WO2003077303A1

WO2003077303A1 - Plasma processing method, seasoning end detection method, and plasma processing device

Info

Publication number: WO2003077303A1
Application number: PCT/JP2003/002932
Authority: WO
Inventors: Naoki Takayama; Bin Wang; Satoshi Harada
Original assignee: Tokyo Electron Limited
Priority date: 2002-03-12
Filing date: 2003-03-12
Publication date: 2003-09-18
Also published as: AU2003221362A1; CN1643664A; JP4173311B2; CN100355040C; JP2003264179A; TW200421411A; TWI297904B

Abstract

With conventional analysis data, it is difficult to decide whether the change serving as a judgment reference for a seasoning end is a change caused by the seasoning, i.e., a change based on the state change in a processing vessel or a change based on a temperature change between dummy wafers and to decide whether the seasoning itself is completed. According to the plasma processing method of the present invention, a method for detecting a seasoning end when performing seasoning by supplying a dummy wafer (W) into a processing vessel (2) of a plasma processing device (1) includes a step of supplying a dummy wafer (W) into the processing vessel (2), cooling the interior of the processing vessel (2), and performing multivariate analysis using a plurality of measurement data obtained when a plurality of dummy wafers (W) are supplied again into the processing vessel (2) so as to create a prediction equation for predicting the seasoning end and a step of detecting a seasoning end when performing the seasoning according to the prediction equation.

Description

Description Plasma processing method, end-of-sizing detection method, and plasma processing apparatus

The present invention relates to a plasma processing method, a seasoning end detecting method, and a plasma processing apparatus used in an etching processing apparatus or the like. Background technology

For example, a processing apparatus such as an etching processing apparatus includes a processing container having an airtight structure, and a holding member disposed in the processing container and holding a processing object, and generates plasma in the processing container to perform processing. It is configured to perform predetermined processing on the body. If the processing of the object to be processed is continued, the inside of the processing container is contaminated with by-products and the like, and internal components are consumed. For this reason, the processing equipment is temporarily stopped, and maintenance such as cleaning of the processing container and replacement of consumables is performed. Then, after the maintenance is completed, the processing device is restarted.

For example, in the case of an etching processing apparatus, when restarting, a predetermined number of dummy wafers are supplied into the processing container and the etching cycle is repeated to perform so-called seasoning in which the processing container is adjusted to a state required for production. After the seasoning, the etching rate and the uniformity of the etching in the wafer surface are examined. Data analysis is performed using measurement data obtained from a plurality of dummy wafers during seasoning, for example, measurement data of the emission spectrum obtained by an end point detector. Then, it is determined whether or not the seasoning has ended by observing the change in the analysis data. However, according to the conventional analysis, the change that is a criterion for ending seasoning is a change due to seasoning, that is, a change based on a state change in the processing chamber, or a change based on a change in temperature between dummy wafers. It is difficult to determine if there is. Therefore, it is difficult to determine whether seasoning has been completed.

That is, for example, the present inventors performed a principal component analysis, which is one of the multivariate analyses, on the basis of the measurement data. The data were analyzed as described below, and two large peaks indicating changes were observed in the analysis results, making it difficult to judge whether or not the seasoning was completed. No. Therefore, this principal component analysis will be described. In this case, measurement data is collected by the same method as before. For example, on the first day, 130 dummy wafers were supplied, and on the second day, the production process was started and 30 dummy wafers were supplied to perform etching. Principal component analysis was performed using the measurement spectrum of the luminescent spectrum obtained from the 51st to 60th dummy wafers and the 121st to 130th dummy wafers on the first day. The analysis results shown in FIGS. 8A, 8B, 9A, and 9B were obtained. In this principal component analysis, 19.3 ηπ! Using a set of 297 wavelengths in the short wavelength range of ~ 19 nm, a single dummy wafer was measured for each wavelength 18 times every 3 seconds for 1 minute T. Principal component analysis was performed based on. Then, the principal component scores and residuals at each measurement are obtained, respectively, and the plots of HOTELLINGS TSQUARE (sum of squares of principal component scores) are shown in Fig. 8A and Fig. 9A. 8B and 9B are plots of the scores. As is evident from the results of these analyses, a large peak was observed in both graphs on the first day and the second day, making it difficult to judge the end of seasoning. The horizontal axis of each graph indicates the number of measurements.

The present invention has been made to solve the above problems, and has as its object to provide a plasma processing method and a plasma processing apparatus and a plasma processing apparatus capable of clearly determining the end of seasoning. .

The present inventors have conducted various studies on the cause of the two beaks, and as a result, it has been found that the cause is in the method of collecting the measured data used for the data overnight analysis. Therefore, by performing a specific treatment on the processing container when sampling the data, the effects of the temperature change between the dummy wafers are eliminated, the change due to seasoning is ascertained, and the end of seasoning is assured. We got the knowledge that we can judge. Disclosure of the invention

The present invention has been made based on the above findings, and the plasma processing method according to claim 1 of the present invention supplies a test object to be processed into a processing vessel of a processing apparatus to provide a system. —In the plasma processing method for detecting the end of the seasoning when performing the seasoning, the test object to be tested is supplied into the processing chamber, the processing chamber is cooled, and then the processing chamber is again cooled. Performing a multivariate analysis using a plurality of measurement data obtained when supplying a plurality of the test objects to the test object to form a prediction formula for predicting the end of the seasoning, based on the prediction formula Detecting the end of seasoning when performing the seasoning.

The plasma processing method according to claim 2 of the present invention is characterized in that, in the invention according to claim 1, principal component analysis is used as the multivariate analysis.

The plasma processing method according to claim 3 of the present invention is characterized in that, in the invention according to claim 1 or 2, a plasma emission spectrum is used as the measurement data. I do.

The plasma processing method according to claim 4 of the present invention is characterized in that, in the invention according to claim 3, among the wavelengths of the emission spectrum, a wavelength having a high contribution to a residual is used. Features.

The plasma processing method according to claim 5 of the present invention is the plasma processing method according to claim 1 or 2, wherein the high-frequency voltage obtained by the electric measurement device is used as the measurement data. Is used.

The plasma processing method according to claim 6 of the present invention uses the high-frequency current obtained by an electric measurement device as the measurement data in the invention according to claim 1 or 2. It is characterized by the following.

The plasma processing method according to claim 7 of the present invention is the plasma processing method according to claim 1 or 2, wherein the high-frequency voltage obtained by the electric measurement device is used as the measurement data. And a high-frequency current.

The method for detecting the end of seasoning according to claim 8 of the present invention is a method for detecting the end of the seasoning when the test object is supplied into the processing container of the processing apparatus to perform the seasoning, After supplying the test object into the processing container, cooling the processing container, and using a plurality of measurement data obtained when supplying the test object into the processing container again. Perform multivariate analysis with The method includes a step of creating a prediction formula for predicting the end of seasoning, and a step of detecting the end of seasoning when performing the seasoning based on the prediction formula.

A plasma processing apparatus according to claim 9 of the present invention includes: a processing container accommodating an object to be processed; a detector for measuring a light emission spectrum of plasma in the processing container; and a connection to the detector. And a control device to which the measurement data from the detector is input, and when the test object is supplied into the processing container and seasoning is performed, the test object is placed in the processing container. Supply, and after cooling the inside of the processing container, a multivariate analysis program is executed based on a plurality of measurement data measured by the detector when supplying a plurality of the test objects to the processing container again. And a controller for detecting the end of the seasoning when performing the seasoning based on the prediction formula. Features.

A plasma processing apparatus according to claim 10 of the present invention includes: a processing container for accommodating an object to be processed; an electric measuring device provided in the processing container; and an electric measuring device connected to the electric measuring device. A control device to which measurement data from a measuring device is input. When seasoning is performed by supplying the test object into the processing container, the test object is stored in the processing container. Supply, and after cooling the inside of the processing container, multivariate analysis is performed based on the multiple measurement data measured by the detector when supplying multiple test objects to the processing container again. A control device that performs a multivariate analysis using a program, creates a prediction formula for predicting the end of the seasoning, and detects the end of the seasoning when performing the seasoning based on the prediction formula. When That. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a configuration diagram showing an example of a plasma processing apparatus to which a seasoning data analysis method, a plasma processing method, and a seasoning end detection method of the present invention are applied.

FIGS. 2A and 2B show the plasma shown in FIG. 1 obtained according to an embodiment of the present invention. Fig. 2A is a graph showing the results of analysis of measurement data for the processing device. Fig. 2A is a graph showing the variation of the sum of squares of the main component scores of the measurement data. It is a graph.

FIGS. 3A and 3B are diagrams showing analysis results obtained by another embodiment of the present invention, and are graphs corresponding to FIGS. 2A and 2B, respectively.

FIG. 4 is a graph showing a contribution ratio of measured data of a light emission spectrum used in still another embodiment of the present invention to a residual.

5A and 5B are diagrams showing analysis results obtained using the average values of the wavelengths shown in FIG. 4, and are graphs corresponding to FIGS. 2A and 2B, respectively.

FIGS. 6A and 6B are diagrams showing analysis results obtained using the wavelengths shown in FIG. 4, and are graphs corresponding to FIGS. 2A and 2B, respectively.

FIG. 7 is a configuration diagram showing another example of a plasma processing apparatus to which the seasoning analysis method and the seasoning end detection method of the present invention are applied.

Figs.8A and 8B show the analysis results obtained by the conventional analysis method, and the results of using the 51st to 60th dummy wafers on the first day of seasoning are shown in Figs. It is a graph corresponding to 2 B.

FIGS. 9A and 9B show other analysis results obtained by the conventional analysis method. Each of the results when using the 121st to 130th dummy wafers on the first day of seasoning is shown in FIG. A, a graph corresponding to Figure 2B. BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, the present invention will be described based on the embodiments shown in FIGS.

As shown in FIG. 1, for example, a plasma container 1 of the present embodiment can maintain a desired high degree of vacuum, and has a processing vessel 2 whose surface is anodized and electrically grounded. A lower electrode 3 which is disposed at the center of the bottom surface of the inside and on which an object to be processed (for example, silicon) W is mounted; an insulating member 2 A which supports the lower electrode 3 from below, and And a lower electrode 3 interposed therebetween and a hollow upper electrode 5 provided with a gap therebetween. For example, a 2 MHz high-frequency power supply 6 is connected to the lower electrode 3 via a matching unit 6 A, and is connected to the upper electrode 5. Is connected to a high-frequency power source 7 having a frequency higher than that of the lower electrode 3, for example, 60 MHz, via a matching unit 7A. A high-pass filter 8 is connected to the lower electrode 3, and a low-pass filter 9 is connected to the upper electrode 5. An exhaust device 11 is connected to an exhaust port 2B on the bottom surface of the processing container 2 through a gas exhaust pipe 11A. The exhaust device 11 evacuates the processing container 2 to maintain a desired degree of vacuum. I do. Hereinafter, the lower electrode 3 and the support 4 will be collectively referred to as a mounting table 10 as needed.

A gas introduction pipe 5A is formed at the center of the upper surface of the upper electrode 5, and this gas introduction pipe 5A penetrates the center of the upper surface of the processing vessel 2 via an insulating member 2C. A processing gas supply source 12 is connected to the gas introduction pipe 5A via a gas supply pipe 13, and supplies an etching gas from the processing gas supply source 12. That is, the processing gas supply source 12, C ₅ F ₈ have a gas supply source 12 A, 0 ₂ gas supply source 12 B and Ar gas supply source 12 C, supply 12 Alpha Each of these gases, 12B 12 C is Each is connected to a branch pipe 13A, 13B, 13C of the gas supply pipe 13. Each branch pipe 13 A, 13 B, 1 3 to C C ₅ F ₈ gas supply source 12 A, 0 ₂ gas supply source 12B and the flow rate control device corresponding to the Ar gas supply source 1 2 C 12 D, 12 E , 12F and valves 12G, 12H, 12I are sequentially provided from the upstream side to the downstream side, and inside the processing vessel 2 via these flow controllers 12D, 12E, 12 and the valves 120, 12H, 12I. The etching gas supplied to the furnace is controlled to a predetermined flow rate.

A large number of holes 5B are formed on the lower surface of the upper electrode 5 so as to be evenly dispersed, and the processing gas is uniformly distributed and supplied into the processing vessel 2 from each hole 5B. Accordingly, while the inside of the processing vessel 2 is evacuated by the exhaust device 11 and a predetermined etching gas is supplied at a predetermined flow rate from the processing gas supply source 12, the lower electrode 3 and the upper electrode 5 are each raised in height. A frequency power is applied to generate plasma of an etching gas in the processing container 2, and a predetermined etching is performed on the wafer W on the lower electrode 3. A temperature sensor (not shown) is mounted on the lower electrode 3, and the temperature of the wafer W on the lower electrode 3 is constantly monitored via the temperature sensor.

A coolant passage 1 OA through which a predetermined coolant (for example, conventionally known fluorine-based fluid, water, etc.) passes is formed in the mounting table 10, and the lower electrode 3 is cooled while the coolant flows through the coolant passage 1 OA. Then, the wafer W is cooled via the lower electrode 3 and the wafer W is controlled to a desired temperature. I do. An electrostatic chuck 14 made of an insulating material is arranged on the lower electrode 3, and a high-voltage DC power supply 15 is connected to an electrode plate 14 A in the electrostatic chuck 14. The electrostatic chuck 14 electrostatically attracts the wafer W by static electricity generated on the surface by the high voltage applied to the electrode plate 14A from the high-voltage DC power supply 15. A focus ring 16 surrounding the electrostatic chuck 14 is arranged on the outer periphery of the lower electrode 3, and plasma is focused on the wafer W via the focusing ring 16.

A gas passage 10B for supplying a thermally conductive gas such as He gas as a backside gas is formed in the mounting table 10, and the gas passage 10B is opened at a plurality of locations on the upper surface of the mounting table 10. are doing. These openings correspond to through holes formed in the electrostatic chuck 14 on the mounting table 10. Therefore, when the backside gas is supplied to the gas passage 10B of the mounting table 10, the backside gas flows out of the through hole of the electrostatic chuck 13 via the gas passage 10B, and the electrostatic chuck 1 It spreads evenly throughout the gap between 4 and the wafer W, increasing the thermal conductivity in the gap. In FIG. 1, reference numeral 17 denotes a gate valve formed in the processing container 2 for opening and closing the loading / unloading port of the wafer W.

For example, an end point detector 18 is attached to the plasma processing apparatus 1. The end point detector 18 is used to measure the emission spectrum of the plasma in the processing vessel 2, and the measured value is stored in the control unit 19. I try to take in. The controller 19 stores, for example, a program for principal component analysis as a multivariate analysis program, and performs principal component analysis via this program. This principal component analysis program is used to analyze the data for seasoning when the processing container 2 is seasoned. As the data analysis data, the measurement data of the emission spectrum of the end point detector 18 is used. As the measurement data, for example, 104 kinds of wavelengths in the range of 193 nm to 950 nm are used.

Therefore, the seasoning data analysis method of the present embodiment, that is, sampling of measurement data used for creating a so-called prediction equation for predicting the end of seasoning will be described. Immediately, after replacing the consumables such as the cleaning inside the processing container 2 and the focus ring (not shown), seasoning is performed in the following procedure to stabilize the processing container 2. First, after starting the plasma processing apparatus 1 on the first day, a dummy wafer (bare silicon) W is supplied into the processing chamber 2. After that, the gas supply pipe 16 into the processing vessel 2 Etching is performed by applying a high frequency power of, for example, 6 OMHz and 2 MHz from the high frequency power supplies 6 and 7 while supplying a etching gas and maintaining a predetermined degree of vacuum. This process is repeated for, for example, 130 dummy wafers W. The processing on the first day is completed with the processing of 130 dummy wafers W.

Thereafter, the etching process is temporarily stopped, and the processing container 2 is left for several hours or more with the power on, that is, in a state where it can be restarted immediately. Then, the internal components such as the processing container 2 itself and the lower electrode 12 that have become high in temperature due to the etching process are cooled to the set temperature.

Next, on the second day, for example, 30 dummy wafers W are supplied one by one into the processing container 2 one by one under the conditions of the production process, and the etching cycle is repeated for each dummy wafer W. At the start of the etching cycle, since the inside of the processing container 2 is cooled, during the etching from the first dummy wafer W to the 30th wafer, the processing container 2 itself and the lower electrodes 12 in the processing container 2 are removed. The temperature of components such as the focus ring gradually rises. In this embodiment, out of the 30 wafers, the light emission spectrum was measured 18 times in about one minute for the dummy wafer W at the time when there was a temperature change from the first to the 20th wafer. The emission intensity of the wavelength is used as measurement data of the principal component analysis. Therefore, the temperature change in the processing container 2 is reflected in these measurement data overnight. Then, the principal component analysis is performed using the above measurement data. For example, 20 dummy wafers are measured m times (in this embodiment, 18 × 20 = 360 times), and n measurements (in this embodiment, 297 wavelengths) are performed for each measurement. If there is measurement data of (emission intensity), the matrix containing the measurement data is expressed by Equation 1. The rows of this matrix have the measurement data of the measurement wavelength obtained by one measurement as its components, and the columns have the measurement data that changes with time at each wavelength as its components. Then, the controller 19 calculates the average value, the variance value, and the standard deviation based on each measurement data, and then normalizes the average value and the standard deviation value. Using the correlation matrix based on these normalized values, the principal component analysis of a plurality of measurement data is performed to find the eigenvalue and its eigenvector. The eigenvalue represents the magnitude of the variance of each measurement data, and is defined as the first principal component, the second principal component,..., The nth principal component in the order of the magnitude of the eigenvalue. Each eigenvalue has its own eigenvector (weight). Usually higher order of principal component The lower the contribution to the evaluation of the data, the lower its utility value, "Equation 1"

^ 11 ^ 12 X In

^ 21 X 22 X 2ιι

X

In the present embodiment, m measurements are performed on 20 dummy wafers, n measurement data are taken for each measurement, and the j-th main value corresponding to the j-th eigenvalue of the i-th measurement is taken. The components are represented by Equation 2. Then, the j-th principal component t to Isseki measuring De obtained by the specific i-th measurement _{(Xn, X i 2, ·} · ·, X m) values obtained al a by substituting the i-th This is the j-th principal component score for the measurement. Therefore, the score tj of the j-th principal component is defined by Expression 3, and the eigenvector P of the j-th principal component is defined by Expression 4. Using the matrix X and the eigenvector P j, the score tj of the j-th principal component is expressed by Equation 5. The matrix X is expressed by Equation 6 using the principal component scores and their respective eigenvectors.

"Formula 2"

= ^X ill jl + ^X i2lj2 + ^{+ X} inPjn

"Formula 3"

2932

Ten

'Formula 4'

Pj2

P

-Equation 5 "

—Equation 6 ”

X = t Pi + t ₂ p ₂ + "" + t _n p _n ' ^] where p _n ^T is the transposed matrix of p _n , so in principal component analysis even if there are many types of measurement data For example, the first principal component, the second principal component, and at most a small number of statistical data up to the third principal component. Can be turned off. For example, in general, if the cumulative contribution ratio of the eigenvalues of the first and second principal components exceeds 90%, the evaluation based on the first and second principal components becomes highly reliable. The first principal component indicates the direction in which the measurement data is most dispersed as described above, and is suitable for grasping the temporal change of seasoning of the processing container 2 and determining the end of seasoning. Changes that cannot be fully grasped by the first and second principal components can be grasped by residual scores. In the present embodiment, the first principal component is obtained.

Therefore, in the present embodiment, the dummy wafer W is etched under the following conditions, and the eigenvalue can be obtained by the principal component analysis of the measured data at this time using the correlation matrix of the measured data. This is the variance of the first principal component score. The eigenvector of the first principal component can be obtained using the eigenvalue and the correlation matrix. And each Fig. 2A is a graph in which the principal component scores of the measurement data are calculated and the sum of squares of each principal component score (HOTELLINGS TSQUARE) is plotted. As is clear from this graph, a large peak was observed only in the measurement data on the first day, and no large peak was observed in the measurement day on the second day. Thus, the end of seasoning can be reliably determined. FIG. 2B is a graph plotting the sum of squares of the residual of each measurement data. In this figure, too, a large peak is observed only on the first day of the measurement, and the end of seasoning can be reliably determined. The horizontal axis of each graph indicates the number of measurements. In the present embodiment, one dummy wafer W is measured 18 times, and since 160 dummy wafers W are processed in two days, there is a scale up to 18 × 160 = 288 °.

[Processing conditions]

Processing equipment: Capacitively coupled parallel plate plasma processing equipment

Dummy wafer (bare silicon): 300mm

Lower electrode power supply high frequency and power: 2MHz, 3800W

Upper electrode power supply high frequency and power: 60MHz ヽ 3300W

Processing pressure: 25mT o r r

Etching gas: C ₅ F ₈ = 29 sc cm,

Ar = 750 sccm, 〇2 = 47 sccm Backside gas: He = 15 Torr (electrode center),

40 T 0 r r (electrode edge)

Processing temperature: upper electrode = 60 ° C, side wall = 6 O lower electrode = 20 ° C

As described above, according to the present embodiment, that is, according to the method of creating the prediction formula for predicting the end of the seasoning, the plasma processing apparatus 1 is temporarily stopped during the stage of entering the production process using the dummy wafer W. After that, the processing container 2 is left for several hours to cool the processing container 2 itself and the components such as the lower electrode 12 therein, and then enter the production process again to process the 20 dummy wafers W while the processing container 2 is being processed. Since the measurement data for judging the end of the sampling is collected, measurement data reflecting temperature changes of the processing container 2 itself and the components such as the lower electrode 3 can be obtained. Peaks based on change can be eliminated. Also, use this analysis result during seasoning. Thus, the end of seasoning can be reliably detected and determined. Therefore, the stable etching process can be performed on the wafer by performing the plasma process after reliably detecting the seasoning.

3A and 3B are views showing a data analysis method according to another embodiment of the present invention. In the present embodiment, unlike the above embodiment, an average value of 18 pieces of measurement data (297 wavelengths) obtained for each dummy wafer W is obtained, and the main component is calculated using these average values. An analysis was performed to determine eigenvalues and eigenvectors. The plots of the sum of squares of the principal component scores and the sum of squares of the residuals of each dummy wafer W are shown in FIGS. 3A and 3B. As is clear from FIGS. 3A and 3B, the end of the seasoning can be determined similarly to the graphs of the above-described embodiment shown in FIGS. 2A and 2B. The values on the horizontal axis indicate the number of dummy wafers.

FIGS. 4, 5A, and 5B are diagrams showing a data analysis method according to still another embodiment of the present invention. In the present embodiment, as shown in FIG. 4, for example, 10 kinds of wavelengths (for example, the wavelengths surrounded by 〇 in FIG. 4) that greatly contribute to the residual of the measurement data are selected, and these 10 kinds are selected. The average value of the wavelength was determined for each dummy wafer W in the same manner as in the above embodiment, and principal component analysis was performed using these average values to determine the eigenvalues and eigenvectors. The first principal component score and the residual score of each dummy wafer W are plotted in the graphs shown in FIGS. 5A and 5B. As is evident from FIGS. 5A and 5B, the jaggedness of the graph is reduced to a smoother curve compared to the embodiment shown in FIGS. 2A and 2B, making it easier to determine the end of seasoning. can do. 6A and 6B are diagrams illustrating a data analysis method according to still another embodiment of the present invention. In this embodiment, as in the embodiment shown in FIGS. 5A and 5B, ten types of wavelengths having a high contribution to the residual are selected. In the embodiment shown in FIGS. 5A and 5B, the time average value of the measurement data of each wavelength for the dummy wafer W is adopted, but in this embodiment, the measurement data itself is used. This is the same as the case shown in FIG. 2B. However, in the case of Fig. 2A and Fig. 2B, the rows of the matrix are the components of the measurement data of the measurement wavelength obtained by one measurement, and the columns are the measurement data that changes with time of each wavelength. In the present embodiment, rows and columns are transposed for each dummy wafer W and each wavelength. One line is 16 times per 10 wavelengths 03 02932

13 Because of the measurement, it has a component of 10 x 16 = 1 60. Since one column contains 20 wafers in the training set, the training set for principal component analysis having 20 components is a matrix of 20 rows and 16 columns. Principal component analysis is performed based on this matrix in the same manner as in the above embodiments, and the sum of squares of the principal component scores and the sum of squares of the residuals are plotted in graphs shown in FIGS. 6A and 6B. As can be seen from Figs. 6A and 6B, the jaggedness is reduced to a smoother curve compared to the graphs shown in Figs. 5A and 5B, making it easier to determine the end of seasoning. be able to.

In addition, the present invention can be applied to the plasma processing apparatus 20 shown in FIG. 7 similarly to the above-described plasma processing apparatus 1, and the same operational effects as those of the above-described plasma processing apparatus 1 can be expected. As shown in FIG. 7, the plasma processing apparatus 20 includes a processing container 21 made of a conductive material such as an aluminum, and a mounting plate disposed on the bottom surface of the processing container 21 and mounting a wafer W thereon. A lower electrode 22 serving also as a mounting table; a hollow grounded upper electrode 23 provided above the lower electrode 22 at a predetermined interval and serving also as a supply part of an etching gas; and a rotating magnetic field. And a rotating magnetic field B generated by the magnetic field forming means 24 acts on an electric field generated between the upper and lower electrodes of the processing vessel 21 under the control of the control device 25, thereby increasing the density. Perform uniform plasma processing on wafer W with plasma.

A gas supply pipe 26 connected to the upper electrode 23 is connected to the upper surface of the processing vessel 21, and a gas supply source (not shown) is connected to the processing vessel 2 via the gas supply pipe 26 and the upper electrode 23. Supply the etching gas into 1. A gas exhaust pipe 27 connected to a vacuum exhaust device (not shown) is connected to the side surface of the processing vessel 21, and the inside of the processing vessel 21 is depressurized through a vacuum exhaust device and a gas exhaust pipe 27 to a predetermined pressure. The degree of vacuum is maintained. A high-frequency power supply 28 is connected to the lower electrode 22, and high-frequency power is applied from the high-frequency power supply 28 to the lower electrode 22 to generate plasma of an etching gas between the two electrodes 22 and 23. For example, a predetermined etching process is performed on the surface of the upper semiconductor wafer W.

For example, an end point detector 29 is attached to the plasma processing apparatus 20. The final inspection detector 29 is used to measure the emission spectrum of the plasma in the processing vessel 21 and the measured value is used as a control unit 2 I try to take it in 5. The controller 25 stores, for example, a program for principal component analysis as a multivariate analysis program. The principal component analysis is performed via the RAM. This principal component analysis program is used to analyze the seasoning data when seasoning the processing vessel 21. As data for data analysis, measurement data of the light emission spectrum of the end point detector 29 is used. As the measurement data, for example, 104 kinds of wavelengths in the range of 193 nm to 950 nm are used.

In the above embodiments, the principal component analysis has been described as an example of the data analysis method for determining the end of seasoning, but other multivariate analysis may be used. In each of the above embodiments, the case of using the plasma emission spectrum has been described. However, other measurement data, such as a high-frequency signal detected by an electric measurement device (VI probe) provided in the plasma processing device, may be used. Measurement data that is easily affected by temperature changes in the processing vessel such as voltage, high-frequency current, and the phase difference between high-frequency voltage and high-frequency current can be used. In each of the above embodiments, the etching apparatus has been described as an example. However, the present invention can be applied to other plasma processing apparatuses. According to the invention set forth in claims 1 to 7 of the present invention, a plasma processing method, a seasoning completion detection method, and a plasma processing apparatus capable of clearly determining the end of seasoning are provided. be able to.

Claims

The scope of the claims

1. In the plasma processing method for detecting the end of the seasoning when the test object is supplied into the processing container of the processing apparatus to perform the seasoning,

A plurality of measurement data obtained when the test object is supplied into the processing container, and after the inside of the processing container is cooled, a plurality of the test objects are supplied again into the processing container. Performing a multivariate analysis using evening to create a prediction formula for predicting the end of the seasoning;

A step of detecting the end of the seasoning when performing the seasoning based on the prediction formula;

A plasma processing method comprising:

2. The plasma processing method according to claim 1, wherein principal component analysis is used as the multivariate analysis.

3. The plasma processing method according to claim 1, wherein a light emission spectrum of a plasma is used as the measurement data.

4. The plasma processing method according to claim 3, wherein a wavelength having a high contribution rate to a residual is used among the wavelengths of the emission spectrum.

5. The plasma processing method according to claim 1, wherein a high-frequency voltage obtained by an electric measurement device is used as the measurement data.

6. The plasma processing method according to claim 1 or 2, wherein a high-frequency current obtained by an electric measurement device is used as the measurement data.

7. The method according to claim 1 or 2, wherein a phase difference between a high-frequency voltage and a high-frequency current obtained by an electric measurement device is used as the measurement data. Zuma treatment method.

8. In the method for detecting the end of seasoning when supplying a test object to be processed into a processing container of a processing apparatus and performing seasoning,

Detecting the end of seasoning when performing the seasoning based on the prediction formula;

A method for detecting the end of a seasoning, comprising:

9. A processing container for storing the object to be processed,

A detector for measuring the emission spectrum of the plasma in the processing container,

A control device that is connected to the detector and receives the measurement data from the detector. When the test object is supplied into the processing container and seasoning is performed, the control device is connected to the processing container. After supplying the test object to the test container and cooling the inside of the processing container, a plurality of measurement data measured by the detector when supplying the test object to the test container again in the process container is supplied. Perform a multivariate analysis using a multivariate analysis program on the basis of the evening, create a prediction formula for predicting the end of the seasoning, and detect the end of the seasoning when performing the seasoning based on the prediction formula. A plasma processing device, comprising: a control device.

10. A processing container for storing the object to be processed,

An electrical measuring device provided in the processing container;

A control device connected to the electric measuring device and receiving a measurement data from the electric measuring device, wherein when performing a seasoning by supplying a test object to be processed into the processing container, After supplying the test object into the processing container, cooling the processing container, and then supplying the plurality of test objects into the processing container again, the test is performed. When performing a multivariate analysis using a multivariate analysis program on the basis of a plurality of measurement data measured by the output device to create a prediction formula for predicting the end of the seasoning, and performing the seasoning based on the prediction formula A control device for detecting the end of seasoning of the

A plasma processing apparatus comprising: