using the least square method.

At this time, the plasma heat input u2 under each condition does not need to be measured accurately, and any three independent matrixes may be used as shown in Equation (11).

Next, in step S3-7, after calculating the plasma heat input u2 in the actual process conditions by a method described later, in step S3-8, using the isomorphic observation, the predicted value of the wafer temperature at the time of plasma heat input (^ X2) (^ added to x is represented by (^ x) for convenience) is calculated. Specifically, a predicted value (^ x2) of the wafer temperature at the time of plasma heat input is calculated using equations (12) and (13). However, the value of the plasma heat input amount u2 is changed from 0 to the value obtained in step S3-7 simultaneously with the start of plasma heat input.

In this embodiment, the unit matrix of Expression (14) is used as the value of L3.

  Next, details of the calculation method of the plasma heat input amount u2 in the actual process conditions shown in step S3-7 will be described according to the procedure of FIG.

In step S4-1, without performing the heater control, the processing of the above-described conditions (1), (2), and (3) is performed, and the wafer temperature and the sensor temperature at the time when the steady state is reached are measured. And the difference (change amount) between the temperature before plasma heat input. Next, in step S4-2, a conversion matrix from the sensor temperature to the wafer temperature is obtained by Expression (15) using the wafer temperature and the change amount of the sensor temperature.

Next, in step S4-3, plasma processing is performed under the desired process conditions without heater control, and a difference Δy between the sensor temperature and the initial temperature when the steady state is reached is obtained. Next, in step S4-4, a temperature increase Δx due to plasma heat input is obtained from the temperature difference Δy according to equation (16).

  Next, in step S4-5, the sensor temperature increment Δy and the temperature rise Δx due to plasma heat input are substituted into equation (10), the differential term on the left side is set to 0, and the plasma heat input amount u2 of the desired process condition is set. Ask for.

  If this method is used, the wafer temperature can be accurately predicted even in the presence of plasma heat input. Therefore, if appropriate PI control is performed on the predicted value (^ x2), the wafer temperature can be controlled at high speed and with high accuracy.

  Using this method, the target temperature of the center / middle / edge portion was controlled from 30/30/30 ° C. to 70/70/70 ° C. and further to 70/60/50 ° C., respectively. FIG. 14 shows the change in the predicted value (^ x2) of the wafer temperature at this time, and FIG. 15 shows the change in the measured value x2 of the wafer temperature. For comparison, FIGS. 16 and 17 show changes in the predicted value (˜x2) and the actual measurement value x2 when controlled by the second wafer temperature predictive control method of the present invention, respectively.

  In the second wafer temperature prediction control method, the actual measured value x2 of the wafer temperature becomes higher than the predicted value (up to x2) at the time of plasma heat input for 50 to 70 s and 110 s to 130 s after the start of processing, so the wafer temperature is higher than the target temperature. turn into. If the third wafer temperature predictive control method is used, since the measured value x2 of the wafer temperature coincides with the predicted value (^ x2) even during the plasma heat input, by controlling the predicted value (^ x2), It can be seen that the wafer temperature x2 can be controlled at high speed and with high accuracy. Further, in this method, even when there is a difference in the measurement position of the sensor temperature and the heat transfer coefficient between the sample stage and the sample, high-speed temperature control without any difference is possible.

  In the present embodiment, the example in which the values of the matrices A41 to A44 and B41 are fixed to one has been described. However, when the refrigerant temperature is greatly different or the refrigerant flow rate is different, the predicted value and the actual measurement value are described. There is a case that shifts. In such a case, high-speed and high-precision control is possible by acquiring a matrix for each refrigerant condition and switching at the same time as changing the refrigerant. In this example, the surface temperature of the sample stage was measured in a vacuum, but a sufficient effect can be obtained even if measured in the air.

  As described above, according to the embodiment of the present invention, the heater power, the wafer temperature, and the sensor temperature are measured, the relationship is approximated by the simultaneous linear differential equation, and the isomorphic observation using the simultaneous differential equation is performed. The wafer temperature is predicted by, and feedback control is performed at the predicted wafer temperature. For this reason, the wafer temperature can be controlled at high speed and stably. Further, according to the first wafer temperature prediction method of the present invention, the wafer temperature can be accurately predicted when the plasma heat input is sufficiently small. Further, according to the second wafer temperature prediction method, a wafer temperature with higher accuracy than that of the first method is possible. Further, according to the third wafer temperature prediction method, the wafer temperature can be predicted with high accuracy even when there is plasma heat input. If the heater power feedback control is performed based on the wafer temperature prediction value thus obtained, the wafer temperature can be controlled at high speed and stably as described above.

DESCRIPTION OF SYMBOLS 1 Pressure reduction processing chamber 2 Pump 3 Pressure control valve 4 Pressure gauge 5 Magnetron 6 Waveguide 7 Quartz window 8 Plasma 9 Gas inlet 10 Sample stage 11 Sample 12 Barium fluoride window 13 Heater 14 Heater power supply 15 Temperature sensor 16 Control arithmetic device 18 High frequency power source 19 Matching device 20 Circulating refrigerant cooling device

Claims (3)

A sample stage having a mounting surface capable of adsorbing and holding a sample on the upper surface and having a heat transfer medium flow path formed therein;
A plurality of heaters provided in the sample table divided into a plurality of regions in the sample table surface;
A plurality of temperature sensors for measuring the temperature of the sample stage corresponding to the plurality of regions,
Time variation of each temperature of the plurality of temperature sensors measured without plasma treatment, time variation of each temperature of the sample corresponding to the plurality of regions, and each temperature obtained by subtracting the temperature of the heat transfer medium The correlation between the temperature of the sensor and the sample and the power supplied to the heater is an approximate expression,
Using the respective approximate equations, the respective temperatures of the plurality of temperature sensors during the plasma processing and the respective powers supplied to the plurality of heaters are used to calculate the respective temperatures of the samples corresponding to the plurality of regions during the plasma processing. Predict,
A control apparatus for controlling the temperature in the surface of the sample during the plasma processing using each predicted temperature of the sample is provided. A mounting surface capable of adsorbing and holding a sample on the upper surface, a flow path of a heat transfer medium formed inside, a plurality of heaters provided inside divided into a plurality of regions within the surface, and corresponding to the plurality of regions A plurality of temperature sensors that measure the temperature of the location, and in a sample stage in which the temperature in the surface of the sample during plasma processing is controlled by a temperature control device,
The temperature control device is configured to measure a time change of each temperature of the plurality of temperature sensors measured in a state where the plasma treatment is not performed, a time change of each temperature of the sample corresponding to the plurality of regions, and a temperature of the heat transfer medium. Correlation between the reduced temperature sensor and sample temperature and the power supplied to the heater is an approximate expression,
Using the respective approximate equations, the respective temperatures of the plurality of temperature sensors during the plasma processing and the respective powers supplied to the plurality of heaters are used to calculate the respective temperatures of the samples corresponding to the plurality of regions during the plasma processing. Predict,
The temperature of the surface of the sample during the plasma processing is controlled using the predicted temperature of each sample. A plasma processing chamber in which the sample is plasma-processed, a high-frequency power source for supplying high-frequency power for generating plasma in the plasma processing chamber, and a mounting surface on which the sample can be adsorbed and held are provided on the upper surface. A sample table in which a flow path of the heat medium is formed; a plurality of heaters provided in the sample table divided into a plurality of regions in the sample table surface; and a plurality of measuring the temperature of the sample table corresponding to the plurality of regions In a plasma processing apparatus comprising: a temperature sensor of: and a temperature control device that controls a temperature in a plane of a sample during plasma processing;
The temperature control device is configured to measure a time change of each temperature of the plurality of temperature sensors measured in a state where the plasma treatment is not performed, a time change of each temperature of the sample corresponding to the plurality of regions, and a temperature of the heat transfer medium. Correlation between the reduced temperature sensor and sample temperature and the power supplied to the heater is an approximate expression,
Using the respective approximate equations, the respective temperatures of the plurality of temperature sensors during the plasma processing and the respective powers supplied to the plurality of heaters are used to calculate the respective temperatures of the samples corresponding to the plurality of regions during the plasma processing. Predict,
The temperature in the surface of the sample during the plasma processing is controlled using the predicted temperature of each sample. The plasma processing apparatus.

Images