JPH06104214A - Plasma treatment device and its failure preventive and diagnostic method - Google Patents

Abstract

PURPOSE:To shorten a non-operating time of a device by predicting failure of an equipment by comparing the transition tendency of a change pattern with time of an article to be treated with that of a previously stored change pattern with time at a normal time. CONSTITUTION:The plasma of a reactant gas is formed into a treating chamber 3, in which an article to be treated 4 is housed, and aimed treatment is carried out to the article to be treated 4 by a reaction between the plasma and the article to be treated 4. The first patterns of changes with time of a plurality of observational data at a time when the article to be treated 4 is treated are monitored and stored while a plurality of the transition tendency of the first patterns of changes with time to a plurality of the articles to be treated 4 are monitored, and compared with the second pattern of a change with time of previously stored observational data at a normal time. A means 19 conducting the predicting operation of the failure of an object to be observed and the diagnostic operation of the presence of the abnormality of treatment is provided. Accordingly, the failure of an equipment can be predicted, thus shortening the nonoperating time of a device.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plasma processing apparatus and a failure preventive diagnosis method therefor, and more particularly to a technology applicable to plasma processing equipment used in a semiconductor manufacturing process.

[0002]

2. Description of the Related Art For example, a dry etching apparatus in a semiconductor manufacturing apparatus excites a reaction gas introduced into a processing chamber into a plasma by exciting it with a high frequency, and converts ions and radicals in the plasma into an electric field, a magnetic field, vacuum exhaust, etc. The driving force of (1) causes a reaction with the workpiece (wafer) to cause a reaction, and an etching process is performed.

In order to perform a highly accurate etching process using a dry etching apparatus, it is essential to control the high frequency, the pressure in the source processing chamber, the flow rate of the reaction gas, the electric field, the magnetic field that determines the driving force, and the vacuum exhaust generation device with high accuracy. For this purpose, technical development has been carried out for stably maintaining control parameters such as current, voltage of the high frequency source, pressure in the processing chamber, and reaction gas flow rate at predetermined set values.

On the other hand, for the abnormality detection of the equipment, for example, as described in Japanese Patent Laid-Open No. 59-16011, the deviation from the set value of the control parameter is monitored, or the abnormality signal from the equipment is detected. It was not possible to break away from the classical method of detecting equipment failure by monitoring.

[0005]

However, in order to meet the recent demands for higher precision and higher functionality of semiconductor devices, the manufacturing equipment becomes complicated, and once a failure occurs, it takes time to find the relevant part. In addition to the above, there is a problem in that the device that has been processed must be discarded when a failure occurs, resulting in a large loss in terms of cost and productivity.

Therefore, an object of the present invention is to provide a plasma processing apparatus capable of shortening the down time of the apparatus by predicting the failure of the equipment in advance, and a failure preventive diagnosis method.

[0007]

[Means for Solving the Problems] Means for solving the above problems according to the present invention will be described below.

That is, the plasma processing apparatus according to the present invention forms plasma of a reaction gas in the processing chamber in which the object to be processed is housed, and the reaction between the plasma and the object causes the object to be processed. The first time-dependent change pattern of a plurality of observation data when processing the object to be processed is monitored and stored, and a plurality of the first time-dependent changes with respect to the plurality of objects to be processed are performed. By monitoring the transition tendency of the change pattern and comparing it with the second time-dependent change pattern of the observation data stored in advance in the normal state, the predicting operation of the failure of the observed object and the diagnosing operation of the presence or absence of abnormality of the processing Is provided.

Further, the plasma processing apparatus according to the present invention is provided with means for performing a failure prediction operation by comparing the transition tendency of the first temporal change pattern of the single observation data with the second temporal change pattern. There is.

In the plasma processing apparatus of the present invention, the failure prediction operation is performed by using a plurality of first observation data pieces.
Means for performing a combinational comparison of the transition tendency of the time-dependent change pattern and the plurality of second time-dependent change patterns are provided.

[0011]

According to the plasma processing apparatus of the present invention, the first temporal change pattern of the plurality of observation data when processing the object to be processed is monitored and stored, and the plurality of the first to plural objects to be processed are stored. By monitoring the change tendency of the time-dependent change pattern of No. 1 and comparing it with the second time-dependent change pattern of the observation data stored in advance at the normal time, it is possible to predict the failure of the observation target and diagnose the abnormality of the processing. Since this can be done promptly, a failure of the device can be predicted in advance, and the downtime of the device can be shortened.

Further, according to the failure preventive diagnosis method of the plasma processing apparatus of the present invention, the first temporal change patterns of the plurality of observation data when the object to be processed is processed and stored, and the plurality of the objects to be processed are also stored. Predicting the failure of the observed object by monitoring the transition tendency of the plurality of first time-dependent change patterns for the processed object and comparing it with the second time-dependent change pattern of the observation data stored in advance in the normal state. Since the abnormality of the process can be quickly diagnosed, the failure of the device can be predicted in advance, and the down time of the device can be shortened.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 is a block diagram showing an apparatus configuration for realizing a plasma processing apparatus and a failure preventive diagnosis method according to an embodiment of the present invention. Each element of the device configuration shown in FIG. 1 will be described.

A processing chamber 3 is constituted by a container 1 and a discharge tube 2 made of transparent quartz. An object 4 to be processed is placed on the lower surface of the container 1, and a high frequency power from a high frequency power source 5 is introduced into the processing chamber 3. A sample table 6 that serves as an electrode is installed.

A part of the wall of the container 1 is connected to a vacuum pump 9 via an exhaust pipe 7 and a pressure control means 8, and the processing chamber 3
The internal pressure can be brought to a desired state.

Further, the container 1 is provided with a gas introduction passage 10 for introducing a treatment gas into the treatment chamber 3, and the gas introduction passage 10 is provided with a gas source 12 via a flow rate control means 11.
It is connected to the.

In the processing chamber 3, microwaves output from the magnetron 13 are supplied to the power monitor 14 and the waveguide 1.
5, introduced through the discharge tube 2. A coil 16 for generating a magnetic field for causing electron cyclotron resonance with respect to the microwave introduced into the processing chamber 3 is provided around the waveguide 15. Here, the power monitor 14
It measures the incident power and reflected power of microwaves.

The microwaves that have caused electron cyclotron resonance in the processing chamber 3 turn the processing gas introduced into the processing chamber 3 into plasma, and etch the object 4 to be processed placed on the sample table 6. Perform processing.

A plasma detector 17 for detecting an emission spectrum from plasma generated in the processing chamber 3 is provided near the side surface of the discharge tube 2 made of transparent quartz or the like.

The end point determining means 18 detects the change in the emission spectrum intensity of a specific substance in plasma by the plasma detector 17
The end point of the etching process for the object to be processed 4 is detected by measuring based on the signal from the.

Further, the pressure control means 8 and the flow rate control means 1
1, the power monitor 14, and the end point determination means 18 are connected to the failure prevention diagnosis means 19. Failure preventive diagnosis means 19
Is composed of a control unit 20, an analog-digital conversion unit 21, a storage unit 22, a display unit 23 and the like. The failure prevention / diagnosis means 19 fetches a status monitor signal for the control parameter output from the pressure control means 8, the flow rate control means 11, the power monitor 14, and the end point determination means 18 via the analog / digital conversion section 21, and stores it in the storage section 22. To do. The control unit 20 performs the failure prevention diagnosis process using the data in the storage unit 22 and displays the result on the display unit 23.

An example of the operation of the plasma processing apparatus shown in this embodiment will be described below.

First, the object 4 to be processed is placed on the sample table 6 by a transfer means from outside the apparatus (not shown). After the processing chamber 3 is sealed, the vacuum pump 9 is activated to evacuate the processing chamber 3. At this time, the pressure control means 8 is used to process the processing chamber 3
Measure and control the degree of vacuum inside. Further, the vacuum degree observation signal in the processing chamber 3 from the pressure control means 8 during the exhaust processing is input to the failure prevention diagnosis means 19. After starting the exhaust, after confirming that the pressure in the processing chamber 3 has reached a predetermined value, the processing gas from the gas source 12 is controlled by the flow rate control means 11 so as to have a predetermined flow rate, and the gas introduction path is controlled. Processing chamber 3 through 10
Coil 16 and magnetron 1
3, the microwave is introduced into the processing chamber 3 through the power monitor 14, the waveguide 15, and the discharge tube 2. Further, the gas flow rate observation signal from the flow rate control means 11 and the microwave incident power observation signal from the power monitor 14 are input to the failure prevention diagnosis means 19.

As described above, the processing gas and the microwave are introduced into the processing chamber 3 to generate plasma in the processing chamber 3. After the plasma is generated, the high frequency power supply 5 is activated to apply a high frequency to the sample stage 6. As a result, the reaction between the predetermined substance of the processing gas excited in the plasma state and the object 4 to be processed is promoted, and the etching process for the thin film formed on the object 4 to be processed proceeds. At this time, the emission spectrum signal of the specific substance emitted from the plasma, which is measured by the end point determination means 18 via the plasma detection source 17, is input to the failure prevention diagnosis means 19.

The operation of the failure preventive diagnosis means 19 will be described with reference to FIG. 2 and subsequent figures.

The failure prevention / diagnosis means 19 stores in the storage unit 22 a temporal change pattern P1 (n) of control parameters and monitor information for each object 4 to be processed (an temporal change pattern for the nth object 4 to be processed). . In addition, the control unit 20
The transitional state from the aging pattern P1 (n-m) to P1 (n) is monitored.

FIG. 2 shows the processing chamber 3 as the processing time elapses.
It is a graph which shows an example of the internal pressure change. In FIG. 2, P2 is a pressure change pattern under normal conditions, and P1
(N), P1 (n-1), and P1 (n-2) are patterns of pressure change in the nth, n-1, and n-2th processes, respectively. Further, T1 indicates a certain time after the etching process is started, and T2, T2 (n) T2 (n-1) T
2 (n-2) indicates the time until the pressure in the processing chamber 3 reaches a specified value at each etching processing number. Further, T2ER is the minimum value of the time until the pressure in the processing chamber 3 reaches the specified value when it is considered as a failure.

Although there are several methods for estimating a failure in the near future using the above data, the method using time will be described here. That is, the time taken for the pressure in the processing chamber 3 to reach the specified value at each number of etching processes starting from T2 is t (n) = T2 (n) -T.
2, t (n-1) = T2 (n-1) -T2, t (n-
2) = T2 (n−2) −T2, and t (m) = F (m)
(For example, F (m) is a quadratic function). t
F (m) using (n), t (n-1), and t (n-2)
Ask for.

Since it is the number of processing times N that the time until the pressure in the processing chamber 3 reaches the specified value is t2er = T2ER-T2, N = f (t2er) (f is
N is obtained as the inverse function of F). When the difference between the number of times of processing n and N becomes equal to or less than a predetermined value, the control unit 20 displays the fact on the display unit 23. Further, the control unit 20 may be provided with an interface function with a host computer (not shown) to transmit the result of the failure prevention diagnosis.

FIG. 3 is a graph showing an example of changes in the flow rate of the gas passing through the gas introducing passage 10 with the lapse of processing time. In FIG. 3, P2 is a pattern of gas flow rate change under normal conditions, and P1 (n), P1 (n-1), P1
(N-2) is a gas flow rate change pattern in the nth, n-1, and n-2th processes, respectively. Also, T
1 and T2 indicate a fixed time after the start of the etching process.
Here, T1 of the gas flow rate at each number of etching processes
To the average value from T2 to G2, G1 (n), G1 (n-
1) and G1 (n-2).

A method of estimating a near future failure using the above data will be described.

That is, the difference between the average value of the gas flow rate and G2 at each etching process is g (n) = G2-G2.
(N), g (n-1) = G2-G2 (n-1), g (n
-2) = G2 (n-2) -G2, and g (m) = F
(M) (for example, F (m) is a quadratic function). g
F (m) using (n), g (n-1), and g (n-2)
Ask for.

Since it is the number of processing times N that the average value of the gas flow rate falls below the specified value GER, it is desired to obtain N
= F (G2-GER) (f is an inverse function of F) is obtained. When the difference between the number of times of processing n and N becomes equal to or less than a predetermined value, the control unit 20 displays the fact on the display unit 23. Further, the control unit 20 may be provided with an interface function with a host computer (not shown) to transmit the result of the failure prevention diagnosis.

FIG. 4 is a graph showing an example of changes in the amount of light received by the plasma detector 17 with the lapse of processing time. In FIG. 4, P2 is a pattern of changes in the amount of received light under normal conditions, and P1 (n), P1 (n-1), and P1 (n-2).
Are patterns of changes in the amount of received light in the nth, n-1, and n-2th processes, respectively. Further, T1 and T2 indicate a fixed time after the start of the etching process. Here, the average value of the amount of received light from each of the number of etching processes from T1 to T2 is H2, H1 (n), H1 (n-1), H1 (n-
2).

A method for estimating a near future failure using the above data will be described. That is, the difference between the average value of the amount of received light and H2 at each etching process number is h (n) = H
2-H2 (n), h (n-1) = H2-H2 (n-
1), h (n-2) = H2 (n-2) -H2, and h
It is assumed that (m) = F (m) (for example, F (m) is a quadratic function).

F (m) is obtained using h (n), h (n-1) and h (n-2). What I want to obtain now is the number of processing times N when the average value of the amount of received light falls below the specified value HER, so N = f (H2-HER) (f is an inverse function of F)
Asking for N. When the difference between the number of times of processing n and N becomes equal to or less than a predetermined value, the control unit 20 displays the fact on the display unit 23. Further, the control unit 20 may be provided with an interface function with a host computer (not shown) to transmit the result of the failure prevention diagnosis.

Further, a plurality of time-dependent change patterns, for example, the product of the pattern shown in FIG. 2 and the pattern shown in FIG. The state transition can be detected over the observation items of and the failure occurrence in the near future can be estimated.

[0039]

As described above, according to the plasma processing apparatus of the present invention, a change in the state of the apparatus can be detected before a failure occurs, and the apparatus can be repaired early. The downtime of the equipment can be shortened, and
Since it is possible to avoid discarding the object to be processed due to a failure, productivity can be improved.

Further, according to the failure preventive diagnosis apparatus of the present invention, since the state change of the apparatus can be detected before the failure occurs and the failure occurrence time can be estimated, the planned maintenance work can be performed. Since the downtime of the device can be shortened, the productivity can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram of a plasma processing apparatus and failure preventive diagnosis means according to the present invention.

FIG. 2 is a graph showing an example of a change in pressure inside a processing chamber with the lapse of processing time.

FIG. 3 is a graph showing an example of changes in gas flow rate with the lapse of processing time.

FIG. 4 is a graph showing an example of changes in the amount of received light with the passage of processing time.

[Explanation of reference numerals] 1 ... Vessel, 2 ... Discharge tube, 3 ... Processing chamber, 4 ... Object to be treated, 5
... high frequency power source, 6 ... sample stage, 7 ... exhaust pipe, 8 ... pressure control means, 9 ... vacuum pump, 10 ... gas introduction path, 11 ... flow rate control means, 12 ... gas source, 13 ... magnetron, 14 ...
Power monitor, 15 ... Waveguide, 16 ... Coil, 17 ... Plasma detector, 18 ... End point determination means, 19 ... Failure prevention diagnosis means, 20 ... Control section, 21 ... Analog-digital conversion section, 22 ... Storage section, 23 ... Display.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Akitaka Makino 794 Azuma Higashitoyoi, Kudamatsu City, Yamaguchi Prefecture Stock Company Hitachi Ltd. Kasado Factory

Claims (4)

[Claims] 1. A plasma processing apparatus in which plasma of a reaction gas is formed in a processing chamber containing an object to be processed, and a target process is performed on the object by a reaction between the plasma and the object. , Monitoring and storing a first temporal change pattern of a plurality of observation data when processing the object to be processed, and monitoring a transition tendency of the plurality of first temporal change patterns with respect to the object to be processed. A means for performing a predicting operation for a failure of an observation target and a diagnosing operation for the presence / absence of an abnormality in the processing by comparing with a second temporal change pattern of the observation data in a normal state stored in advance, Characteristic plasma processing device. 2. A plasma processing apparatus in which plasma of a reaction gas is formed in a processing chamber accommodating an object to be processed, and the object processing is performed on the object by a reaction between the plasma and the object. , Monitoring and storing a first temporal change pattern of a plurality of observation data when processing the object to be processed, and monitoring a transition tendency of the plurality of first temporal change patterns with respect to the object to be processed. A means for performing a predicting operation for a failure of an observation target and a diagnosing operation for the presence / absence of an abnormality in the processing by comparing with a second temporal change pattern of the observation data in a normal state stored in advance, A method for preventing and diagnosing a failure in a plasma processing apparatus. 3. The failure preventive diagnosis of the plasma processing apparatus according to claim 2, wherein the failure prediction operation is performed by comparing the transition tendency of the first temporal change pattern of the single observation data with the second temporal change pattern. Method. 4. The plasma processing apparatus according to claim 2, wherein the failure prediction operation is performed by comparing the transition tendency of the plurality of first time-dependent change patterns of the plurality of observation data and the plurality of second time-dependent change patterns. Failure Prevention Diagnosis Method