JPH07258853A - Method and device for discriminating state of process - Google Patents

Abstract

PURPOSE: To provide a method for identifying the state of a plasma process and an apparatus therefor.

CONSTITUTION: A plasma treatment apparatus 10 for production of semiconductor wafers is disclosed. This apparatus includes a plasma treatment tool 12 and an RF energy source 20 coupled to this plasma treatment tool 12. An arbitrarily selective matching circuit net 22 can be included between the RF energy source 20 and the plasma treatment tool 12. A circuit 18 for obtaining a measurement characteristic by monitoring RF energy is also provided. At least one transducer 14 or 16 is coupled between the plasma treatment tool 12 and the circuit 18 for monitoring the RF energy. The RF energy is usually impressed at a basic frequency and the electrical characteristic is monitored at the second frequency different from this basic frequency. Further, a circuit 19 of a computer or the like for detaining the state of the treatment apparatus 10 by interpreting the measurement characteristic is also included.

COPYRIGHT: (C)1995,JPO

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION The present invention relates to a method and apparatus for identifying process and machine conditions, especially in plasma processors.

[0002]

2. Description of the Related Art Integrated circuit manufacturing processes using reactive plasma are commonplace in today's semiconductor manufacturing lines. The term reactive plasma is meant to describe an electrical discharge in which gas ionization and dissociation occurs, producing a class of chemically active substances, often oxidizing and / or reducing agents. Such plasmas are reactive both in the gas phase and on the solid surfaces exposed to them.
When reactive plasmas were first used in semiconductor factories in the early 1970s, little was known about the chemistry and physics of these discharges. Today, we have a broad base of empirical knowledge, some qualitative understanding of the mechanism of activation, and even a detailed understanding of certain segregated phenomena. With this knowledge, plasma processes can now be designed to meet the more stringent requirements of today's submicron device structures.

When using plasma processing, many problems can occur that limit yield or reduce manufacturing throughput. For example, water vapor entering a plasma processing chamber can adversely affect the manufacture of semiconductor wafers. This problem requires prompt detection, without which there is a high risk of yield degradation.

Another problem with plasma processing is the formation of contaminants on the chamber walls. Nowadays, in a typical processing plant, the chamber wall is
For example, every time a certain number of wafers are processed, they must be cleaned periodically. This technique presents some great difficulties if the rate of pollutant formation is unknown, but it is usually unknown. For example, if the chamber walls are not cleaned frequently enough, some of the processed wafers may become contaminated at a later stage, causing yield degradation. On the other hand, if the chamber is cleaned too often, a loss of throughput will occur. Therefore, there is currently an inherent need for a real-time detector of chamber wall cleanliness.

[0005]

Accordingly, improvements are currently desired that overcome any or all of these problems. Other purposes and advantages are self-explanatory,
Some of which appear below and are accomplished by the present invention which provides a method and apparatus for identifying process / machine conditions.

[0006]

SUMMARY OF THE INVENTION One feature of the present invention is to optimize the chamber conditions that can be implemented within a process control scheme to prevent chamber condition variations and increase wafer yields. It provides a way to identify. At present, there is essentially no technology that comprehensively identifies the state of the machine, such as the state of the chamber walls and electrodes. This flaw creates many problems.
The lack of native sensor measurements requires the use of pilots or test wafers to determine if the chamber conditions are correct for material processing. This additional work reduces throughput and increases operating costs. The lack of natural sensors further precludes the use of controls to maintain optimum chamber conditions. The ability of SQC (Statistical Quality Control) is also diminished in the absence of direct machine measurements. The reason is
This is because the pilot is affected by the previous treatment and does not completely show the problem of the machine. Further, if the state of the chamber cannot be identified, the chamber cannot be properly cleaned. Excessive cleaning reduces throughput, while incomplete cleaning results in reduced wafer yield due to particles and altered process conditions. In addition, accidental unnoticed changes in chamber conditions, such as exposure to water vapor, can lead to product disposal as scrap.

Disclosed herein is a plasma processing apparatus for semiconductor wafer manufacturing that overcomes many of the problems. In an embodiment, the apparatus includes a plasma treatment tool and an R coupled to the plasma treatment tool.
RF energy source for applying F energy.
An optional matching network may be included between the RF energy source and the plasma processing tool. An RF energy monitoring circuit for obtaining the measurement characteristic is also provided. At least one transducer includes a plasma processing tool and an R
It is coupled to the F energy monitoring circuit.
RF energy is typically applied at the fundamental frequency, and electrical properties are monitored at the relevant frequency, which is different from the fundamental frequency, and the fundamental frequency. Also included is a circuit for interpreting the measurement, such as a computer, for determining the state of the processor.

Also disclosed herein is a method of determining the state of the processor and controlling as much as possible. A processing apparatus is provided that includes an electromagnetic power source and a processing chamber. Electromagnetic energy, for example RF power, is applied to the processing device and at least one electrical characteristic of the electromagnetic energy is monitored to obtain a measured characteristic. The electrical characteristics are power, phase,
One or all of RMS voltage, RMS current, peak-peak voltage, peak-peak current, and impedance can be measured over time. Normally, power is applied at the fundamental frequency and electrical properties are monitored at the relevant frequency. At that time, a measurement of the measurement characteristic is performed in order to determine the state of the processing device.

An advantage of the present invention is that the method is not affected by non-critical variables such as dirty optics. Moreover, this application of the sensor serves as a direct measurement of chamber and process conditions. Furthermore, the present invention provides a true, real-time method. Therefore, the present invention can be used for real-time control of processing steps. The methods and apparatus described herein may also serve as a final assembly inspection tool.

The above features of the present invention will be more clearly understood upon consideration of the following description in conjunction with the accompanying drawings. The same reference numbers and symbols in different figures refer to the same parts unless otherwise noted.

[0011]

EXAMPLES The construction and use of the examples employed herein will be described in detail below. However, it should be appreciated that the present invention provides a number of applicable inventive concepts that may be embodied in a wide variety of specific contexts. The particular embodiments described herein are merely illustrative of particular ways of making and using the device of the invention, and do not define the scope of the invention.

The apparatus and method of the present invention are described below. First, the problem solved by the present invention will be briefly described. Next, examples are described, followed by modifications. Next, a summary of some experimental results is explained.

The present invention proposes a method and apparatus for observing plasma conditions by means of current and voltage characteristics in order to identify and / or evaluate the condition of the machine. In this patent, the state of the machine includes items such as wall material or formations on the wall, gas absorbed on the wall, electrode material and configuration, electrode spacing, type of power supply, etc. Be done. Process conditions include power supply, plasma composition,
Items such as pressure and transient (temporary) response of conditions are included. Evaluation of these conditions can be used to identify optimal machine conditions. As a result, it is possible to change the conditions of the chamber, for example by cleaning or adjusting, and thereby achieve optimum mechanical conditions. It is proposed to use machine state identification in an automated feedback control loop so that chamber changes can be made efficiently. Finally, the sensor output can be used in an SQC (Statistical Quality Control) environment or to determine when chamber conditions have changed.

In one example, solution studies have shown that changes in the state of the chamber appear in harmonics, but not in the fundamental electrical signal. For example, pressure problems, regular maintenance, and atmospheric exposure of the chamber affect the phase of the harmonic signal and can therefore be detected. On the other hand, the phase of the fundamental signal is not affected by these same effects.

As is known, plasma cleaning of electromechanical components, optical surfaces, and aerospace components includes
Widely used, because it is particularly effective in removing the last traces of organic pollution.
In addition, the effect of surface contamination on the measured ellipsometry parameters was investigated, and it was shown that the ellipsometry measurements were a function of plasma cleaning time. However, these studies do not deal with RF electrical signals (at the fundamental frequency or other frequencies).

Optical emission spectroscopy can be used to monitor the signals from the projectiles within the plasma. When chemometrics are applied to the optical emission spectra obtained from plasma processes, a comprehensive empirical mapping of plasma processing conditions to optical signals is obtained. If a broad emission spectrum is examined using a multivariate statistical algorithm (chemometrics) and the spatial dependence of the emission is taken into account, the change in light emission intensity correlates close to 1 with probe measurements at the fundamental frequency. It has been reported that The problem with this method is that the plasma state must emit light. These emissive materials are not directly affected by the constituent effects of the machine like RF signals. This is because the condition of the chamber walls affects the ability of the plasma to impinge on the wafer, but the optical signal is only sensitive to the firing from the material within the plasma volume. These wall effects disrupt the correlation of the optical signal to the chamber state. Furthermore, the optical signal is affected by the dirty optical device, which does not affect the process.

In addition, the number of wafers that can be processed until a plasma cleaning process is required is determined by monitoring the particle level associated with the process.
When the particle count is increased to an unacceptable level of yield, dry cleaning is performed. The number of wafers processed during successive dry cleanings is therefore fixed to keep the particle level below a certain threshold. However, an improper cleaning process is usually only noticed by a reduction in product yield. Unfortunately, the product yield is
It does not have a good correlation with the particle levels associated with the process. Therefore, when yields decrease, corrective action, such as dry cleaning, is performed until yields increase to an acceptable level.

After the plasma cleaning process, eg dry cleaning, a high light emission intensity is observed until several wafers have been processed. These first few wafers put the chamber in optimal operating condition,
Break it in and get it back.

Since dry cleaning is performed throughout the chamber, the chamber is in a condition of increased strength, allowing better coupling of RF power and brighter plasma rather than clean optics. This confuses the use of optical radiation to indicate whether the chamber conditions are acceptable for product processing.

In addition, wet cleaning is usually performed when the optical signal has deteriorated to a state where it is not useful for indicating the end point of the process. Wet cleaning is typically required to restore the optical signal to allow process endpoint indication before the particle count is high enough to reduce yield.

In addition, a DC bias signal may be used to indicate acceptable machine conditions for product processing. The plasma cleaning process is performed until the DC bias is at a level that maps to an acceptable mechanical state. It has been found that the effect of the DC bias signal is the same at different machine conditions for the same process settings. Therefore, the correlation to acceptable machine conditions becomes confusing. It is further desirable to produce and control a DC bias. If DC
If the bias is part of the control loop, it cannot be used for monitoring.

FIG. 1 shows a general block diagram of the device 10 of the present invention. The apparatus includes a processing tool 12 to which voltage and current transducers 14 and 16 for making electromagnetic characteristic measurements are coupled. The outputs of the transducers 14 and 16 are coupled to the electronic device 18 and the electronic device 1
8 includes circuitry for determining the electrical characteristics of the electromagnetic signal. Transducer 14 or transducer 16
It should be noted that either or both can be used. Electromagnetic power is, for example, 13.56 MHz RF
It is supplied from a power supply 20, which may be a power supply. This power is then coupled into the matching network 22. Matching network 22 is optional and may be used if there is an impedance mismatch between power supply 20 and processing tool 12. Matching network 22 is not necessary in the present invention.

In general, the apparatus 10 can be utilized for all processes, including the application of electromagnetic power sources to electrodes, such as plasma etching or deposition equipment. The device 10 may also be used for monitoring and controlling plasma cleaning machines.

The processing tool 12 may include a capacitively coupled plasma etching tool, such as a parallel plate apparatus. In this device, the semiconductor wafer being processed is grounded or receives an RF signal. The device can be, for example, a reactive ion etching device.

The present invention also applies to high density plasma devices. Examples of these devices are helicon
on) or inductively coupled or transformer coupled processing device with or without a magnetron device such as a spiral wave generator and an ECR (electron cyclotron resonance) device. The device may optionally include a high density plasma device. An example of a high density plasma device is a magnetron device.

Transducers 14 and 16 are included to provide voltage and current information to the electronic device. The voltage probe is a capacitively coupled voltage transducer. The current probe is a transformer coupled current transducer. Both the current and voltage probes are housed together in a housing. The housing assembly is arranged in-line and is configured to maintain the coaxiality of power transmission.

Transducers 14 and 16 are coupled to electronic circuitry 18. The electronic circuit 18 operates to obtain a measurement characteristic signal from the electrical measurement.
The electronic circuit 18 sends the measurement signal to a circuit 19 for interpreting the signal and determining the state of the processing unit. In one example, the circuit 19 for interpreting the measurement characteristic is a personal computer (or other computer terminal, workstation or mainframe) or other signal recording hardware. The electronic circuit 18 may typically include electrical components such as filters, multipliers, transducers, digitizers, and other circuitry. The output of the electronic circuit 18 can be digital, analog, or a combination of both. Impedance, power, phase,
Measurement information such as RMS voltage, RMS current, peak-peak voltage, and peak-peak current may be measured over time to determine characteristics of the signal, such as waveform. This information is typically measured at the fundamental or drive frequency (eg, power supply frequency) and the harmonic or subharmonic frequencies. An example of a typical measurement circuit 18 is shown in FIG.

In FIG. 2, transducer 15 (which represents one or both of transducers 14 and 16) is coupled to the input of measurement circuit 18. The voltage and current inputs are coupled to the filter 40. Filter 40 is typically a notch filter for each of the current and voltage signals. This filter is centered on the fundamental frequency (eg 13.56 MHz). Filtered current output 42 is coupled to a first power divider 46, while filtered voltage output 44 is coupled to a second power divider. One output of each of the power dividers 46 and 48 is coupled to an integrator circuit 50 and a Fourier transform (eg, fast Fourier transform) circuit 52. The integrating circuit 50 and the Fourier transform circuit 52 are formed on separate printed wiring boards. The outputs of the integrator circuit 50 and the Fourier transform circuit 52 are coupled to a signal conditioning and routing circuit 54. The signal conditioning and routing circuit 54 prepares the signals representing the items from the set of measurements. Examples of these items include RMS voltage, RMS current, power, peak-peak voltage, and peak-peak current. These quantities can be given at different frequencies.

The circuit 18 operates as follows. Notch filter 40 and power dividers 46 and 48 are used to precondition the signal. The preconditioned signal is sent to the next integrator circuit 50 and Fourier transform circuit 52. The integrator circuit 50 integrates the product of the voltage signal and the current signal to obtain the transducer 15
Get a measure of the power passing through. Fourier transform circuit 52
Performs a Fourier transform on the voltage signal and the current signal to obtain a measure of the frequency spectrum of these signals. The signal conditioning and routing circuit 54 provides scaling and selection of the signal to be used as an input for the applications disclosed herein. The selection is done because not all possible signals are needed for each application.

The output of the measurement circuit 18 is coupled to a circuit 19 which provides an interpretation of the measurement characteristics, which circuit 19 can be a computer or other signal recording hardware, as described above. The computer 19 can use the signals as inputs for control algorithms, diagnostic decisions, and process monitoring.

The measurement information can be interpreted separately in interpretation or in combination. In other words, the measurement information can be used in univariate or multivariate analysis. For example, the signals can be combined and correlated using the T ² technique.

In analyzing frequency spectra, fundamental frequencies, harmonic frequencies (ie, integer multiples of the fundamental frequency), subharmonic frequencies (ie, frequencies below the fundamental frequency), and other frequencies (generated by non-linearities in the device) Possible) can be analyzed. These frequencies can be transient or steady state.

In interpreting the measurement data, the characteristics of the entire device can be analyzed. For example, the device may include a power source 20,
It may include the matching network 22, the processing chamber 12, and the wafer being manufactured (not shown in FIG. 1). It is the function of the electronic device in the other circuit that determines the meaning of each of these measured quantities.

The method and apparatus of the present invention can be used for many purposes. For example, the device may be used for checking diagnostic measurements such as pressure problems, water vapor leaks into the chamber, power supply problems, or matching network problems. The device can also be used for regularly scheduled maintenance, such as cleaning the chamber. Furthermore, the apparatus can also be used for monitoring and / or controlling process conditions such as etching and / or deposition. For example, the method and apparatus can be used to check process step uniformity and endpoint measurements. Of course, when making uniformity measurements, one must consider the uniformity of the starting material.

The device is also used to monitor the process and to verify that a problem may have occurred. When a meaningful change occurs, a problem may have occurred because the state of the process has changed. At this stage, the process is stopped and the diagnosis is started.

Although primarily described herein with respect to semiconductor processing, the present invention may also be used in numerous plasma applications. Examples of non-semiconductor applications include processes for coating medical devices or mirrors. In particular, the present invention is extremely useful in plasma polymerization processes.

It should be noted that different machines usually have different measurement characteristics. That is, different electrical characteristics should be interpreted as determining different things.
These measured properties are usually determined empirically. In the future, common laws will be discovered that determine common trends in different processing units, and will therefore require less empirical research for each particular processing machine or machine type.

FIG. 3 shows in detail the processing chamber capacitively coupled to the parallel plates. As shown in this figure, wafer 224 is mounted on chuck 226. The chuck 226 is electrically grounded or receives a signal. The second chuck 228 is spaced from the wafer 224 and forms the second electrode of the device. If not shown, the second electrode is provided by the chamber wall and thus the second chuck 228 is unnecessary. The second chuck 228 may also be electrically grounded or may receive a signal. (Of course, if plasma is generated, both electrodes cannot be grounded.)

As shown in FIG. 3, each electrode 226 and 228 is connected to a transducer 216 or 214 for measuring electrical properties. As described above, the transducer 214 or 21
Only one of 6 needs to be included. Also shown in FIG. 3 is an optional magnet 230, which is E
It is used to generate a CR (electron cyclotron resonance) mode.

Referring now to FIG. 4, the inductively coupled processing chamber is shown in detail. In this case, the wafer 32
8 is a chuck 3 connected to the transducer 316.
Mounted on 26. The inductive coil 332 is provided to generate plasma for processing. Transducer 314 is placed in series with inductive coil 232 to measure RF electrical characteristics.

Although illustrated in two specific examples, the present invention
It should be appreciated that it applies to all processes including applying an electromagnetic power source to the electrodes. Other examples include antenna impedance monitoring and RF power quality inspection.

The method and apparatus of the present invention have been experimentally tested in a standard advanced vacuum processor (AVP), a single wafer processing apparatus configured in a reactive ion etching (RIE) mode with a wide electrode gap. It was A light emitting endpoint scheme may be used. Current and voltage probes are used at the electrode connection locations of the power supply / matching network, and the outputs from these probes feed a network analyzer such as the HP4195A automatic network analyzer. This analyzer uses a discrete table to measure the amplitude and phase of the frequencies listed in that table.

The core composite design consists of the following in the following areas:
Executed by variables: about 100 mtorr and 200 m
pressure between torr, about 300W to 500W
RF power between, between about 30 sccm and 50 sccm
CHF₃, CF between about 40 sccm and 80 sccm
_Four, And between about 10 sccm and 20 sccm
O ₂. Argon was kept constant at about 100 sccm,
The pressure of the helium chuck was about 2.5 torr.
The temperature of the substrate chuck was kept at about 0 ° C.

In this case, the experiment was carried out by (on blanket) oxide on silicon nitride on polysilicon,
Nitride on oxide (on blanket) on silicon,
And (patterned) nitride on polysilicon on oxide on silicon. Wafer cleaning was performed approximately every 2 wafers.

To be referred to in later data analysis,
The phase at a frequency is the phase measured for voltage minus the phase measured for current. The power (W _i ) is defined as:

[0046]

Where W _i = I _i ^* V _i ^* cos θ _i where I _i is the current at frequency i, V _i is the voltage at frequency i, W _i is the power at frequency i, and θ _i Is the phase at frequency i. The impedance (Z _i ) is defined as follows.

[0047]

[Equation 2] The quantities I _i ^* V _i and V _i / I _i are also examined.

No calibration method was used in this particular example. The calibration for amplitude should be constant and is therefore absorbed by the coefficients that fit the model. However, the phase is usually more difficult to calibrate. The phase calibration is added to the measured phase. Since the power and impedance equations include the cosine of the phase, the effect of uncalibrated phase data on power and impedance is not clear.

An example of the response as a function of the number of processes is shown in FIG. 5a for the fundamental frequency (13.56 MHz in this case).
, And for the first two harmonics (in this case 27 MHz and 40 MHz) are shown in FIG. 5b. Figure 5
It should be noted that in a and FIG. 5b, each successive pair was treated in the same state, eg treatment numbers 5 and 6 were in the same experimental state. In Figures 5a and 5b, the change in response after exposure of the chamber was not the same for all exposures. This difference is due both to the time the pump is stopped (eg, overnight vs. 1 hour) and the confusion with pressure problems.

Based on the response of all signals, it is possible to get an indication of which signal should be monitored for a given problem. For example, pressure problems were monitored by measuring one or more of the following: 1) 2
Impedance (or V _i / at 7 MHz and / or 40 MHz (ie first harmonic))
I _i ), 2) Phase at 40 MHz, and 3) 27
Phase at MHz and / or 54 MHz (minor). The exposure of the chamber to air is dependent on the phase at 27 MHz and / or 40 MHz and possibly 40
It was monitored by measuring the impedance (or V _i / I _i ) at MHz.

Although this data is for blanket wafers only, the conclusions are also valid for patterned wafers where exposure to the atmosphere in the chamber near the end of the experiment was the only accident. is there.
The signal at 40 MHz is more sensitive (ie, shows a larger change) than the signal at 27 MHz. Due to the noise level due to the narrow bandwidth of the circuit analyzer,
It was difficult to come to a conclusion about harmonics. However, previous work with analog spectrum analyzers has shown that 50 MH
It is suggested that z is extremely sensitive to changes. In addition, in this previous study, a "semi-harmonic" was found to indicate process stabilization. Although these signals were not investigated in this particular study, they too may be important for diagnosis.

It appears that diagnostic information has never been obtained from a signal at 13.56 MHz (ie the fundamental frequency or power supply frequency). Information is harmonic (or subharmonic)
Can be obtained from, but not from the fundamental frequency,
I don't understand enough.

Another function of the method and apparatus of the present invention is to determine the functional form of the relationship of supply power to power setpoint and other processing conditions. Small changes in process flow and composition do not result in significant changes in power supply compared to natural variations due to errors in the same experimental iteration. However, no large compositional or flow changes that affect the signal were induced in this experiment. It has been shown that the gain, defined as the change in response to a change in the setpoint of supply power, is a function of pressure. That is, the pressure affects both the magnitude of the supplied power and the gain at the set point of the given power.

The present invention can also be used to design etch rates using RF signals. Attempts have been made to design the etch rate as a function of its RF signal in order to determine if the signal is indicative of the etch rate. Stepwise regression was used to select key variables and incorporate them into the design, and some "goodness of fit" tests were performed. 0.8% with a standard error of evaluation of 9%, using only data from the fundamental frequency (13.56 MHz)
A prepared R ^{2 of} 0 was obtained as a match for the etch rate. By repeating the same experiment, the theoretical limit of the standard error of evaluation was 6%. Incorporating harmonic measurements into the design increased the fit slightly.

In conclusion, in a particular set of experiments,
RF signals were collected for experiments using the nitride etcher. The results of this study can be used to assess the utility of RF signal monitoring and possibly control. The RF power delivered to the electrodes has been found to be a function of power and pressure. Therefore, if either the power supply or the pressure controller causes a problem, it is difficult to diagnose the source of the problem. Furthermore, pressure and power are confused when performing RSM (Response Surface Method) using the power source settling point. Since the etch rate is known to be a function of current and voltage, rather than power, the implementation of a controller for holding voltage and voltage at specific values seems justified.

It has also been found that the phase and impedance do not respond to changes in the processing conditions that occur in this experiment, only to changes in the mechanical conditions. According to the analysis,
The chamber walls / electrodes appear to be the major factors determining the measured phase and impedance. This observation suggests that phase and / or impedance are indicators of changes in chamber conditions, such as polymer deposition or water vapor adsorption. If so, a control loop for pump down time or cleaning step time could be realized.

Although the present invention has been described above with reference to examples, this description should not be construed in a limiting sense. Various modifications and combinations of the embodiments, and other embodiments of the present invention should be apparent to those skilled in the art with reference to this description. Accordingly, the claims are intended to cover any such modifications or examples.

With respect to the above description, the following items will be further disclosed. (1) Measured by installing a processing apparatus including an electromagnetic power source and a processing chamber, applying electromagnetic energy to the processing apparatus at a fundamental frequency, and monitoring at least one electrical characteristic of the electromagnetic energy. A step of obtaining a characteristic, wherein the power is applied at the fundamental frequency and the electrical characteristic is monitored at at least one associated frequency different from the fundamental frequency. And a step of determining the state of the processing device by interpreting the measurement characteristic, the method of determining the state of the processing device.

(2) The method of claim 1, wherein the measured characteristic indicates information about impurities in the processing chamber. (3) The method of claim 1, wherein the measured characteristic indicates information about the pressure in the processing chamber or its control.

(4) A step of disposing a wafer having a first material formed thereon in the processing chamber, and a step of etching the first material, wherein the measurement characteristic is that of the first material. An etching step of the first material showing information about etching rate or uniformity,
The method of claim 1, further comprising:

(5) disposing a wafer in the processing chamber and depositing a first material on the surface of the wafer, wherein the measured characteristic is a deposition rate or a first deposition rate of the first material. The method of claim 1, further comprising the step of depositing the first material, the step of indicating information about characteristics.

(6) Fifth, wherein the information includes end point information
Method described in section. (7) The method of claim 1, further comprising cleaning the processing chamber, the cleaning being controlled at least in part based on the measured characteristic.

(8) A method according to claim 7, wherein the cleaning step includes a step of keeping the processing chamber in a substantially vacuum state. (9) The method of claim 1, wherein the processing chamber comprises a plasma processing chamber.

(10) The method according to claim 1, wherein the electrical characteristic is selected from the group consisting of power, phase, effective voltage, effective current, peak peak voltage, peak peak current, and impedance. (11) The method of claim 1, further comprising the step of monitoring the at least one electrical characteristic at the fundamental frequency.

(12) The method according to item 1, wherein the measurement characteristic is obtained by monitoring a combination of signals. (13) A method according to item 12, wherein the combination of signals is measured at a single frequency.

(14) A processing apparatus comprising a processing tool, an electromagnetic energy source coupled to the processing tool for applying electromagnetic energy to the processing tool at a fundamental frequency, and the electromagnetic energy at the fundamental frequency. A circuit for obtaining a measurement characteristic by monitoring at a different associated frequency, a transducer coupled between the processing tool and the monitoring circuit for the electromagnetic energy, and the processing by interpreting the measurement characteristic. A processing unit that includes a circuit that determines a state of the device.

(15) The apparatus of claim 14 wherein the processing tool comprises a capacitively coupled plasma processing chamber. (16) The apparatus of claim 14, wherein the processing tool comprises an inductively coupled plasma processing chamber.

(17) The apparatus of claim 14 further including a matching network coupled between the source of electromagnetic energy and the processing tool. (18) The electrical characteristics include power, phase, effective voltage,
15. The apparatus of claim 14, selected from the group consisting of RMS current, peak peak voltage, peak peak current, and impedance.

(19) The apparatus according to item 14, wherein the related frequency is a rational multiple of the fundamental frequency. (20) The related frequency is lower than the fundamental frequency,
The apparatus according to claim 19. (21) The apparatus according to Item 14, wherein the transducer is coupled between the processing tool and a reference potential.

(22) The apparatus of claim 14 wherein said transducer is coupled between said processing tool and said source of electromagnetic energy. (23) The apparatus of claim 22, further comprising a second transducer coupled between the processing tool and a reference potential.

(24) A method of controlling the mechanical state of a processing apparatus, comprising the steps of installing a processing apparatus including an electromagnetic power source and a processing chamber, and applying electromagnetic energy to the processing apparatus at a fundamental frequency. Obtaining a measurement characteristic by monitoring at least one electrical characteristic of the electromagnetic energy at an associated frequency different from the fundamental frequency, the measurement characteristic providing information about the state of the machine. Monitoring the electrical characteristics, determining the information about the condition of the machine by interpreting the measured characteristics, and influencing the condition of the machine. Influencing the condition of the machine, wherein steps are controlled at least in part on the basis of said information And a step of controlling the machine state of the processing apparatus.

(25) The method of claim 24, wherein the machine condition comprises contamination on a wall of the processing chamber and the influencing step comprises cleaning the wall. (26) In the twenty-fifth aspect, the step of cleaning the wall includes a step of maintaining the processing chamber at a low pressure.
Method described in section.

(27) The method of claim 24, wherein the condition of the machine comprises the pressure in the processing chamber and the influencing step comprises the step of varying the pressure. (28) A method according to item 27, wherein the changing step includes a step of reducing the pressure.

(29) Installing a processing apparatus including an electromagnetic power source and a processing chamber, arranging a device to be manufactured in the processing chamber, and applying electromagnetic energy to the processing apparatus at a fundamental frequency. A step of performing a processing step on the device, and obtaining a measurement characteristic by monitoring at least one electrical characteristic of the electromagnetic energy at the fundamental frequency, the measurement characteristic being manufactured. Monitoring the electrical properties, providing information about the device to be determined, determining the information about the device by interpreting the measured properties, and controlling the processing steps And said controlling step is based at least in part on said information. And a control step of the processing step.

(30) A method according to claim 29, wherein performing the processing steps comprises depositing material on the device. (31) The method according to Item 30, wherein the information includes deposition uniformity information.

(32) A method according to item 30, wherein the information includes information on a deposition speed and an end point. 33. The method of claim 29, wherein performing the processing step comprises etching material from the device. (34) A method according to item 30, wherein the information includes etching uniformity information.

(35) The plasma processing apparatus 10 for manufacturing the semiconductor wafer 24 is disclosed. This device
A plasma processing tool 12 and an RF energy source 20 coupled to the plasma processing tool 12 are included. Between the RF energy source 20 and the plasma processing tool 12,
Optional matching network 22 may be included. A circuit 18 for monitoring the RF energy and obtaining the measurement characteristic is also provided. At least one transducer 14 or 16 is coupled between the plasma processing tool 12 and a circuit 18 for monitoring RF energy.
RF energy is usually applied at the fundamental frequency and the electrical properties are monitored at a second frequency that is different from the fundamental frequency. Furthermore, by interpreting the measurement characteristics, the processing device 1
A circuit 19 such as a computer for determining the state of 0
Is also included. Other devices and methods are also disclosed.

This invention is part of Contract No. F under US Air Force Arbitration.
It was conducted under US Government support under 33615-88-C-5448. The US Government has certain rights in this invention.

[Brief description of drawings]

FIG. 1 is a simplified block diagram of a processing apparatus of the present invention.

FIG. 2 is a block diagram of an exemplary circuit for monitoring electromagnetic energy signals.

FIG. 3 shows an example of a capacitively coupled processing device.

FIG. 4 shows an example of an inductively coupled processing device.

FIG. 5a is a graph of experimental phase cosine at 27 MHz and 40 MHz as a function of blanket wafer throughput, and b is at 13.56 MHz for the same wafer as in a. It is a graph of the cosine of the phase of the experimental result.

[Explanation of symbols]

 10 Processing Device 12 Processing Chamber 14 Voltage and Current Transducer 16 Voltage and Current Transducer 18 Electronic Device 19 Personal Computer 20 RF Power Supply

Claims (2)

[Claims] 1. A step of installing a processing apparatus including an electromagnetic power source and a processing chamber, a step of applying electromagnetic energy to the processing apparatus at a fundamental frequency, and monitoring at least one electrical characteristic of the electromagnetic energy. Of the electrical characteristic, wherein the electrical power is applied at the fundamental frequency and the electrical characteristic is monitored at at least one associated frequency different from the fundamental frequency. A method of determining the state of a processing device, comprising: a monitoring step; and a step of determining the state of the processing device by interpreting the measurement characteristic. 2. A processing apparatus comprising: a processing tool; an electromagnetic energy source coupled to the processing tool to apply electromagnetic energy to the processing tool at a fundamental frequency; and the electromagnetic energy to the fundamental frequency. A circuit for obtaining a measurement characteristic by monitoring at different relevant frequencies; a transducer coupled between the processing tool and the monitoring circuit for the electromagnetic energy; and the processing device by interpreting the measurement characteristic. A circuit for determining the state of the processor.