JP2005531913A - Method and apparatus for non-invasive measurement and analysis of plasma parameters - Google Patents

Abstract

Method and apparatus for non-intrusive measurement and analysis of plasma parameters United States Patent Application 20070274305 Kind Code: A1 An RF sensor for detecting and analyzing plasma process parameters. The RF sensor is provided on an antenna for receiving RF energy emitted from a plasma process tool and a plasma process tool. The antenna is positioned adjacent to the plasma process tool because of its non-intrusive nature. In addition, the RF sensor can be configured for multiple harmonic wideband reception of RF energy emitted from the plasma process tool. The RF sensor may be coupled to a high pass filter and a processor for processing the received RF energy. The antenna can be placed in an enclosure with an absorber to reduce the interference experienced by the RF sensor. Tool control can be coupled to a processor that is provided to adjust and maintain various parameters of the plasma process.

Description

(Cross-reference of related applications)
This application is based on US Provisional Application No. 60 / 393,103 filed on July 3, 2002, and claims the benefit of the US Provisional Application, the contents of which are hereby incorporated by reference As incorporated here.

  The present invention relates to a plasma process tool, and more particularly, the present invention relates to a sensing device for non-invasive measurement and parameter analysis of the plasma process tool.

  Plasma processing systems are quite useful for material processing, and for the manufacture and processing of semiconductors, integrated circuits, displays and other devices, both for etching and layer deposition on substrates such as semiconductor wafers. In general, the basic configuration of a plasma processing system consists of a chamber in which plasma is formed, an exhaust region connected to a vacuum port for injecting and removing process gas, and a power source for forming plasma in the chamber. including. In addition, the configuration can include a chuck that supports the wafer and a power source that accelerates plasma ions that have the desired energy for etching or depositing the wafer and that bombard the wafer surface. .

  The power source used to generate the plasma can also be used to accelerate the ions, or different power sources can be used for each task.

  Generally, to ensure that an error-free wafer has been produced, the plasma processing system is monitored using sensors that determine the state of the plasma processing system.

  In general, in such systems, sensors are installed in the plasma to monitor certain parameters or in transmission lines coupled to electrodes in the process chamber.

  The present invention provides a new method and apparatus for measurement and analysis of plasma process parameters.

  An RF sensor for sensing of plasma process parameters is provided with a plasma process tool and an antenna for receiving RF emitted from the plasma process tool. The antenna is placed adjacent to the plasma process tool because of its non-intrusive nature. The antenna may be a broadband monopole antenna.

  In an aspect of the invention, the RF sensor can be coupled to a processor that includes a high pass filter, an amplifier, and a data processing device. Further, the data processing device may be coupled to a user interface for interaction by the user, or may be coupled to a network that allows remote access to the data processing device.

  The present invention is described in further detail below with respect to the specific embodiments that are disclosed. FIG. 1 is an explanatory diagram of an RF sensor according to an embodiment of the present invention. The plasma process tool includes a chamber 110. The plasma process tool is typically moved by an RF power source (not shown). RF energy 120 from an RF power source generates and maintains a plasma 130 within a chamber 110 of a plasma process tool commonly used in substrate processing. The plasma process tool can be assembled in any of a variety of known configurations, all of which include a chamber 110 in which a plasma 130 is shown for processing. Some of these configurations include, for example, inductively coupled plasma (ICP) sources, electrostatic shielded radio wave (ESRF) plasma sources, transformer coupled plasma (TCP) sources and capacitively coupled plasma (CCP) sources. . Regardless of the source of RF energy, the plasma 130 is excited within the chamber 110 by the RF energy generated by the RF power source. Thus, RF energy radiates from the chamber 110 at a fundamental RF frequency and a harmonic of the fundamental RF frequency. The harmonic frequency is generated by the plasma 130. The magnitude and phase of the harmonic frequency provide information on the state of the plasma 130 and the chamber 110. For example, experiments at various powers, pressures, and flow rates show a high degree of correlation between radiated energy and process parameters. Specifically, the analysis shows that the first and second harmonics are better than 99% agreement and correlate with the plasma electron density.

  The antenna 140 is provided outside the plasma chamber 110 to receive RF energy radiated from the plasma 130, and converts the RF energy into an RF signal. In FIG. 1, the antenna 140 is shown outside the chamber 110. Alternatively, it may be placed inside the chamber 110 but outside the process area of the plasma 130. With this configuration, the antenna has the benefit of being non-intrusive to the plasma 130, since intrusive sensors are known to change process parameters. Antenna 140 is coupled to processor 150. The processor 150 is configured to receive the RF signal from the antenna 140 and process the RF signal to provide the desired information of the plasma condition. In addition, since the fundamental frequency of the energy source can be on the order of megahertz, the antenna 140 can be a monopole antenna that is capable of receiving a wide bandwidth and a large bandwidth of radiated RF energy. For example, the Antenna Research Model RAM-220 can be used as a broadband monopole antenna.

  FIG. 2 is a simplified block diagram of an antenna and a processor according to an embodiment of the present invention. In the illustrated embodiment, the antenna 140 is coupled to the high pass filter 210. Alternatively, the antenna 140 may be coupled to other types of filters, such as a band reject, band pass, or low pass filter. The output of the high pass filter 210 is coupled to a low noise amplifier (LNA) 220 and the amplified signal becomes an input to the processor 230. Conventionally, high pass filters do not have any useful information contained in the fundamental frequency that can be utilized to remove the fundamental frequency from the received signal, but rather useful information is contained within the harmonics of the RF energy. Of course, data regarding the fundamental frequency can be collected by removing or adjusting the cutoff frequency of the high pass filter 210. Typical attenuation of the signal below the high pass filter cutoff can be in the range of 40 dB. The LNA 220 amplifies the RF signal provided from the high pass filter so that the signal can be properly processed by the processor 230. The typical gain of LNA can be in the range of 20-30 dB. The processor 230 may be configured to support multiple inputs (multiple inputs) as shown in FIG. In this case, multiple processes can be monitored independently and processed by a single processor 230. The processor 230 can include an analog to digital (A / D) converter to convert the received analog signal into digital samples. The sampling rate of the signal can be determined in various ways. For example, if the fundamental frequency of the RF energy is 13.56 MHz, then a 125 MHz bandwidth will measure 8 harmonics (8th (8th order) harmonics having a frequency of 122.04 MHz). Would be appropriate to. If the sampling interval of the A / D converter is 100 ms and a frequency bin of 10 kHz is selected, the sampling rate is calculated as at least 250 MS / s by the Nyquist criterion. The sample size is 25,000.

  User interface 240, external computer 250, and network 260 are coupled to processor 230. User interface 240 may comprise a variety of known components for the purpose of allowing a user to interact with processor 230. For example, after sampling, if the processor is to perform an FFT (Fast Fourier Transform) of the sampled data, the result is displayed on a touch screen that allows the user to interface with the system. The The external computer 250 can provide various purposes including process parameters and real time control of the chamber 110. Network 260 provides for allowing remote access by the user to and from the processor. For example, the FFT information may be made available to the external computer 250 or to the network 260.

  As an example, such as antenna and processor, chamber parameters can be characterized during calibration conditions, and the data collected by antenna 140 can be applied to a model that relates various parameters of the chamber and plasma. Can do.

  For example, some of the parameters can include electron density, assembly cleanliness, electron temperature, and endpoint detection. The use of such a model can allow the use of the antenna regardless of the absolute calibration of the antenna that can simplify the sensor design parameters.

  FIG. 3 is a simplified block diagram of an antenna according to an embodiment of the present invention. The chamber 110, plasma 130, antenna 140, and processor 150 may be the same as those revealed in FIGS. The antenna 140 is placed in an enclosure 340 that connects to the chamber 110 through the connecting wall 310. The connecting wall 310 is designed to allow RF energy emitted from the plasma 130 to pass, which may be quartz, alumina, or any other suitable material. Alternatively, holes may be provided in the connecting wall 310 to allow RF energy to pass therethrough. Absorbers 320 and 330 not only reduce distortion due to the resonance of enclosure 340, but are also utilized to absorb RF energy from unwanted sources, ie, without absorbers 320 and 330, The antenna can experience unnecessary resonances that distort the signal that must be received. In general, the absorber can comprise a material that absorbs energy at discrete or broadband frequencies.

  Although shown behind the enclosure 340, the absorbers 320 and 330 are placed around the enclosure 340 on five of the sides (if the enclosure is considered to be a rectangular box). This arrangement of the absorber allows RF energy to radiate from the plasma 130 through the connecting wall 310 and inside the enclosure while the absorber is on the other five sides of the box.

  In an embodiment, the absorbers 320 and 330 can be selected so that the absorber 320 can absorb the fundamental frequency, and the absorber 330 can be selected to absorb the first harmonics. Can be done.

  The quarter wave arrangement can provide maximum attenuation of the selected frequency. Furthermore, additional absorbent layers can be used when needed. Although a particular arrangement of absorbers has been described above, any configuration of absorber that reduces unnecessary interference can be utilized.

  FIG. 4 is a simplified block diagram of a plasma processing system according to an embodiment of the present invention. For purposes of illustration, chamber 110 was identified as a capacitively coupled chamber with top electrode 125, however, any type of system could be used in the same way. The plasma 130, antenna 140, and processor 150 may be the same as described above. As previously mentioned, the plasma 130 is excited by an RF power source (generator) 420. The RF power source 420 may be coupled directly to the chamber 110 or may be coupled to the chamber 110 via a matching network 410 or 440 as shown in FIG.

  In FIG. 4, two RF power supplies are shown for illustrative purposes. However, it is possible to use a single RF power source 420 depending on the configuration of the chamber 110. Upper electrode (UEL) matching network 410 is coupled to upper electrode 125, and lower electrode (LEL) matching network 440 is coupled to lower electrode 450. Plasma 130 is excited by RF power source (s) 420. Accordingly, the plasma 130 emits RF energy at the fundamental frequency and harmonics of the fundamental frequency. RF energy is radiated from the chamber 110 and received by the antenna 140 positioned outside the plasma 130. The antenna 140 is coupled to the processor 150 described in part above. The arrangement described with reference to FIG. 1 and described above provides a non-invasive method of receiving plasma process parameters.

  The processor 150 receives the RF energy and converts the analog signal to a digital signal through an analog-to-digital (A / D) converter. Typically, the sampling rate of an analog signal depends on a significant bandwidth (ie, the bandwidth is a function of the fundamental frequency and significant harmonics). For example, a 500 MHz bandwidth can typically be sampled at a sample rate of 1 billion times per second. Of course, if desired, the sampling rate can be determined without being limited to the above example. The magnitude and phase of the RF energy, including the harmonics, can provide information about the state of the plasma 130 and thus the state of the chamber 110. Data can then be processed by processor 150, and operations such as fast Fourier transform (FFT) and principal component analysis (PCA) can generally be used to gather information from the RF signal. . Information obtained by the processor 150 can provide insights into parameters such as assembly cleanliness, plasma density, electron temperature and endpoint detection.

  As one embodiment of the processor, the received RF energy trace data may be converted to a frequency domain output signal using conventional techniques including FFT. Information at the harmonic frequency can be extracted and multiplied by the resulting coefficient during calibration of the plasma processing system and determined by PCA. PCA is useful in determining coefficients because it allows a large set of correlated values to be converted to a smaller set of principal values. The reduction in the size of the set can make it possible to transform the original set of values into an uncorrelated linear combination of the original (larger) set.

  Using the fundamental frequency magnitude and the harmonic frequency of the received RF energy, it is possible to perform a number of different analyses, including power analysis, flow analysis and pressure analysis. By processing the information obtained from the magnitude values, it is possible to determine which of the harmonics has the largest correlation and, as a result, to determine the acceptable coefficients for each frequency component. It is even possible. Dependency analysis can also be determined because even if one parameter in the system affects the other, the initial result indicates that the parameter can be adjusted independently.

  Furthermore, end point detection may be possible from analysis of trace data. Once plotted, it becomes clear that there is a significant variation in the harmonics of the received RF energy. More specifically, it is possible for the main harmonic contribution to change upon completion of the process.

  For example, as shown in FIG. 5, which shows simplified and expected data, the change in the third (third order) harmonic is clear at T1, both the basic and third (third order) harmonics. The change in is clear at T2. Process analysis indicates that these changes are due to process completion. Such a method of endpoint detection can be an accurate and cost effective method of endpoint detection.

  The processed data is then sent to the tool control 430. Tool control 430 may be configured to fulfill multiple tasks. Some of the tasks that the tool control 430 can perform include endpoint determination, power control and gas control (flow, pressure, etc.). As shown in FIG. 4, the tool control 430 is coupled to the chamber 110 and the RF power source 420. In this way, it is possible for the tool control to adjust the parameters of these devices with data received from the processor 150 so that a reproducible process can be maintained in the chamber 110.

  As mentioned earlier, PCA is a multivariate statistical method that allows a large set of correlated variables to be reduced to a smaller set of major components. Thereby, during phase calibration, the PCA can be used to initially generate a covariance matrix from a data set comprising various harmonic data. The eigensolution can then be obtained from the covariance matrix, and thus a set of eigenvectors can be calculated. From the eigensolution, the percent contribution of each major component can be calculated. Using percentages, the coefficients can be selected according to the weighted total number of eigenvectors in the resulting percentage. For various parameters including power, gas flow, and chamber pressure, this calculation can be performed. If the calibration is complete and various coefficients are determined, the tool control can use the information in the control loop as will be apparent to those skilled in the art. In this type of feedback loop, a reproducible process can be maintained.

  The processor 150 can be coupled to multiple devices as shown in FIG. Some of the devices that are important in this embodiment include a user interface 240 and an external computer 250. Further, the user interface 240 and the external computer 250 can be a single device (eg, a personal computer).

  Finally, as will be appreciated by those skilled in the art, the amount of data processed by the processor 150 can be quite large. For this point, it is necessary to use an external storage device (not shown). One possible configuration for connecting storage devices is to be directly connected to the processor 150. Alternatively, it may be beneficial to use remote storage through network 260 (shown in FIG. 2). However, any method of storing data is acceptable. One benefit of accumulating data is for future processing and analysis. Further, the archived data can be utilized to model an acceptable control system for operating the tool control 430 and thus control the plasma processing.

  The previous description of the described embodiments is provided to enable any person skilled in the art to utilize the invention. Various modifications to these embodiments are possible, and the general principles of RF sensors for the measurement of semiconductor process parameters presented here can be applied to other embodiments. Thus, the present invention is not intended to be limited to the embodiments shown above, but rather the principles and novelty of the forms disclosed in any way are consistent with the broadest scope. Yes.

It is explanatory drawing of RF sensor by embodiment of this invention. FIG. 3 is a simplified block diagram of an antenna and a processor according to an embodiment of the present invention. 1 is a simplified block diagram of an antenna according to an embodiment of the present invention. 1 is a simplified block diagram of a plasma processing system according to an embodiment of the present invention. 6 is a simplified graph of expected harmonics data according to an embodiment of the present invention.

Claims (13)

An RF sensor for detecting a parameter of a plasma process,
A plasma process tool having a plasma process region;
An RF sensor comprising an antenna provided outside the plasma process region for receiving RF energy emitted from the plasma process tool and having a fundamental frequency and a plurality of harmonic frequencies.   The RF sensor of claim 1, further comprising a processor coupled to the antenna to process the RF energy received at the antenna. The processor is
A filter coupled to the antenna;
An amplifier coupled to the filter;
The RF sensor of claim 2, further comprising a data processing device coupled to the amplifier.   The RF sensor of claim 3, wherein the data processing device may be configured to support at least two input signals independently.   The RF sensor according to claim 3, wherein the filter is a high-pass filter.   The RF sensor according to claim 3, wherein the amplifier is a low noise amplifier. A user interface coupled to the data processing device;
An external computer coupled to the data processing device,
The RF sensor of claim 3, wherein the user interface and the external computer are configured to allow a user to interact with the data processing device.   The RF sensor according to claim 7, wherein the user interface is a touch screen monitor.   The RF sensor of claim 3, wherein the data processing device is coupled to a network to allow a user to communicate remotely with the data processing device.   The RF sensor according to claim 2, wherein the processor is configured to perform at least one of spectral analysis and harmonic content analysis of the RF energy.   The RF sensor according to claim 1, wherein the antenna is a broadband monopole antenna. A method for detecting plasma process parameters comprising:
Installing an antenna outside the plasma process area and adjacent to the plasma process tool;
Sensing RF energy emitted from the plasma process tool.   The method of claim 12, further comprising processing the RF energy, including at least one of spectral analysis and harmonic content analysis of the RF energy.

Images