Systems and methods for tuning mhz RF generator within cycle of operation of khz RF generator

By regulating kilohertz RF generators through frequency and phase detection, and using computer-based models to adjust higher frequency RF generators with periodic fluctuations, the method addresses the challenge of power reflection in plasma tools, improving efficiency and reducing damage while eliminating the need for high-speed data acquisition devices.

JP2024156989AActive Publication Date: 2024-11-06LAM RES CORP
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
JP2024135174
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2019-06-07
Filing Date
2024-08-14
Publication Date
2024-11-06
Estimated Expiration
2040-05-06

AI Technical Summary

Technical Problem

In plasma tools, the reflection of power from the plasma chamber towards the RF generators reduces processing efficiency and causes damage, particularly due to the strong modulation of load impedance by lower RF frequencies, making it difficult to determine a single combination of variable capacitors and RF frequencies that work across the entire cycle.

Method used

A method is provided to regulate the kilohertz RF generator by detecting frequency and phase waveforms, using sensors and high-speed data acquisition to measure complex voltages and currents, and calculating optimal frequency parameters to minimize power reflection coefficients through a computer-based model, adjusting the higher frequency RF generator with periodic fluctuations and predetermined functional forms.

Benefits of technology

This approach significantly reduces power and voltage reflection towards the megahertz RF generator, minimizing tool-to-tool and wafer-to-wafer variations without the need for high-speed data acquisition devices during processing, thereby enhancing processing efficiency and reducing damage.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024156989000001_ABST
    Figure 2024156989000001_ABST
Patent Text Reader

Abstract

To provide systems and methods for tuning a megahertz radio frequency (RF) generator within a cycle of operation of a kilohertz (kHz) RF generator.SOLUTION: In a system 100, a predetermined periodic waveform is provided to a processor 128 of a computer 118. The processor uses a computer-based model to determine a plurality of frequency parameters for the predetermined periodic waveform. The frequency parameters are applied to a megahertz RF generator 114 to generate an RF signal having the frequency parameters during one or more cycles of operation of a kilohertz RF generator 112.SELECTED DRAWING: Figure 1A
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present embodiment relates to a system and method for regulating a megahertz (MHz) radio frequency (RF) generator during an operating cycle of a kilohertz (kHz) RF generator. [Background technology]

[0002] The background description provided herein is intended to provide a general overview of the contents of the present disclosure. Work by the currently named inventors within the scope of what is described in this Background section, as well as aspects of the description that may not otherwise be regarded as prior art at the time of filing, are not admitted, expressly or impliedly, as prior art against the present disclosure.

[0003] In a plasma tool, one or more radio frequency (RF) generators are coupled to an impedance matching circuit. The impedance matching circuit is coupled to a plasma chamber. An RF signal is provided from the RF generator to the impedance matching circuit. The impedance matching circuit outputs an RF signal upon receiving the RF signal. The RF signal is provided from the impedance matching circuit to the plasma chamber for processing a wafer in the plasma chamber. During processing of a wafer, an amount of power is reflected from the plasma in the plasma chamber through the impedance matching circuit towards the one or more RF generators. The reflected power reduces wafer processing efficiency and also causes damage to the one or more RF generators.

[0004] It is in this context that the embodiments described in this disclosure arise. Summary of the Invention

[0005] The embodiments of the present disclosure provide an apparatus, method, and computer program for regulating a megahertz (MHz) radio frequency (RF) generator during an operating cycle of a kilohertz (kHz) RF generator. It should be understood that the embodiments can be implemented in many ways, such as a process, an apparatus, a system, a piece of hardware, or a method on a computer readable medium. Some embodiments are described below.

[0006] Some plasma etching systems operate at two different RF frequencies, one low, such as 400 kilohertz (kHz), and one that is more than 100 times higher, such as 60 megahertz (MHz). The load impedance for the higher RF frequency can be strongly modulated by the lower RF frequency. The matching network for each RF frequency is adjusted using one variable capacitor and a variable RF frequency. For example, the higher RF frequency is adjusted to a value in the range of 57 MHz to 63 MHz. A single value of the load impedance can be adjusted so that the voltage reflection coefficient Γ is close to zero, but since the impedance of the higher RF frequency varies with the phase of the lower RF frequency, it is difficult to determine a single combination of the variable capacitor and the higher RF frequency that works for all parts of the cycle of the lower RF frequency. In the absence of a single combination, the time-average power reflection coefficient Γ 2 could be as high as 50%.

[0007] In one embodiment, a method is provided where the higher RF frequency has a mean value as well as a variation around the mean value. The variation is more periodic at the lower RF frequency and has a predefined functional form. For example, the variation is a square wave with an amplitude and phase relative to the lower RF frequency. As another example, the variation is a sinusoidal or rectangular function. As yet another example, the variation is a trapezoidal function with a high frequency segment and a low frequency segment connected by a frequency ramp.

[0008] In a first embodiment, a method for determining frequency parameters and match network parameters for controlling a higher frequency RF generator is provided. In the first embodiment, the following operations are performed: 1. Plasma tools detect the frequency and phase of lower frequency waveforms. As an example, the lower frequency is detected by interrogating a lower frequency RF generator, and the phase of the lower frequency is inferred from detecting the voltage zero crossings of the lower frequency. 2. The plasma tool uses sensors and high speed data acquisition to store the higher frequency waveform measured between the higher frequency RF generator and the matching network over a period of one or several cycles of the lower frequency. As an example, the sensor measures complex voltage and complex current. 3. The computer calculates the instantaneous complex voltage reflection coefficient Γ at short time intervals, for example, about 0.1 microseconds. The computer also calculates the average power reflection coefficient Γ associated with the higher frequency RF generator over a period of time at the lower frequency. 2 Calculate. 4. The computer calculates a set of optimal parameters that minimizes the predicted average power reflection coefficient. For example, the computer may: a. Divide the lower frequency measured voltage waveform into short segments of approximately 0.1 microseconds each, producing approximately 25 short segments per lower frequency period. b. Calculate the higher frequency complex voltage reflection coefficient Γ for each segment. c. Applying a computer-based model of the plasma tool's RF path to calculate the load impedance for each segment from the complex voltage reflection coefficient Γ. Inputs to the calculation include the known RF frequency and the variable capacitor value. d. For a given functional form at the higher frequency, calculate a set of frequency parameters that minimizes the period-averaged reflection coefficient associated with the higher frequency RF generator. By way of example, frequency parameters include: i. Constant average frequency of the higher frequency RF generator. ii.The amplitude of the trapezoidal frequency variation. iii. the duration of the high and low sections of the trapezoidal frequency variation; and iv. The phase of the trapezoidal frequency variation relative to the phase of the lower frequency; 5. Optimization to minimize the average power reflection coefficient involves variable capacitor values ​​as well as higher frequencies. 6. The computer applies a frequency variation. 7. The computer repeats steps 2 to 5. After applying this method, the power reflection coefficient in the higher frequency RF generator is reduced from about 35% to about 3%.

[0009] In a second embodiment, a method is provided for reducing voltage and power reflection coefficients at the input of a computer-based model. The method is performed during recipe development to determine operational parameters that are applied during processing of a substrate. In the method, the following operations are performed: 1. Eliminates issues and time associated with the use of high speed data acquisition devices and sensors because they are installed during recipe development and are not used on the production tool during processing. One example of a high speed data acquisition device is an oscilloscope. 2. During recipe development, two actions take place: a. An adjustment action that determines the value of the average frequency applied to the higher frequency RF generator. b. A high speed data sensing, acquisition, and computer analysis operation that determines the frequency variation at the average frequency. As an example, the two operations are performed to minimize the average voltage reflection coefficient, which can have positive and negative values ​​averaging to zero. This means that the average voltage reflection coefficient can be zero even if the average power reflection coefficient is large. To further illustrate, the operation of selecting the frequency variation coincides with the operation for determining the average frequency. The frequency variation around the average frequency is determined simultaneously with the operation for determining the average frequency. In the case of multiple RF generators, the average frequency is determined by minimizing the portion of the reflected power at the fundamental frequency and ignoring the remaining portion of the reflected power at one or more modulated sideband frequencies of the reflected power. This is equivalent to minimizing the Fourier peak of the reflected power at the fundamental frequency, or minimizing the average voltage reflection coefficient, which is a complex number with positive and negative parts canceling each other out. 3. During recipe development, the average frequency and frequency variation are determined in the same manner as described above in the first embodiment, with the additional constraint that the average voltage reflection coefficient is simultaneously minimized. 4. The computer stores the frequency parameters. 5. During processing, for a particular recipe, tool-to-tool or wafer-to-wafer variations are averaged at a higher frequency. The average frequency determined during recipe development is applied during processing to minimize the average voltage reflection coefficient. Then, in addition to the average frequency, the frequency variation previously determined for that recipe is applied tool-to-tool or wafer-to-wafer. The timing and duration of the frequency variation are synchronized with the zero crossings and period of the lower frequency RF generator. In the second embodiment, no high speed data acquisition devices and / or sensors are used during processing.

[0010] In one embodiment, a method is provided in which the higher RF frequency is adjusted in two parts. The first part determines the average value of the higher RF frequency, which remains constant for many cycles of the lower RF frequency, and the second part determines the fast variation around the average value that is applied within a cycle of the lower RF frequency.

[0011] In the third embodiment, a method is described to reduce the power reflection coefficient for each bin within a cycle of a lower RF frequency. The lower RF frequency is divided into multiple bins. There is no predefined function form used in the third embodiment. The following operations are performed in the third embodiment: 1. During recipe development, the computer performs a fast frequency variation operation by subdividing the period or cycle of the lower RF frequency into several short time bins. For example, the period of the lower RF frequency is divided into 16 bins. The length or duration of each bin depends on the value of the lower RF frequency. An example of a lower RF frequency is a frequency in the range of 340 kHz to 440 kHz. 2. During recipe development, the computer applies a computer-based model to minimize the average power reflection coefficient for a bin by modifying: a. A variable capacitor that determines the capacitance; b. The average frequency of the higher RF frequencies; c. The frequency variation from the average frequency, and d. Subject to the constraint that the average voltage reflection coefficient is also minimized 3. During tool operation in which a substrate is processed, for a particular recipe, tool-to-tool or wafer-to-wafer variations are assumed to be at the average value of the higher RF frequency. 4. Also during tool operation, the frequency variations previously determined for that recipe are applied by the computer from bin to bin in addition to the average frequency of the higher RF frequency. The timing and duration of the frequency variations are synchronized by the computer with the zero crossings and period of the lower RF frequency. Also during tool operation, the capacitance determined during recipe development is applied to all bins of the lower RF frequency.

[0012] In one embodiment, an adjustment method is described. The adjustment method includes accessing a plurality of reflection parameter values ​​associated with a second radio frequency generator for a first set of one or more cycles of operation of the first radio frequency generator. The adjustment method further includes calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path. The radio frequency path is between the second radio frequency generator and an electrode of the plasma chamber. The adjustment method also includes receiving a plurality of frequency modulation parameters of a radio frequency signal generated by the second radio frequency generator. The adjustment method further includes determining values ​​of the plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model. The values ​​of the plurality of frequency modulation parameters are determined to minimize a reflection coefficient parameter at an input of the computer-based model. The adjustment method includes controlling the second radio frequency generator according to the values ​​of the plurality of frequency modulation parameters during a second set of one or more cycles of operation of the first radio frequency generator.

[0013] In one embodiment, an adjustment method is described. A portion of the adjustment method is performed during recipe development, and another portion of the adjustment method is performed during processing. The portion performed during recipe development includes accessing a plurality of reflection parameter values ​​associated with a second radio frequency generator for a set of one or more cycles of operation of the first radio frequency generator. Additionally, the portion performed during recipe development further includes calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path. The radio frequency path is between the second radio frequency generator and an electrode of the plasma chamber. The portion performed during recipe development also includes receiving a plurality of frequency modulation parameters of a radio frequency signal generated by the second radio frequency generator. The portion performed during recipe development further includes determining values ​​of the plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model. The values ​​of the plurality of frequency modulation parameters are determined to minimize one or more reflection coefficient parameters at an input of the computer-based model. Another portion of the method is performed during processing of the substrate in another plasma chamber. Another portion of the method includes controlling a third radio frequency generator according to values ​​of a plurality of frequency modulation parameters determined during recipe development, the operation of controlling the third radio frequency generator being performed during a set of one or more cycles of operation of the fourth radio frequency generator.

[0014] In one embodiment, an adjustment method is described. A portion of the adjustment method is performed during recipe development, and another portion of the adjustment method is performed during processing. The portion performed during recipe development includes accessing a plurality of reflection parameter values ​​associated with a second radio frequency generator for a set of one or more cycles of operation of the first radio frequency generator. The portion performed during recipe development further includes calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path between the second radio frequency generator and an electrode of the plasma chamber. The portion performed during recipe development also includes receiving a plurality of frequency modulation parameters of a radio frequency signal generated by the second radio frequency generator, the plurality of frequency modulation parameters including a frequency variation of the radio frequency signal. The portion performed during recipe development further includes determining values ​​of the plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model. The values ​​of the plurality of frequency modulation parameters are determined to minimize one or more reflection coefficient parameters at an input of the computer-based model. Another portion of the method is performed during processing of the substrate in another plasma chamber. Another portion of the method includes controlling a third radio frequency generator according to values ​​of a plurality of frequency modulation parameters determined during recipe development, the operation of controlling the third radio frequency generator including applying the values ​​of the plurality of frequency modulation parameters to a baseline frequency of operation of the third radio frequency generator.

[0015] In one embodiment, an adjustment method is described. A portion of the adjustment method is performed during recipe development, and another portion of the adjustment method is performed during processing. The portion performed during recipe development includes accessing a plurality of reflection parameter values ​​associated with a second radio frequency generator for a set of one or more cycles of operation of the first radio frequency generator. Each of the plurality of reflection parameter values ​​corresponds to a respective bin of the one or more cycles of operation of the first radio frequency generator. The portion performed during recipe development further includes calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path. The radio frequency path is between the second radio frequency generator and an electrode of the plasma chamber. The portion performed during recipe development also includes determining values ​​of a plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model. The values ​​of the plurality of frequency modulation parameters are determined to minimize a plurality of values ​​of a reflection coefficient parameter at an input of the computer-based model for each of the bins. Another portion of the method is performed during processing of the substrate in another plasma chamber. Another portion of the method includes controlling a third radio frequency generator according to values ​​of a plurality of frequency modulation parameters determined during recipe development, the operation of controlling the third frequency generator being performed during a set of one or more cycles of operation of the fourth radio frequency generator.

[0016] Some advantages of the systems and methods described herein include finding higher RF frequencies and variable capacitor values ​​that reduce the power and / or voltage reflected back to the megahertz RF generator during each cycle of operation of the kilohertz RF generator. Additional advantages include reducing the number of sensors and high speed data acquisition devices during the processing of the substrate. For example, in the second and third embodiments, the use of high speed data acquisition devices and sensors is not required after recipe development. Once a recipe with higher RF frequencies and variable capacitor values ​​is determined, the recipe is applied during the processing of the substrate. There is no need to measure complex voltages and complex currents again during the processing of the substrate.

[0017] Further advantages include applying a predetermined functional form, which simplifies the determination of frequency parameters and the application of operating parameters, and it is easy to control a higher frequency RF generator to follow the predetermined functional form.

[0018] Furthermore, when using a computer-based model, it is not necessary to use sensors in the RF path. For example, sensors are not coupled to the output of the RF transmission line or impedance matching network. Rather, complex voltages and currents are measured by sensors coupled to the output of the megahertz RF generator and propagated through the computer-based model to facilitate the determination of higher frequencies and variable capacitor values. It is more difficult and time consuming to use sensors in the RF path than using a computer-based model.

[0019] Other aspects will become apparent from the following detailed description taken in conjunction with the accompanying drawings. [Brief description of the drawings]

[0020] The embodiments may be best understood by referring to the following description taken in conjunction with the accompanying drawings.

[0021] [Figure 1A]FIG. 1A is a diagram of one embodiment of a system illustrating the determination of a value of a voltage reflection coefficient over one or more cycles of operation of an X kilohertz radio frequency (RF) generator.

[0022] [Figure 1B-1] FIG. 1B-1 is a diagram illustrating one embodiment of a table illustrating that the value of the voltage reflection coefficient is determined based on complex voltages and complex currents obtained at periodic time intervals over a cycle of operation of an X kilohertz RF generator.

[0023] [Figure 1B-2] FIG. 1B-2 shows one embodiment of a graph illustrating the correspondence between voltage reflection coefficient values ​​and time intervals during a cycle of an RF signal generated by an X kilohertz RF generator.

[0024] [Figure 1C] FIG. 1C is a diagram of one embodiment of an RF path model illustrating the calculation of load impedance values ​​from voltage reflection coefficient values.

[0025] [Figure 1D-1] FIG. 1D-1 is a diagram of one embodiment of an RF path model illustrating the determination of operational parameters, including frequency parameters and match network parameters, based on application of a load impedance value at the output of the RF path model that minimizes the average power reflection coefficient at the input of the RF path model.

[0026] [Figure 1D-2] FIG. 1D-2 shows one embodiment of a graph illustrating multiple cycles of an RF signal generated by an X kilohertz RF generator.

[0027] [Figure 1D-3] FIG. 1D-3 is a diagram of one embodiment of a system that applies values ​​of operational parameters during a set of cycles of operation of the X kilohertz RF generator of FIG. 1 during processing of a substrate.

[0028] [Figure 2A] FIG. 2A shows one embodiment of a graph illustrating cycles of an RF signal.

[0029] [Figure 2B] FIG. 2B shows one embodiment of a graph illustrating a forward voltage waveform of an RF signal provided by a Y megahertz (MHz) RF generator.

[0030] [Figure 2C] FIG. 2C shows one embodiment of a graph illustrating the reverse voltage waveform of an RF signal provided by a Y MHz RF generator.

[0031] [Figure 2D] FIG. 2D is an embodiment of a graph illustrating forward and reverse voltage waveforms on one graph.

[0032] [Figure 2E] FIG. 2E shows one embodiment of a Smith chart illustrating the value of the voltage reflection coefficient at the output of a Y MHz RF generator before applying the methods described herein.

[0033] [Figure 3A] FIG. 3A is an embodiment of a graph illustrating a comparison between the frequency of an RF signal output from a Y megahertz RF generator after applying a method described herein and the RF signal output from the Y megahertz RF generator before applying the method.

[0034] [Figure 3B] FIG. 3B shows one embodiment of a Smith chart illustrating the value of the voltage reflection coefficient at the output of a Y MHz RF generator after applying the methods described herein.

[0035] [Figure 4A]FIG. 4A is a diagram of one embodiment of a system illustrating that the method described above with respect to FIG. 1A can be applied to the system of FIG. 4A during recipe development, rather than during processing of a substrate.

[0036] [Figure 4B] FIG. 4B is an illustration of one embodiment of a table that is generated by a processor during recipe development in the same manner that the table of FIG. 1B-1 is generated by a processor during processing of a substrate.

[0037] [Figure 4C] FIG. 4C is a diagram of one embodiment of an RF path model illustrating the generation of load impedance values ​​from voltage reflection coefficient values ​​for or during recipe development.

[0038] [Figure 4D] FIG. 4D is a diagram of one embodiment illustrating the determination of values ​​of operational parameters for or during recipe development.

[0039] [Figure 4E] FIG. 4E is a diagram of one embodiment of a system illustrating the application of operational parameter values ​​determined during recipe development during processing.

[0040] [Figure 5A] FIG. 5A is a diagram of one embodiment of a system illustrating the generation of voltage reflection coefficient values ​​by a processor during recipe development for each bin of an operating cycle of an X kilohertz RF generator.

[0041] [Figure 5B] FIG. 5B is an illustration of one embodiment of a table generated by a processor for recipe development for each bin of one or more cycles of operation of an X kilohertz RF generator.

[0042] [Figure 5C]FIG. 5C is a diagram of one embodiment of an RF path model illustrating the generation of load impedance values ​​for each bin of an operating cycle of an X kilohertz RF generator from voltage reflection coefficient values ​​during recipe development.

[0043] [Figure 5D] FIG. 5D is a diagram of one embodiment of an RF path model illustrating the determination of values ​​of operational parameters at which the value of the power reflection coefficient for a bin of an operating cycle of an X kilohertz RF generator is minimized during recipe development.

[0044] [Figure 5E] FIG. 5E is a diagram of one embodiment of a system illustrating the use of operational parameters for each bin during processing of a substrate. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0045] The following embodiments describe systems and methods for regulating a megahertz (MHz) radio frequency (RF) generator during an operating cycle of a kilohertz (kHz) RF generator. It will be apparent that the embodiments may be practiced without some or all of these specific details. In other instances, well known process operations have not been described in detail so as not to unnecessarily obscure the embodiments.

[0046] FIG. 1A is a diagram of an embodiment of a system 100, such as a plasma tool, illustrating the determination of a value of a voltage reflection coefficient over one or more cycles of operation of an X kilohertz RF generator 112. The system 100 includes an X kilohertz RF generator 112, a Y megahertz RF generator 114, a computer 118, and a plasma chamber 108. The system 100 also includes an impedance matching network 110, a sensor 120, and a high-speed data acquisition device 121. An example of a sensor 120 is a voltage sensor, and an example of a high-speed data acquisition device is a digital storage oscilloscope. Another example of a high-speed data acquisition device is a high-speed data acquisition circuit board. The sensor 120 and the high-speed data acquisition device 121 both measure forward and reverse voltages. Illustratively, the sensor 120 and the high-speed data acquisition device 121 measure a forward voltage amplitude, a reverse voltage amplitude, and a phase between the forward and reverse voltage amplitudes. It should be noted that the terms reverse voltage and reflected voltage are used interchangeably herein.

[0047] An example of an X kilohertz RF generator as described herein includes a 400 kHz RF generator or another RF generator operating at kilohertz frequencies. An example of a Y megahertz RF generator as described herein includes a 2 MHz RF generator or a 13.56 MHz RF generator or a 27 MHz RF generator or a 60 MHz RF generator. Examples of computers as described herein include desktop computers, laptop computers, smartphones, and tablets.

[0048] The computer 118 includes a processor 126 and a memory device 128. The processor 126 is coupled to the memory device 128 and is also coupled to the RF power sources 122 and 124. As used herein, a central processing unit (CPU), controller, application specific integrated circuit (ASIC), or programmable logic device (PLD) is used in place of a processor, and these terms are used interchangeably herein. Examples of memory devices include read only memory (ROM), random access memory (RAM), hard disk, volatile memory, non-volatile memory, redundant array of storage disks, flash memory, and the like. The processor 126 is coupled to a high speed data acquisition device 121 coupled to the sensor 120 via a cable, such as a serial transfer cable, a parallel transfer cable, or a universal serial bus (USB) cable.

[0049] An example of a high-speed data acquisition device described herein includes a device for acquiring data related to the oscillation or frequency of oscillation of the voltage or power of an RF signal output at an output terminal of an RF generator. The sensor 120 is coupled to the output O1 of the X megahertz RF generator 112 via a directional coupler. The Y megahertz RF generator 114 includes a sensor 131, such as a complex voltage and complex current probe or a complex voltage sensor. The sensor 131 is coupled to the output terminal O2 of the Y megahertz RF generator 114. The sensor 131 is also coupled to the high-speed data acquisition device 123, which is coupled to the processor 126 via a cable. Examples of cables are described above.

[0050] The impedance matching network 106 includes one or more circuit components coupled to each other, such as one or more inductors, or one or more capacitors, or one or more resistors, or a combination or two or more thereof. For example, the impedance matching network 106 includes two branches. Each of the two branches includes multiple circuit components coupled to each other in series or parallel. One of the two branches is coupled to the X kilohertz RF generator 112, and the other of the two branches is coupled to the Y megahertz RF generator 114, and the branch is coupled to the output of the impedance matching network 106. It should be noted that the terms impedance matching network, match, impedance matching circuit, match circuit, and match network are used interchangeably herein.

[0051] One example of the plasma chamber 108 is a capacitively coupled plasma (CCP) chamber, in which the top electrode 106 and the bottom electrode of the plasma excitation electrode 104 are mounted to be capacitively coupled to each other. For example, the top electrode 106 is mounted on the plasma excitation electrode 104, forming a gap between the top electrode 106 and the plasma excitation electrode 104. The top electrode 106 is coupled to a ground potential. Radio frequency power is capacitively coupled between the top electrode 106 and the plasma excitation electrode 104 through the gap. Each of the bottom electrode and the top electrode 106 is made of a metal, for example, anodized aluminum, an aluminum alloy, or the like.

[0052] An output terminal O1 of the X kilohertz RF generator 112 is coupled to an input terminal I1 of the impedance matching network 110 via an RF cable RFC1, and an output terminal O2 of the Y megahertz RF generator 114 is coupled to another input terminal I2 of the impedance matching network 110 via another RF cable RFC2. The output of the impedance matching network 110 is also coupled to the bottom electrode of the plasma excitation electrode 104 via an RF transmission line RFT1.

[0053] The X kilohertz RF generator 112 includes an RF power supply 122 that is an RF oscillator. Similarly, the Y kilohertz RF generator 114 includes an RF power supply 124 that is also an RF oscillator. The RF oscillator 122 is coupled to an output terminal O1, and the RF oscillator 124 is coupled to an output terminal O2. The combination of the RF cable RFC2, the branch of the impedance matching network 110 coupled between an input I2 and the output of the impedance matching network 110, the RF transmission line RFT1, and the plasma excitation electrode 104 is referred to herein as the RF path 102. Another branch of the impedance matching network 110 is coupled between an input I1 and the output of the impedance matching network 110.

[0054] A substrate S, such as a semiconductor wafer, is placed on top of the plasma excitation electrode 104 for processing, which may include etching a layer of the substrate S, or depositing material onto the substrate S, or sputtering the substrate S, or cleaning the substrate S, or a combination thereof. During processing of the substrate S, the processor 126 provides one or more power values ​​and one or more frequency values ​​to the RF power source 122 of the X kilohertz RF generator 112 and provides one or more power values ​​and one or more frequency values ​​to the RF power source 124 of the Y megahertz RF generator 114.

[0055] Upon receiving the power and frequency values, RF power source 122 generates an RF signal 130 having a power and frequency value and provides RF signal 130 via output O1 and RF cable RFC1 to input I1 of impedance matching network 110. Similarly, upon receiving the power and frequency values, RF power source 124 generates an RF signal 132 having a power and frequency value and provides RF signal 132 via output O2 and RF cable RFC2 to input I2 of impedance matching network 110.

[0056] The impedance matching network 110 receives the RF signals 130 and 132, matches the impedance of a load coupled to the output of the impedance matching network 110 with the impedance of a source coupled to the inputs I1 and I2 of the impedance matching network 110, and combines the RF signals 130 and 132 to output a modified RF signal 134 at the output of the impedance matching network 110. An example of a load coupled to the output of the impedance matching network 110 includes the plasma chamber 108 and the RF transmission line RFT1. An example of a source coupled to the inputs I11 and I21 includes the X kilohertz RF generator 112, the Y megahertz RF generator 114, and the RF cables RFC1 and RFC2. The modified RF signal 134 is supplied to the lower electrode of the plasma excitation electrode 104 via the RF transmission line RFT1 to strike or maintain a plasma in a gap formed between the plasma excitation electrode 104 and the upper electrode 106 of the plasma chamber 108. The plasma in the plasma chamber 108 has RF power to process the substrate S.

[0057] During processing of the substrate S, the sensor 120 and high speed data acquisition device 121 measure the voltage or power of the RF signal 130 output from the X kilohertz RF generator 112 and output a waveform, such as a voltage waveform or a power waveform. The waveform measured by the sensor 120 and high speed data acquisition device 121 has a phase φ of the RF signal 130. The phase φ of the RF signal 130 provides crossing points or times at which the waveform of the RF signal 130 crosses a zero value of voltage or power. The waveform of the RF signal 130 is sinusoidal or substantially sinusoidal and crosses a zero value of voltage or power at periodic time intervals.

[0058] The processor 126 receives the values ​​of the waveform measured by the sensor 120 and the high speed data acquisition device 121 via the cable and determines the phase φ of the waveform of the RF signal 130. For example, the processor 126 determines the time when the waveform of the RF signal 130 becomes zero at a positive crossing point, which defines the phase φ. At the positive crossing point, the waveform of the RF signal 130 transitions from a negative value to a positive value.

[0059] Also, during processing of the substrate S, the sensor 131 and high speed data acquisition device measure the forward voltage amplitude, the reverse voltage amplitude, and the phase between the forward and reverse voltage amplitudes at the output O2 of the Y megahertz RF generator 114 and provide the amplitudes and phases to the processor 126 via the cable. The processor 126 determines multiple values ​​Γ11, Γ21 through Γn1 of the voltage reflection coefficient over one cycle (e.g., one period) of the RF signal 130 based on the forward voltage amplitude, the reverse voltage amplitude, and the phase, where n is an integer greater than 2. For example, the processor 126 determines the value of the voltage reflection coefficient as the ratio of the voltage reflected towards the Y megahertz RF generator 114 at the output O2 to the voltage supplied from the Y megahertz RF generator 114 at the output O2. The reflected voltage is a complex number having an amplitude and a phase, and the supply voltage is also a complex number having an amplitude and a phase. The terms supply voltage and forward voltage are used interchangeably herein. The voltage reflected back towards the Y megahertz RF generator 114 is reflected out of the plasma chamber 108 via the RF path 102. The reflected voltage and the voltage provided at the output O2 are identified by the processor 126 from the forward voltage amplitude, the reverse voltage amplitude, and the phase between the amplitude measured by the sensor 131 and the high speed data acquisition device 123. The processor 126 stores the voltage reflection coefficient values ​​Γ11, Γ21 to Γn1 in the memory device 128 and accesses (e.g., for retrieval or reading) the values ​​Γ11, Γ21 to Γn1 from the memory device 128. The voltage reflection coefficient is one example of a reflection parameter or reflection coefficient parameter. Another example of a reflection parameter is a complex voltage and a complex current, or another parameter having a complex value.

[0060] In one embodiment, the processor 126 is coupled to the RF generator's RF power supply via the RF generator's digital signal processor, and the RF generator's one or more power controllers, and the RF generator's one or more frequency tuners. The digital signal processor is coupled to the processor 126, the one or more power controllers, and the one or more frequency tuners. The power controller is a controller that controls the power value output from the RF generator's RF power supply during different states, such as a high logic state and a low logic state, and the frequency tuner is a controller that controls the frequency value output from the RF generator's RF power supply during different states.

[0061] In one embodiment, the sensor 120 and high speed data acquisition device 121 measure the voltage of the RF signal 130 at the output O1. The measured voltage is provided from the voltage sensor to the processor 126 via a cable coupling the voltage sensor to the processor 126. The processor 126 determines the time when the measured voltage is zero during a positive crossing and determines the phase φ of the voltage of the RF signal 130.

[0062] In one embodiment, the sensor 120 is coupled to the output O1 of the X kilohertz RF generator 112 for a predetermined number of initial operating cycles of the X kilohertz RF generator 112 or the RF signal 130. After the predetermined number of initial cycles, the sensor 120 is disconnected from the output O1. The processor 126 determines that the phase of the RF signal 130 remains the same as determined by a measurement of the voltage of the RF signal 130 received from the sensor 120.

[0063] In one embodiment, sensor 131 is located outside of Y megahertz RF generator 114 and is coupled to output O2 of Y megahertz RF generator 114. In one embodiment, sensor 120 is located within X kilohertz RF generator 112.

[0064] It should be noted that in one embodiment, instead of applying the modified RF signal 134 to the lower electrode of the plasma excitation electrode 104 and coupling the upper electrode 106 to ground potential, the modified RF signal 134 is applied to the upper electrode 106 and the lower electrode of the plasma excitation electrode 104 is coupled to ground potential.

[0065] In one embodiment, instead of coupling the upper electrode 106 to ground potential, the upper electrode is coupled to one or more RF generators through an impedance matching network.

[0066] In one embodiment, a voltage and current probe (VI probe) is used instead of sensor 120 to measure complex voltages and currents, which include a current amplitude, a voltage amplitude, and a phase between the current amplitude and the voltage amplitude.

[0067] In one embodiment, a cycle of operation of the RF generator occurs to generate a cycle of the RF signal, e.g., during one cycle of operation of the RF generator described herein, one period of the RF signal is generated by the RF generator.

[0068] In one embodiment, the sensor 131 and high speed data acquisition device 123 measure complex voltages and complex currents, including voltage amplitudes, current amplitudes, and phases between the voltage and current amplitudes. The sensor 131 and high speed data acquisition device 123 provide the amplitudes and phases to the processor 126 via a cable. The processor 126 determines multiple values ​​of voltage reflection coefficients Γ11, Γ21 through Γn1 over one cycle of the RF signal 130 based on the voltage amplitudes, current amplitudes, and phases.

[0069] In one embodiment, both sensors 120 and 131 are coupled to the same high speed data acquisition device 121 or 123 .

[0070] In one embodiment, the terms value and amount are used interchangeably herein. For example, the terms value of a frequency modulation parameter and amount of a frequency modulation parameter are used interchangeably herein. Illustratively, an amount represents an amount of a frequency modulation parameter and a value represents another amount of a frequency modulation parameter.

[0071] 1B-1 is one embodiment of table 130 illustrating that the values ​​of voltage reflection coefficients Γ11, Γ21 through Γn1 are determined based on complex voltages and complex currents obtained at periodic time intervals, such as every 0.1 microseconds (μs) or every 0.2 μs, over an operating cycle of X kilohertz RF generator 112 (FIG. 1). For example, value Γ11 is calculated by processor 126 (FIG. 1) from complex voltages and complex currents measured 0.1 μs after a positive crossing in a cycle of RF signal 130 (FIG. 1), and value Γ12 is calculated by processor 126 (FIG. 1) from complex voltages and complex currents measured 0.2 μs after a positive crossing in a cycle of RF signal 130. Similarly, value Γn1 corresponds to complex voltages and complex currents measured 0.n microseconds after a positive crossing in a cycle of RF signal 130.

[0072] In one embodiment, each of the values ​​Γ11, Γ21 through Γn1 is an average calculated by the processor 126 over a set of multiple cycles of the RF signal 130. For example, the processor 126 calculates multiple values ​​Γ11A, Γ21A through Γn1A of the voltage reflection coefficient from the complex voltages and complex currents measured at the output O2 in the same manner that the values ​​Γ11, Γ21 through Γn1 are calculated. The value Γ11A is calculated from the complex voltages and complex currents measured 0.1 μs from the start of the first cycle of the RF signal 130, the value Γ21A is calculated from the complex voltages and complex currents measured 0.1 μs from the start of the second cycle of the RF signal 130, and so on until the value Γn1A is calculated. The second cycle of the RF signal 130 is contiguous to the first cycle of the RF signal 130. Processor 126 calculates value Γ11 as the average of values ​​Γ11A, Γ21A through Γn1A. Similarly, value Γn1 is the average of multiple values ​​of the voltage reflection coefficient, the values ​​of which are calculated from complex voltages and complex currents measured 0.n μs from the beginning of each corresponding cycle of RF signal 130.

[0073] In one embodiment, instead of calculating the voltage reflection coefficient values ​​Γ11, Γ21 through Γn1, measurements of the complex voltage and complex current over one cycle of the RF signal 130 are used by the processor 126.

[0074] 1B-2 is one embodiment of a graph 140 illustrating the correspondence between values ​​Γ11, Γ21 through Γn1 and time intervals during cycle 1 of RF signal 130 (FIG. 1). Graph 140 plots a voltage waveform 142 of RF signal 130 versus time t. Voltage waveform 142 is sinusoidal or substantially sinusoidal and oscillates over multiple cycles, such as cycle 1 and cycle 2. Cycle 2 is consecutive to cycle 1. For example, there are no other cycles of RF signal 130 between cycles 1 and 2.

[0075] The value Γ11 corresponds to a time interval between 0 μs and 0.1 μs. At 0.1 μs, voltage waveform 142 has a positive intersection. Similarly, the value Γ21 corresponds to a time interval between 0.1 and 0.2 μs, and so on, until the value Γn1 corresponds to a time interval between (n-1) μs and 0.n μs.

[0076] FIG. 1C is a diagram of an embodiment of an RF path model 150 illustrating the calculation of load impedance values ​​ZL11-ZLn1 from voltage reflection coefficient values ​​Γ11, Γ21-Γn1. RF path model 150 is a computer-based model that is generated or executed, or generated and executed, by processor 126 from at least a portion of RF path 102. For example, RF path model 150 includes a number of circuit elements that are connected in the same manner as a portion of RF path 102 or circuit components of RF path 102. Examples of circuit elements include capacitors, inductors, and resistors. Any two adjacent circuit elements of RF path model 102 are coupled to each other via connections in the same manner that corresponding two adjacent circuit components of the portion of RF path 102 are coupled to each other. To illustrate, when two capacitors of RF path 102 are coupled to each other in series, two capacitors of RF path model 150 are also coupled to each other in series. Each of the two capacitors in RF path model 150 has the same capacitance as a corresponding one of the two capacitors in RF path 102, or the two capacitors in RF path model 150 have the same total or combined capacitance as the total or combined capacitance of the two capacitors in RF path 102. As another example, when the two capacitors in RF path 102 are coupled in parallel with each other, the two capacitors in RF path model 150 are also coupled in parallel with each other, and the two capacitors in RF path model 150 have the same total or combined capacitance as the total or combined capacitance of the two capacitors in RF path 102.

[0077] As yet another example, the RF path model 150 has the same impedance or substantially the same impedance as the combined impedance of the portion of the RF path 102 or the circuit components of the RF path 102. The RF path model 150 has substantially the same impedance as the combined impedance of the portion of the circuit components of the RF path 102 when the impedance of the RF path model 150 is within a predetermined range from the combined impedance. The predetermined range is stored in the memory device 128 and accessed by the processor 126. As an example, the RF path model 150 is received by the processor 126 via an input device and executed by the processor 126. The RF path model 150 is generated by a user operating the input device 150. The RF path model 150 is generated by the user by using a computer program executed by the processor 126. As another example, the RF path model 150 is generated by the processor 126 and executed by the processor 126. As another example, the RF path model 150 includes several modules, each module including one or more resistors, or one or more capacitors, or one or more inductors, or a combination thereof. Illustratively, each module is represented as an equation involving a resistance and a reactance.

[0078] Examples of portions of RF path 102 include RF cable RFC2 (FIG. 1), or a branch of impedance matching network 110 (FIG. 1) coupled between input I2 (FIG. 1) of impedance matching network 110 and output of impedance matching network 110, or RF transmission line RFT1 (FIG. 1), or plasma excitation electrode 104 (FIG. 1). Additional examples of portions of RF path 102 include a combination of RF cable RFC2 and a branch of impedance matching network 110 coupled between input I2 of impedance matching network 110 and output of impedance matching network 110, or a combination of RF cable RFC2 and a branch of impedance matching network 110 and RF transmission line RFT1, or a combination of RF cable RFC2 and a branch of impedance matching network 110 and RF transmission line RFT1 and plasma excitation electrode 104.

[0079] Processor 126 provides values ​​Γ11, Γ21 through Γn1 to input In of RF path model 150 to generate load impedance values ​​ZL11 through ZLn1 at output Out of RF path model 150. For example, processor 126 propagates value Γ11 forward from input In through circuit elements of RF path model 150 to generate load impedance value ZL11 at output Out. Before and during forward propagation, RF path model 150 is initialized to have a capacitance Cknown1 and a radio frequency RFknown. Capacitance Cknown1 is the capacitance of a capacitor in a branch of impedance matching network 110 between input I2 and the output of impedance matching network 110 during processing of substrate S (FIG. 1A), and radio frequency RFknown is the operating frequency of Y megahertz RF generator 114 (FIG. 1A). Illustratively, radio frequency RFknown is the operating frequency of Y megahertz RF generator 114. Y megahertz RF generator 114 operates, by way of example, between 57 MHz and 63 MHz. To further illustrate, frequency RFknown varies over time intervals, such as between 0 and 0.1 μs, between 0.1 μs and 0.2 μs, etc. The values ​​Cknown1 and RFknown are provided as input to processor 134 by a user via an input device connected to processor 126 via an input / output interface, such as a serial interface, a parallel interface, a universal serial bus (USB) interface, etc. Examples of input devices include a mouse, a keyboard, a stylus, a keypad, a button, and a touch screen.

[0080] Continuing with the example, processor 126 determines impedance values ​​Z11, Z21-Zn1 at input In of RF path model 150 from values ​​Γ11, Γ21-Γn1. Illustratively, processor 126 accesses a correspondence, such as a mapping or one-to-one relationship, between impedance values ​​Z11 and values ​​Γ11 from memory device 128 to identify impedance value Z11. In a similar manner, impedance values ​​Z21-Zn1 at input In of RF path model 150 are determined by processor 126 from values ​​Γ21-Γn1.

[0081] Continuing with the example, processor 126 propagates impedance value Z11 through the circuit elements of RF path model 150 to calculate a value ZL11 at output Out of RF path model 150. Illustratively, processor 126 calculates a directional sum of value Z11 and the impedance values ​​of the circuit elements of RF path model 150 to calculate a load impedance value ZL11 at the output of RF path model 150. The impedance values ​​of the circuit elements are stored in memory device 128. In a similar manner, processor 126 determines load impedance values ​​ZL21-ZLn1 at output Out of RF path model 150 from impedance values ​​Z21-Zn1 at input In of RF path model 150.

[0082] In one embodiment, instead of receiving the voltage reflection coefficient values ​​Γ11, Γ21 to Γn1 at the input In of the RF path model 150, the complex voltage and complex current values ​​are received at the input In by the RF path model 150 from the sensor 131 and the high-speed data acquisition device 123 (FIG. 1A) coupled to the output of the Y MHz RF generator 114 (FIG. 1A). In this embodiment, the processor 126 does not need to calculate the voltage reflection coefficient values ​​Γ11, Γ21 to Γn1 from the complex voltage and complex current values. The load impedance values ​​ZL11 to ZLn1 are calculated by the processor 126 by propagating the complex voltage and complex current values ​​measured by the sensor 131 and the high-speed data acquisition device 123 through the RF path model 150. For example, the processor 126 accesses from the memory device 128 impedance values ​​or complex voltage and complex current values ​​associated with each of the circuit elements of the RF path model 150, calculates the directional sum of each of the complex voltage and complex current values ​​at the input In and the impedance or complex voltage and complex current values ​​associated with the circuit elements of the RF path model 150, and calculates the load impedance values ​​ZL11 to ZLn1 at the output Out of the RF path model 150.

[0083] FIG. 1D-1 shows the average power reflection coefficient Γ at the input of the RF path model 150. 2 FIG. 1 is a diagram of an embodiment of RF path model 150 illustrating the determination of operational parameters, including frequency parameters or frequency parameters and match network parameters, based on application of load impedance values ​​ZL11-ZLn1 at the output of RF path model 150 that minimizes . For example, processor 126 is provided with an input via an input device indicating a type of periodic waveform of RF signal 132 output from Y MHz RF generator 114 (FIG. 1A). By way of example, processor 126 is provided with an input via an input device, the input including initial values ​​of operational parameters described below. The input is provided by a user operating the input device. As another example, the operational parameters are provided by a user as input, and processor 126 initializes values ​​of the operational parameters. The type of periodic waveform or the operational parameters define the shape of the envelope of RF signal 132. For example, the type of periodic waveform or the operational parameters provide the shape of the peak-to-peak amplitude of RF signal 132. Examples of the type of periodic waveform or the shape of the periodic waveform include a sinusoidal waveform, a trapezoidal waveform, a sawtooth waveform, a rectangular waveform, and a square waveform. In this example, sine, trapezoidal, rectangular, sawtooth, and square are shapes of the envelope of the RF signal 132. In some cases, the terms periodic waveform and periodic function are used interchangeably herein. The power reflection coefficient is another example of a reflection parameter or reflection coefficient parameter.

[0084] Continuing with the example, values ​​ZL11-ZLn1 of the load impedance at the output of RF path model 150 are back-propagated through the circuit elements of RF path model 150 to determine multiple values ​​ZL1a-ZLna of the load impedance at the input of RF path model 150. Processor 126 calculates a directional sum of value ZL11 and the values ​​of the impedance of the circuit elements of the RF path model to determine value ZL1a, and calculates a directional sum of value ZL21 and the values ​​of the impedance of the circuit elements of RF path model 150 to determine value ZL2a. In a similar manner, value ZLna is determined by processor 126 from value ZLn1.

[0085] Continuing with the example, the processor 126 derives a value of the power reflection coefficient Γ at the input of the RF path model 150 from the values ​​of the load impedance ZL1a-ZLna at the input of the RF path model 150. 2 ~Γna 2 For example, the processor 126 accesses a correspondence, such as a mapping or link, between the voltage reflection coefficient values ​​Γ1a-Γna and the load impedance values ​​ZL1a-ZLna from the memory device 126 to identify or determine the voltage reflection coefficient values ​​Γ1a-Γna. Illustratively, the memory device 126 stores a correspondence between the values ​​Γ1a and ZL1a and another correspondence between the values ​​Γna and ZLna.

[0086] Continuing with the example, the processor 126 calculates the square of each of the voltage reflection coefficient values ​​Γ1a through Γna at the inputs of the RF path model 150 and calculates the power reflection coefficient value Γ1a at the input In of the RF path model 150. 2 ~Γna 2 The processor 126 determines the value of the power reflection coefficient Γa 2 ~Γna 2 The average of ΓAavmin is calculated. 2 The processor 126 generates the first average value ΓAavmin 2Determine values ​​for operating parameters such as yMHzavfreqA, yMHzfreqvariationA, thighA, tlowA, φrelativeA, and CA for which φrelativeA is calculated, where yMHzavfreqA is the average operating frequency at which the Y megahertz RF generator 114 operates, yMHzfreqvariationA is the variation in the average frequency, thighA is the high dwell time that the power or voltage level of the Y megahertz RF generator 114 remains at a high level, and tlowA is the time that the power or voltage level of the Y megahertz RF generator 114 remains at a high level. where φ is the low dwell time that the power or voltage level of 114 remains at a low level, φrelativeA is the relative phase of the RF signal 132 (FIG. 1A) output by the Y megahertz RF generator 114 compared to the phase of the RF signal 130 (FIG. 1A) output by the X kilohertz RF generator 112, and CA is the total capacitance applied to the branches of the impedance matching network 110 coupled between the input I2 of the impedance matching network 110 (FIG. 1A) and the output of the impedance matching network 110. The low level is a voltage or power level that is lower than the voltage or power level that corresponds to the high level. The high level may be referred to herein as a high state, and the low level may be referred to herein as a low state. Also, the low level is the bottom envelope or bottom boundary of the multiple values ​​of the RF signal, and the high level is the upper envelope or upper boundary of the multiple values ​​of the RF signal. The RF signal oscillates between the low level and the high level.

[0087] To illustrate the determination of values ​​for operational parameters such as yMHzavfreqA, yMHzfreqvariationA, thighA, tlowA, φrelativeA, and CA, processor 126 provides values ​​of the operational parameters to RF path model 150. By providing the values ​​to RF path model 150, processor 126 makes the values ​​available to RF path model 150. Once RF path model 150 receives the values ​​of the operational parameters from processor 126 and is characterized by the values, processor 126 back-propagates the values ​​of the load impedances ZL11-ZLn1 at the output of RF path model 150 to determine the values ​​of the voltage reflection coefficients Γ1a-Γna at the input of RF path model 150. Processor 126 accesses the values ​​of the operational parameters from memory device 126 to make the values ​​available and provides the values ​​to RF path model 150, for example, where the values ​​of the voltage reflection coefficients Γ1a-Γna are determined.

[0088] Continuing with the example, values ​​ZL11-ZLn1 of the load impedance at the output of RF path model 150 are back-propagated through the circuit elements of RF path model 150 to determine multiple values ​​ZL1x-ZLnx of the load impedance at the input of RF path model 150. Processor 126 calculates a directional sum of value ZL11 and the values ​​of the impedance of the circuit elements of RF path model 150 to determine value ZL1x, and calculates a directional sum of value ZL21 and the values ​​of the impedance of the circuit elements of RF path model 150 to determine value ZL2x. In a similar manner, value ZLnx is determined by processor 126 from value ZLn1.

[0089] Continuing with the example, the processor 126 derives the value of the power reflection coefficient Γ at the input of the RF path model 150 from the values ​​of the load impedance ZL1x to ZLnx at the input of the RF path model 150. 2 ~Γn1 2For example, the processor 126 accesses a correspondence, such as a mapping or link, between the voltage reflection coefficient values ​​Γ11 through Γn1 and the load impedance values ​​ZL1x through ZLnx from the memory device 126 to identify or determine the voltage reflection coefficient values ​​Γ11 through Γn1. Illustratively, the memory device 126 stores a correspondence between values ​​Γ11 and values ​​ZL1x and stores another correspondence between values ​​Γn1 and values ​​ZLnx.

[0090] Continuing with the example, the processor 126 calculates the square of each of the voltage reflection coefficient values ​​Γ11 through Γn1 at the input of the RF path model 150 and calculates the power reflection coefficient value Γ11 2 ~Γn1 2 The processor 126 determines the value of the power reflection coefficient Γ 2 ~Γn1 2 The average of Γ1avmin is then calculated. 2 Generate.

[0091] The processor 126 calculates the first average value Γavmin 2thigh1 is the high dwell time that the power or voltage level of Y megahertz RF generator 114 remains at a high level; and tlow1 is the time that the power or voltage level of Y megahertz RF generator 114 remains at a high level. y is the low dwell time that the power or voltage level of yMHz is at a low level, φrelative1 is the relative phase of the RF signal 132 (FIG. 1A) output by the Y megahertz RF generator 114 compared to the phase of the RF signal 130 (FIG. 1A) output by the X kilohertz RF generator 112, and C is the total (combined) capacitance of the branches of the impedance matching network 110 between the input I2 (FIG. 1A) of the impedance matching network 110 and the output of the impedance matching network 110. Illustrating the determination of values ​​such as yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C, of ​​the operating parameters, the processor 126 provides values ​​of the operating parameters to the RF path model 150. By providing the values ​​to the RF path model 150, the processor 126 makes the values ​​available to the RF path model 150. Once RF path model 150 receives and is characterized by the values ​​of the operational parameters from processor 126, processor 126 back-propagates the load impedance values ​​ZL11-ZLn1 at the output of RF path model 150 to determine the voltage reflection coefficient values ​​Γ11-Γn1 at the input of RF path model 150. Processor 126 accesses the values ​​of the operational parameters from memory device 126 and makes the values ​​available, e.g., provides the values ​​to RF path model 150 where the voltage reflection coefficient values ​​Γ11-Γn1 are determined.

[0092] The processor 126 calculates a second average value Γ of the power reflection coefficient at the input of the RF path model 150. 2is the first average value ΓAavmin of the power reflection coefficient at the input of the RF path model 150. 2 In response to the determination, processor 126 determines to apply frequency parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1 instead of applying yMHzavfreqA, yMHzfreqvariationA, thighA, tlowA, φrelativeA to Y megahertz RF generator 114, and determines to apply match network parameter C1 instead of applying match network parameter CA to impedance matching network 110. Once processor 126 has determined the values ​​yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1, processor 126 updates the initial values ​​of the operating parameters with the determined values ​​of the operating parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1.

[0093] The determined values ​​of the operating parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are applied during a next set of one or more cycles of operation of the X kilohertz RF generator 112 following a set of one or more cycles of operation of the X kilohertz RF generator 112 for which the operating parameters to be applied are determined by the processor 126, as follows: For example, the processor 126 receives the complex voltages and complex currents measured at the output O2 (FIG. 1A) of the Y megahertz RF generator 114 for ten cycles of operation of the X kilohertz RF generator 112, determines the operating parameters from the complex voltages and complex currents, and applies the operating parameters to the Y MHz RF generator 114 during the subsequent ten cycles of operation of the X kilohertz RF generator 112. The subsequent ten cycles occur after the amount of time it took the processor 126 to determine the operating parameters. Example amounts of time include a portion of ten cycles of operation of the X kilohertz RF generator 112, and one or more cycles that are contiguous to the ten cycles of operation of the X kilohertz RF generator 112. The one or more contiguous cycles are the operation of the X kilohertz RF generator 112. The processor 126 stores values ​​of operating parameters in the memory device 128, such as values ​​of yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1.

[0094] In one embodiment, processor 126 repeats the operation or the method of determining the operating parameters until the values ​​of the operating parameters converge within the predetermined range. For example, during 10 cycles following the 10 cycles during which the values ​​of the operating parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are determined, processor 126 controls Y megahertz RF generator 114 to apply the values ​​yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and the value C1, and controls one or more variable capacitors in a branch of impedance matching network 110 between input I2 (FIG. 1A) and the output of impedance matching network 110 to apply the value C1. During the 10 cycles following the 10 cycles in which the values ​​of the operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are applied, or during the same 10 cycles in which the values ​​of the operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are applied, processor 126 determines the values ​​of the operational parameters yMHzavfreqVAL1, yMHzfreqvariationVAL1, thigh1, tlow1, φrelative1, and CVAL1 in the same manner as the values ​​yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are determined except that processor 126 initiates RF path model 150 to have a radio frequency that is the directional sum of the values ​​yMHzavfreq1 and yMHzfreqvariation1, and has a capacitance C1. The processor 126 determines, for the remaining values ​​thighVAL1, tlowVAL1, φrelativeVAL1, and CVAL1 and thigh1, tlow1, φrelative1, and C1, whether the value yMHzavfreqVAL1 is within a predetermined range of the value yMHzavfreq1 or whether the value yMHzfreqvariationVAL1 is within a predetermined range of the value yMHzfreqvariation1, etc.Upon determining that for the remaining values ​​thighVAL1, tlowVAL1, φrelativeVAL1, and CVAL1 and thigh1, tlow1, φrelative1, and C1, the value yMHzavfreqVAL1 is within a predetermined range from the value yMHzavfreq1, the value yMHzfreqvariationVAL1 is within a predetermined range from the value yMHzfreqvariation1, etc., processor 126 determines convergence of the values ​​of the operating parameters and does not further determine values ​​of the operating parameters.

[0095] However, continuing with the embodiment, upon determining that for the remaining values ​​thighVAL1, tlowVAL1, φrelativeVAL1, and CVAL1 and thigh1, tlow1, φrelative1, and C1, that the value yMHzavfreqVAL1 is not within the predetermined range of the value yMHzavfreq1 or that the value yMHzfreqvariationVAL1 is not within the predetermined range of the value yMHzfreqvariation1, etc., the processor 126 continues to apply the operations and methods for further operating cycles of the X kilohertz RF generator 112 to determine further values ​​of the operating parameters. Processor 126 applies the values ​​yMHzavfreqVAL1, yMHzfreqvariationVAL1, thighVAL1, tlowVAL1, φrelativeVAL1, and CVAL1 during the next 10 cycles following the 10 cycles in which the values ​​yMHzavfreqVAL1, yMHzfreqvariationVAL1, thighVAL1, tlowVAL1, φrelativeVAL1, and CVAL1 are determined in the same manner that processor 126 applies the values ​​of operating parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlowVAL1, φrelativeVAL1, and CVAL1 as described below with reference to FIG. 1D-3.

[0096] It should be noted that the values ​​of the operating parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1 are for a trapezoidal waveform. In one embodiment, after applying the method described herein, different types of values ​​are calculated for different types of waveforms output by the Y megahertz RF generator described herein. For example, when the type of waveform is sinusoidal, the operating parameters include average frequency, frequency variation such as frequency amplitude from the average frequency, period equal to the period of the RF signal generated by the X kHz RF generator described herein, and phase relative to the phase of the RF signal generated by the X kHz RF generator. As another example, when the type of waveform is rectangular, the operating parameters include high frequency value, high frequency time during which the high frequency value is maintained, low frequency value, low frequency time during which the low frequency value is maintained, period equal to the period of the RF signal generated by the X kHz RF generator described herein, and phase relative to the phase of the RF signal generated by the X kHz RF generator. The high frequency value is greater than the low frequency value. As yet another example, when the waveform type is triangular, the operating parameters include a high frequency value, a low frequency value, a first ramp rate or first ramp time between the high frequency value and the low frequency value, a second ramp rate or second ramp time between the low frequency value and the high frequency value, a period equal to the period of an RF signal generated by an X kHz RF generator as described herein, and a phase relative to the phase of an RF signal generated by an X kHz RF generator. The high frequency value is greater than the low frequency value. The first ramp rate is lower, the same as, or greater than the second rate.

[0097] In one embodiment, the match network parameter C1 is not determined by the processor 126, and the determination of the match network parameter C1 is optional.

[0098] In one embodiment, the terms frequency parameter and frequency modulation parameter are used interchangeably herein.

[0099] FIG. 1D-2 is one embodiment of a graph 170 illustrating multiple cycles of an RF signal 172 generated by an X kilohertz RF generator described herein. For example, the RF signal 172 is an example of the RF signal 130 of FIG. 1A. The graph 170 plots the voltage of the RF signal 172 versus time t. The RF signal 172 is a voltage waveform that represents the voltage output by the X kilohertz RF generator. The RF signal 172 is sinusoidal or substantially sinusoidal, and periodically oscillates between positive and negative values ​​of voltage. For example, the RF signal 172 has a positive intersection point PC1, which is the time when the value of the voltage waveform goes from negative to positive. Similarly, after a period of time, the RF signal 172 has another positive intersection point PC2, which is another time when the value of the voltage waveform goes from negative to positive.

[0100] Cycle 1 of RF signal 172 represents a period of occurrence of the voltage waveform. Another cycle 2 of RF signal 172 represents another example of a period of occurrence of the voltage waveform. Cycle 2 is contiguous with cycle 1. For example, there is no cycle between cycles 1 and 2. Additionally, additional cycles 3 through (m-1) of RF signal 172 are contiguous with cycle 2. Cycle 3 is contiguous with cycle 2. Cycles 1 through (m-1) form set 1 of cycles of RF signal 172, where m is an integer greater than 2.

[0101] Further, cycles m, (m+1), and so on, where q is an integer greater than 1, until (m+q) form set 2 of cycles of RF signal 172. Set 2 is contiguous to set 1. For example, there is no set of cycles between sets 1 and 2. The number of cycles m~(m+q) is the same as the number of cycles 1~(m-1) of set 1. As an example, the above-mentioned 10 cycles in which complex voltages and complex currents are measured at output O2 of Y megahertz RF generator 114 (FIG. 1A) to determine values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are applied, is an example of cycles m~(m+q) of set 2.

[0102] In one embodiment, instead of a voltage waveform, a power waveform of the RF signal 172 is used.

[0103] FIG. 1D-3 is a diagram of an embodiment of a system 180, such as a plasma tool, that applies values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 during set 2 of cycles of operation of X kilohertz RF generator 112 during processing of substrate S. System 180 is similar in structure and function to system 100, except that system 180 excludes high speed data acquisition device 123, sensor 131, and high speed data acquisition device 121, and includes voltage sensor 182, comparator 184, motor system 186, and driver system 188. An example of comparator 184 includes a processor or application specific integrated circuit or programmable logic device. An example of motor system 186 includes one or more electric motors, each having a stator and a rotor. An example of driver system 188 includes one or more transistors coupled together to output one or more current signals upon receiving one or more signals from motor system 186. The comparator 184 is coupled to the voltage sensor 182 and is also coupled to the processor 126. The voltage sensor 182 is coupled to the output O1 of the X kilohertz RF generator 112.

[0104] The driver system 188 is coupled to the processor 126, and the motor system 186 is coupled to the driver system 188. The motor system 186 is coupled to one or more circuit components of the impedance matching network 110 via one or more corresponding connection components. An example of each connection component may include one or more rods, or a combination of one or more rods and one or more gears.

[0105] The RF signal 130 is generated by the X kilohertz RF generator 112 in a manner similar to that described above with reference to FIG. 1A and transmitted to the input I1 of the impedance matching network 110 via the output O1 of the X kilohertz RF generator 112 and the RF cable RFC1. For example, set 2 of cycles of the RF signal 130 is output from the X kilohertz RF generator 112. The voltage sensor 182 measures the voltage of the RF signal 130 at the output O1 of the X kilohertz RF generator 112. The comparator 184 compares the voltage measured by the voltage sensor 182 with a value of zero to output a comparison result and provides the comparison result to the processor 126. The comparison result provides whether the voltage is above or below zero or zero, e.g., whether the voltage is positive, negative, or zero. The processor 126 determines from the comparison result a number of times or instances at which the voltage is zero and is likely to go from negative to positive. Based on the number of times or instances determined by the processor 126, the phase of the RF signal 130 is determined by the processor 126.

[0106] Additionally, the processor 126 accesses the values ​​of the frequency parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1 from the memory device 128 and provides the values ​​to the RF power supply 124. The relative phase φrelative1 is determined by the processor 126 from the phase of the RF signal 130 determined during processing of the substrate S. Upon receiving the frequency parameters, the RF power supply 124 generates an RF signal 190 having values ​​yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, and φrelative1. For example, the RF signal 190 has an envelope or peak-to-peak amplitude that is trapezoidal or pulse-shaped or square-wave or sinusoidal. The relative phase φrelative1 of the RF signal 190 is relative to the phase of the RF signal 130, such as a delay, determined using the voltage sensor 182, the comparator 184, and the processor 126. The RF signal 190 is itself a sinusoidal or substantially sinusoidal signal. An RF signal 190 is provided by RF power supply 124 via output O2 and RF cable RFC2 to input I2 of impedance matching network 110.

[0107] Further, the processor 126 accesses the value C1 from the memory device 128, generates a command signal, and transmits the command signal to the driver system 188. An example of a command signal includes one or more amounts of current output by the driver system 188 to control the motor system 186 to achieve the combined capacitance of the branches of the impedance matching network 110 between the input I2 and the output of the impedance matching network 110. Upon receiving the command signal, the driver system 186 generates one or more current signals and transmits the current signals to the motor system 186.

[0108] The motor system 186 operates according to one or more current signals to control one or more variable capacitors in a branch between the input I2 and the output of the impedance matching network 110 to achieve a capacitance C1 of the branch of the impedance matching network 110. For example, the motor of the motor system 186 operates to rotate or move a plate of a variable capacitor in the impedance matching network 110 in a linear direction to change the distance or area between a plate of the capacitor and another plate to change the capacitance of the capacitor and achieve the capacitance C1 of the branch between the input I2 and the output of the impedance matching network 110.

[0109] An impedance matching network 110 having a capacitance C1 receives the RF signals 130 and 190 and processes the RF signals 130 and 190 in the manner described above with respect to the RF signals 130 and 132 (FIG. 1A) to output a modified RF signal 192. For example, the impedance matching network 110 matches the impedance of a load coupled to the output of the impedance matching network 110 with the impedance of a source coupled to the inputs I1 and I2 of the impedance matching network 110 to output a modified RF signal 192. When one or more process gases are supplied to the gap between the upper electrode 106 and the lower electrode of the plasma excitation electrode 104 and the lower electrode of the plasma excitation electrode 104 receives the modified RF signal 192 from the output of the impedance matching network 110 via the RF transmission line RFT1, a plasma is struck or sustained in the plasma chamber 108 to process the substrate S during a cycle of set 2 of operation of the X kilohertz RF generator 112. An example of a process gas includes an oxygen-containing gas such as O2. Other examples of process gases include fluorine-containing gases, such as tetrafluoromethane (CF4), sulfur hexafluoride (SF6), hexafluoroethane (C2F6), and the like.

[0110] In one embodiment, comparator 184 is part of processor 126 and the functions described herein as being performed by comparator 184 are performed by processor 126 .

[0111] In embodiments in which the capacitance C1 is not determined, the processor 126 does not control the impedance matching network 110 to achieve the capacitance C1.

[0112] In one embodiment, sensor 120 (FIG. 1A) is decoupled from output O1 of X kilohertz RF generator 112 after the values ​​of operating parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are determined or determined after convergence by processor 126. Sensor 131 is also decoupled from output O2 of Y megahertz RF generator 114.

[0113] Figure 2A is one embodiment of a graph 200 illustrating an operating cycle of an RF signal 202. RF signal 202 is an example of RF signal 130 of Figure 1A. Graph 200 plots the voltage of RF signal 130 versus time t in microseconds.

[0114] 2B is one embodiment of a graph 206 illustrating a forward voltage waveform 208 of an RF signal provided by a Y MHz RF generator as described herein. For example, the forward voltage waveform 208 is an example of the forward voltage of the RF signal 132 of FIG. 1A. The graph 206 plots the forward voltage or voltage provided at the output of the Y MHz RF generator versus time t in microseconds during one cycle of the RF signal 202 (FIG. 2A). It should be noted that the forward voltage waveform 208 has multiple cycles during one cycle of the RF signal 202 provided by the X kilohertz RF generator, as described above with reference to FIG. 2A.

[0115] 2C is one embodiment of a graph 210 illustrating a reverse voltage waveform 212 of an RF signal provided by a Y MHz RF generator as described herein. For example, the reverse voltage waveform 212 is an example of the reverse voltage of the RF signal 132 of FIG. 1A. The graph 210 plots the reverse voltage or voltage reflected toward the output of the RF generator versus time t in microseconds during one cycle of the RF signal 202 (FIG. 2A). It should be noted that the reverse voltage waveform 212 has multiple cycles during one cycle of the RF signal 202 provided by the X kilohertz RF generator, as described above with reference to FIG. 2A.

[0116] The forward voltage waveform 208 of FIG. 2B and the reverse voltage waveform 212 of FIG. 2C show that the voltage reflection coefficient is a complex number having an amplitude and a phase. The forward voltages have a similar peak-to-peak amplitude or substantially constant envelope. The peak-to-peak amplitude or envelope of the reverse voltage is periodically low and periodically high. The forward and reverse voltages together produce a complex voltage reflection coefficient. The complex number includes an amplitude and a phase.

[0117] 2D is an embodiment of a graph 214 illustrating the forward voltage waveform 208 and the reverse voltage waveform 212 on one graph. Again, the variation in peak-to-peak amplitude of the forward voltage waveform 208 and the reverse voltage waveform 212 produces a complex voltage reflection coefficient.

[0118] 2E is one embodiment of a Smith chart 220 illustrating values ​​of the voltage reflection coefficient at the output of a Y MHz RF generator described herein. For example, the Smith chart 220 illustrates values ​​of the voltage reflection coefficient at the output O2 of the Y MHz RF generator 114 (FIG. 1A) when the methods described herein are not applied to one cycle of operation of the X kilohertz RF generator 112 (FIG. 1A). For example, as shown in FIG. 2E, most values ​​of the voltage reflection coefficient are not close to the center of the Smith chart 220. Thus, a large amount of voltage is reflected back toward the Y megahertz RF generator without applying the methods described herein.

[0119] 3A is an illustration of one embodiment of a graph 300 illustrating a comparison between the frequency of an RF signal output from a Y megahertz RF generator described herein after applying a method described herein and the RF signal output from the Y megahertz RF generator before applying the method. Graph 300 plots a frequency 302 of an RF signal and another frequency 304 of another RF signal against time t in microseconds. Frequency 302 is for the RF signal output by the Y megahertz RF generator before applying a method described herein, and frequency 304 is for the RF signal output by the Y megahertz RF generator after applying the method.

[0120] Frequency 304 is an example of a frequency of RF signal 190 (FIG. 1D-3) after applying the method described herein, and frequency 302 is the frequency of RF signal 132 before applying the method. Frequency 302 is the frequency or peak-to-peak amplitude of an envelope of an RF signal output from a Y megahertz RF generator, and frequency 304 is the frequency or peak-to-peak amplitude of another envelope of an RF signal output from a Y megahertz RF generator. As shown, the envelope of the RF signal represented by frequency 302 is substantially constant or substantially the same. On the other hand, the envelope of the RF signal represented by frequency 304 has a trapezoidal shape.

[0121] The graph 300 also illustrates a frequency parameter of the RF signal output from the Y megahertz RF generator after applying the method described herein. For example, the frequency parameter 306 is an example of a value yMHzavfreq1. To illustrate, the frequency parameter 306 is the average frequency of the RF signal 190 during a cycle or set of cycles of the X MHz RF generator 112. As another example, the frequency parameter 308 is an example of a value yMHzfreqvariation1. To illustrate, the frequency parameter 308 is a value of a positive or negative variation from the average frequency of the RF signal output from the Y megahertz RF generator after applying the method described herein. A positive direction is a direction in which the voltage value of the RF signal output from the Y megahertz RF generator is positive compared to the average frequency yMHzavfreq1, and a negative direction is a direction in which the voltage value of the RF signal output from the Y megahertz RF generator is negative compared to the average frequency yMHzavfreq1. As yet another example, the frequency parameter 310 is an example of a value thigh1. Illustratively, frequency parameter 310 is the duration during which the frequency of an RF signal output from a Y megahertz RF generator after application of a method described herein is high, e.g., within a predetermined high frequency range. As yet another example, frequency parameter 312 is an example of a value tlow1. Illustratively, frequency parameter 312 is the duration during which the frequency of an RF signal output from a Y megahertz RF generator after application of a method described herein is low, e.g., within a predetermined low frequency range.

[0122] FIG. 3B is an illustration of one embodiment of a Smith chart 350 illustrating values ​​of voltage reflection coefficients at the output of a Y megahertz RF generator described herein after applying the methods described herein. For example, Smith chart 350 shows values ​​of voltage reflection coefficients at output O2 of Y megahertz RF generator 114 of FIG. 1D-3 when RF signal 190 (FIG. 1D-3) is generated by Y megahertz RF generator 114. Smith chart 350 is plotted for one cycle of operation of an X kilohertz RF generator, such as X kilohertz RF generator 112 (FIG. 1A) described herein. As shown in FIG. 3B, most of the values ​​of voltage reflection coefficients are closer to the center of Smith chart 350 compared to the values ​​of voltage reflection coefficients shown in Smith chart 220 of FIG. 2E.

[0123] FIG. 4A is a diagram of one embodiment of a system 400, such as a plasma tool, illustrating that the method described above with respect to FIG. 1A can be applied to the system 400 during recipe development, rather than during processing of a substrate S (FIG. 1A). The system 400 is similar in structure and function to the system 100 of FIG. 1A, except that in the system 400, instead of processing a substrate S, a dummy substrate 402 is used for recipe development. For example, the dummy substrate 402 is placed on top of the plasma excitation electrode 104 in the plasma chamber 108. The X kilohertz RF generator 112 generates the RF signal 130, and the Y megahertz RF generator 114 generates the RF signal 132. The RF signals 130 and 132 are modified by the impedance matching network 110 in the manner described above to generate a modified RF signal 134. The modified RF signal 134 is supplied to a bottom electrode embedded in the plasma excitation electrode 104 to generate a plasma in the plasma chamber 108, but without processing of the dummy substrate 402. For example, one or more process gases are not supplied to the plasma chamber 108 during recipe development.

[0124] Sensor 120 and high speed data acquisition device 121 measure the voltage at output O1 to produce a voltage waveform that is analyzed by processor 126 in the manner described above with reference to Figure 1A to determine the phase of RF signal 130 provided by X kilohertz RF generator 112. Sensor 131 and high speed data acquisition device 123 also measure complex voltages and currents at output O2 that are analyzed by processor 126 to determine voltage reflection coefficient values ​​Γ11, Γ21 through Γn1 for one cycle of operation of X kilohertz RF generator 112 in the manner described above with reference to Figure 1A.

[0125] 4B is an illustration of one embodiment of table 130 generated by processor 126 (FIG. 4A) during recipe development in the same manner that table 130 of FIG. 1B-1 is generated by processor 126 during processing of substrate S (FIG. 1A). Processor 126 determines the values ​​of voltage reflection coefficients Γ11, Γ21 through Γn1 for one cycle of operation of X kilohertz RF generator 112 (FIG. 4A) during recipe development.

[0126] As described above, in one embodiment, instead of determining each value Γ11, Γ21 through Γn1 for one cycle of operation of the X kilohertz RF generator 112, each value of the voltage reflection coefficient Γ11, Γ21 through Γn1 is an average of values ​​of the voltage reflection coefficient determined by processor 126 from complex voltages and complex currents measured at output O2 during a set of cycles of operation of the X kilohertz RF generator 112 during recipe development.

[0127] 4C is a diagram of one embodiment of RF path model 150 illustrating the generation of load impedance values ​​ZL11-ZLn1 from voltage reflection coefficient values ​​Γ11, Γ21-Γn1 for or during recipe development. The load impedance values ​​ZL11-ZLn1 are generated by processor 126 (FIG. 4A) from values ​​Γ11, Γ21-Γn1 in the same manner as described above with respect to FIG. 1C, except that the load impedance values ​​are generated during recipe development rather than during processing of substrate S (FIG. 1A).

[0128] 4D is a diagram of one embodiment illustrating the determination of values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 for or during recipe development. The values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 for or during recipe development are determined in the same manner as described above with reference to FIG. 1D-1 in which the average power reflection coefficient is minimized.

[0129] In one embodiment, processor 126 determines values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 for or during recipe development to minimize both the average power reflection coefficient and the average voltage reflection coefficient. For example, processor 126 determines values ​​of voltage reflection coefficients Γ1a-Γna and power reflection coefficients Γ1a-Γn at the inputs of RF path model 150 for load impedance values ​​ZL11-ZLn1 at the outputs of RF path model 150 in the same manner as described above with reference to FIG. 1D-1. 2 ~Γna 2 Determine the voltage reflection coefficients Γ1a to Γna and the power reflection coefficient Γ1a 2 ~Γna 2 Prior to determining I2, RF path model 150 is initialized by processor 126 to have a capacitance Cknown1 and a radio frequency RFknown, where Cknown1 is the capacitance of the capacitor in the branch of impedance matching network 110 between input I2 of impedance matching network 110 and the output of impedance matching network 110, and RFknown is the value at which Y megahertz RF generator 114 is operating during recipe development.

[0130] The processor 126 calculates the value of the power reflection coefficient Γ at the input of the RF path model 150. 2 ~Γna 2 The average of the power reflection coefficients is then calculated as ΓAavmin 2The processor 126 also calculates the average of the values ​​Γ1a to Γna of the voltage reflection coefficients at the input of the RF path model 150 to generate a first average value ΓAavmin of the voltage reflection coefficients. The processor 126 calculates the first average value ΓAavmin 2 and determine values ​​for operating parameters such as yMHzavfreqA, yMHzfreqvariationA, thighA, tlowA, φrelativeA, and CA for which ΓAavmin is calculated, where yMHzavfreqA is the average operating frequency at which the Y megahertz RF generator (FIG. 4E) operates during processing of the substrate, yMHzfreqvariationA is the variation of the average frequency, thighA is the high dwell time that the power level or voltage level of the Y megahertz RF generator remains at a high level, tlowA is the low dwell time that the power level or voltage level of the Y megahertz RF generator remains at a low level, φrelativeA is the relative phase of the RF signal output by the Y megahertz RF generator compared to the phase of the RF signal output by the X kilohertz RF generator used to process the substrate, and CA is the total capacitance applied to a branch of an impedance matching network used to process the substrate. The branch of the impedance matching network is between an input of the impedance matching network coupled to the Y megahertz RF generator and an output of the impedance matching network.

[0131] Additionally, the processor 126 may calculate the power reflection coefficient values ​​Γ11 through Γn1 at the input of the RF path model 150 for the load impedance values ​​ZL11 through ZLn1 at the output of the RF path model 150 in the same manner as described above with reference to FIG. 1D-1. 2 ~Γn1 2 and determine the values ​​of the voltage reflection coefficients Γ11 to Γn1. 2 ~Γn1 2 Prior to determining the values ​​of the voltage reflection coefficients Γ11 through Γn1, the RF path model 150 is initialized by the processor 126 to have a capacitance Cknown1 and a radio frequency RFknown.

[0132] The processor 126 calculates the value of the power reflection coefficient Γ 2 ~Γn1 2 The average of is further calculated to obtain the second average value Γ1avmin 2 and also calculates the average of the voltage reflection coefficient values ​​Γ11 through Γn1 to generate a second average value Γ1avmin of the voltage reflection coefficient at the input of the RF path model 150. The processor 126 calculates the second average value Γ1avmin 2 and determining values ​​for operating parameters such as yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 for which Γavmin is calculated, where yMHzavfreq1 is the average operating frequency at which the Y megahertz RF generator operates to process the substrate, yMHzfreqvariation1 is the variation of the average frequency, thigh1 is the high dwell time that the power level or voltage level of the Y megahertz RF generator remains at a high level, tlow1 is the low dwell time that the power level or voltage level of the Y megahertz RF generator remains at a low level, φrelative1 is the relative phase of the RF signal output by the Y megahertz RF generator compared to the phase of the RF signal output by the X kilohertz RF generator to process the substrate, and C1 is the total capacitance applied to the branches of the impedance matching network used to process the substrate.

[0133] The processor 126 calculates a second average value Γ of the power reflection coefficient at the input of the RF path model 150. 2 is the first average value of the power reflection coefficient at the input of the RF path model 150, 2and determines that the second average value Γ1avmin of the voltage reflection coefficient at the input of the RF path model 150 is less than or lower than the first average value ΓAavmin of the voltage reflection coefficient at the input of the RF path model 150. In response to the determination, the processor 126 determines to apply frequency parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1 instead of applying yMHzavfreqA, yMHzfreqvariationA, thighA, tlowA, φrelativeA to the Y megahertz RF generator, and determines to apply match network parameter C1 instead of applying match network parameter CA to an impedance matching network used to process the substrate. The processor 126 stores the values ​​of the operational parameters, such as values ​​yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 in the memory device 128 for later access by the processor 464 (FIG. 4E) from another memory device 468 (FIG. 4E).

[0134] 4E is a diagram of an embodiment of a system 450, such as a plasma tool, illustrating application of values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 determined during recipe development. The values ​​of operational parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1 are applied during processing of a substrate SU in a plasma chamber 458. As an example, the processing of the substrate SU is performed by an entity performing recipe development and is performed after recipe development. The system 450 includes an X kilohertz RF generator 452, a Y megahertz RF generator 454, a plasma chamber 458, an impedance matching network 456, a computer 463, a driver system 470, a motor system 472, a comparator 480, and a voltage sensor 482.

[0135] The X kilohertz RF generator 452 is similar in structure and function to that of the X kilohertz RF generator 112 (FIG. 4A) and includes an RF power source 474. For example, the X kilohertz RF generator 452 is a 400 kHz RF generator. The RF power source 474 is also similar in structure and function to the RF power source 122 of the X kilohertz RF generator 112 (FIG. 4A). Illustratively, the RF power source 474 is an RF oscillator.

[0136] Similarly, the Y megahertz RF generator 454 is similar in structure and function to that of the Y megahertz RF generator 114 (FIG. 4A) and includes an RF power source 476. For example, the Y megahertz RF generator 454 is a 60 MHz RF generator. The RF power source 476 is also similar in structure and function to the RF power source 124 (FIG. 4A) of the Y megahertz RF generator 114. Illustratively, the RF power source 476 is an RF oscillator.

[0137] The comparator 480 is similar in structure and function to the comparator 184 of FIG.

[0138] The impedance matching network 456 is also similar in structure and function to that of the impedance matching network 110 of Figure 4A. For example, the branch of the impedance matching network 456 between the input I21 of the impedance matching network 456 and the output of the impedance matching network 456 has a combined impedance that is within a predetermined range from the combined impedance of the branch between the input I2 and the output of the impedance matching network 110. As an example, as described herein, the combined impedance of the branch of the impedance matching network is the combination or sum of the impedances of all the electrical components of the branch of the impedance matching network.

[0139] Motor system 472 is similar in structure and function to motor system 186 (FIG. 1D-3), and driver system 470 is also similar in structure and function to driver system 188 (FIG. 1D-3). For example, motor system 72 includes one or more electric motors, and driver system 70 includes a network of transistors.

[0140] Examples of computers, such as computer 463, are provided above. Computer 463 includes a processor 464 and a memory device 468. Examples of processors, such as processor 464, are provided above. Examples of memory devices, such as memory device 468, are provided above. Processor 464 is coupled to memory device 468 and is also coupled to RF power sources 474 and 476.

[0141] The plasma chamber 458 includes a plasma excitation electrode 462 and an upper electrode 460 coupled to a ground potential. The plasma excitation electrode 462 is similar in structure and function to the plasma excitation electrode 104 of the plasma chamber 108 (FIG. 4A). For example, the lower electrode of the plasma excitation electrode 462 is made of aluminum or an alloy of aluminum. Also, the upper electrode 460 is similar in structure and function to the upper electrode 106 of FIG. 4A, and is placed on the plasma excitation electrode 462, forming a gap between the upper electrode 460 and the plasma excitation electrode 462. For example, the upper electrode 460 is made of aluminum or an alloy of aluminum.

[0142] An RF power source 474 of the X kilohertz RF generator 452 is coupled to an input I11 of the impedance matching network 456 via an output O11 of the X kilohertz RF generator 452 and an RF cable RFC11. In a similar manner, an RF power source 476 of the Y megahertz RF generator 454 is coupled to an input I21 of the impedance matching network 456 via an output O21 of the Y megahertz RF generator 454 and an RF cable RFC21.

[0143] An output of the impedance matching network 456 is coupled to the lower electrode of the plasma excitation electrode 462 via RF transmission line RFT11. The processor 464 is coupled to a driver system 470 and a comparator 480. The drive system 470 is coupled to a motor system 472, which is connected to the impedance matching network 456 via one or more connection components, examples of which are provided above. A voltage sensor 482 is coupled to the output O11 of the X kilohertz RF generator 452 and is coupled to the comparator 480.

[0144] The RF power supply 474 of the X kilohertz RF generator 452 generates one or more cycles of an RF signal 484 and provides the RF signal 484 to an input I11 of the impedance matching network 456 via an output O11 and an RF cable RFC11. The voltage sensor 482 measures the voltage of the RF signal 484 at the output O11 of the X kilohertz RF generator 452. The comparator 480 compares the voltage measured by the voltage sensor 482 with a value of zero, outputs a comparison result, and provides the comparison result to the processor 464. The comparison result provides whether the voltage is above or below zero or zero, e.g., whether the voltage is positive, negative, or zero. The processor 464 determines from the comparison result a number of times or instances at which the voltage is zero and is likely to go from negative to positive. Based on the times or instances determined by the processor 464, the phase of the RF signal 484 is determined by the processor 464.

[0145] 4A. For example, the values ​​of the operating parameters are received by processor 464 of computer 463 from processor 126 of computer 118 over a computer network, such as the Internet or an intranet or a combination thereof. Processor 464 stores values ​​of the operating parameters, such as yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1, and C1, in memory device 468 of computer 463.

[0146] Further, the processor 464 provides the RF power supply 476 with values ​​of the frequency parameters yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, φrelative1. The relative phase φrelative1 is determined by the processor 464 from the phase of the RF signal 484 determined during processing of the substrate SU. Upon receiving the frequency parameters, the RF power supply 476 generates an RF signal 486 having values ​​yMHzavfreq1, yMHzfreqvariation1, thigh1, tlow1, and φrelative1. For example, the RF signal 486 has an envelope or peak-to-peak amplitude that is trapezoidal or pulse-shaped or square-wave or sinusoidal. The RF signal 486 is itself a sinusoidal signal or a substantially sinusoidal signal. The RF signal 486 is supplied by the RF power supply 476 to the input I21 of the impedance matching network 456 via the output O21 and the RF cable RFC21. The RF signal 486 is generated for one or more cycles of the RF signal 484.

[0147] Further, the processor 464 generates a command signal and transmits the command signal to the driver system 470. An example of a command signal is one that includes one or more amounts of current output by the driver system 470 to control the motor system 472 to achieve the combined capacitance of the branches between the input I21 and the output of the impedance matching network 456. Upon receiving the command signal, the driver system 470 generates one or more current signals and transmits the current signals to the motor system 472.

[0148] The motor system 472 operates according to the one or more current signals to control one or more variable capacitors of the impedance matching network 456 to achieve the capacitance C1 of the impedance matching network 456. For example, the motor of the motor system 472 operates to rotate or move a plate of a capacitor in the impedance matching network 456 in a linear direction to change the area or distance between one plate of the capacitor and another plate to change the capacitance of the capacitor and achieve the capacitance C1 of the impedance matching network 456.

[0149] An impedance matching network 456 having a capacitance C1 receives the RF signals 484 and 486 and processes the RF signals 484 and 486 in the manner described above with respect to the RF signals 130 and 132 (FIG. 4A) to output a modified RF signal 488. For example, the impedance matching network 456 matches the impedance of a load coupled to the output of the impedance matching network 456 to the impedance of a source coupled to the inputs I11 and I21 of the impedance matching network 456 and outputs the modified RF signal 488. Examples of loads coupled to the output of the impedance matching network 456 include the plasma chamber 458 and the RF transmission line RFT11. Examples of sources coupled to the inputs I11 and I21 include the X kilohertz RF generator 452, the Y megahertz RF generator 454, and the RF cables RFC11 and RFC21.

[0150] The lower electrode of the plasma excitation electrode 462 receives a modified RF signal 488 from the output of the impedance match network 456 via RF transmission line RFT11. In addition to receiving the modified RF signal 488, one or more process gases are received by the plasma chamber 458 to strike or sustain a plasma within the plasma chamber 458. The plasma facilitates processing of the substrate SU during one or more cycles of operation of the X kilohertz RF generator 452.

[0151] In one embodiment, the method described herein with reference to FIG. 4E is used to process multiple substrates in plasma chamber 458 or in a different system similar in structure and function to system 450.

[0152] In one embodiment, a baseline frequency is used or determined instead of the average frequency of the Y megahertz RF generator. Examples of baseline frequencies include average and median frequencies.

[0153] In one embodiment, during processing of substrate SU, processor 126 continues to apply frequency parameters determined during recipe development to the baseline frequency of Y megahertz RF generator 454. For example, during processing of substrate S, values ​​yMHzfreqvariation1, thigh1, tlow1, and φrelative1 are applied by processor 126 to the average operating frequency of Y megahertz RF generator 454. To illustrate, processor 126 achieves a relative phase of φrelative1 between RF signal 486 generated by Y MHz RF generator 454 and RF signal 484 generated by X MHz RF generator 452 while modifying the average frequency of Y MHz RF generator 454 during processing of substrate SU by values ​​yMHzfreqvariation1, thigh1, and tlow1.

[0154] In an embodiment, the baseline frequency of the Y megahertz RF generator 454 is determined by the processor 464 by applying one of many methods. For example, a sensor, such as a complex voltage and complex current sensor, is coupled to the output O21 of the Y megahertz RF generator 454. The sensor is also coupled to the processor 464. The sensor coupled to the processor 464 measures the complex voltage and current and provides the complex voltage and current to the processor 464. The processor 464 calculates or determines power-based parameters from the complex voltages and currents, such as the average power reflected towards the Y megahertz RF generator 454 at the fundamental operating frequency of the Y megahertz RF generator 454, the average power reflected towards the Y megahertz RF generator 454 at the fundamental frequencies of the X kilohertz RF generator 452 and the Y megahertz RF generator 454 including sidebands, the average power reflected towards the Y megahertz RF generator 454 at the fundamental frequencies of the X kilohertz RF generator 452 and the Y megahertz RF generator 454 without sidebands, the average power reflection coefficient, and the average power reflection coefficient and the average voltage reflection coefficient. The sidebands include, for example, harmonic frequencies related to, e.g., generated therefrom, based thereon, etc., the fundamental frequencies of the X kilohertz RF generator 452 and the Y megahertz RF generator 454. As an example, the processor 454 accesses a look-up table stored in the memory device 468 to determine the power-based parameters from complex voltages and complex currents measured by sensors coupled to the processor 454 .

[0155] Continuing with the embodiment, processor 464 determines whether the value of one or more of the determined power-based parameters received from the sensor are within a preset range from the power-based parameter values ​​stored in memory device 468. Upon determining that the value of one or more of the determined power-based parameters are not within the preset range from the power-based parameter values ​​stored in memory device 468, processor 464 alters the operating frequency of Y megahertz RF generator 454. Processor 464 continues to alter the operating frequency of Y megahertz RF generator 454 until the value of one or more of the determined power-based parameters received from the sensor are within the preset range from the power-based parameter values ​​stored in memory device 468. Processor 464 determines an average of the values ​​of the operating frequency of Y megahertz RF generator 454 at which the value of one or more of the determined power-based parameters received from the sensor are within the preset range from the power-based parameter values ​​stored in memory device 468 to determine a baseline frequency. As another example, instead of the average of the values ​​of the operating frequency of Y megahertz RF generator 454, the median of the values ​​of the operating frequency of Y megahertz RF generator 454 is determined by processor 468 to be the baseline frequency. Determining the average frequency based on a power-based parameter allows for variation in baseline frequency from one plasma tool to another, e.g., from one plasma tool's RF generator to another plasma tool's RF generator. The RF generator of a plasma tool operates at a different baseline frequency than the RF generator of another plasma tool. Applying the values ​​yMHzfreqvariation1, thigh1, tlow1, and φrelative1 to the different baseline frequencies of the plasma tool's RF generators allows for tool-to-tool variation in baseline frequency.

[0156] 5A is a diagram of one embodiment of a system 500, such as a plasma tool, illustrating the generation of voltage reflection coefficient values ​​Γ11, Γ21 through Γn1 by processor 126 during recipe development. System 500 is similar to system 100 (FIG. 1A) or system 400 (FIG. 4A), except that in system 500, values ​​Γ11, Γ21 through Γn1 are determined for each bin or segment or portion of the duration of a cycle of RF signal 130 output by X kilohertz RF generator 112. The system 500 has the same structure and the same functionality as the system 100 or the system 400, except that in the system 500, a value Γ11 is generated from values ​​of complex voltages and complex currents measured by the sensor 131 and the high speed data acquisition device 123 during a first bin of a cycle of the RF signal 130, a value Γ12 is generated from values ​​of complex voltages and complex currents measured by the sensor 131 and the high speed data acquisition device 123 during a second bin of a cycle of the RF signal 130, and so on, until a value Γn1 is generated from values ​​of complex voltages and complex currents measured by the sensor 131 and the high speed data acquisition device 123 during the nth bin of a cycle of the RF signal 130. The second bin is contiguous with the first bin, the third bin is contiguous with the second bin, and so on, until the nth bin is contiguous with the (n-1)th bin of a cycle of the RF signal 130. As an example, a time segment of the RF signal 142 (FIG. 1B-2) from 0 to 0.1 microseconds is an example of a first bin, a time segment of the RF signal 142 from 0.1 microseconds to 0.2 microseconds is an example of a second bin, a time segment of the RF signal 142 from 0.2 microseconds to 0.3 microseconds is an example of a third bin, and a time segment of the RF signal 142 from 0.(n-1) microseconds to 0.n microseconds is an example of an nth bin. The system 500 operates similarly to the systems 100 or 400, except that in the system 500, a dummy substrate 402 is used and values ​​Γ11, Γ21 to Γn1 are generated for the bins of a cycle of the RF signal 130. In the system 500, there is no processing of the substrate S.

[0157] FIG. 5B is an illustration of one embodiment of a table 510 generated by the processor 126 (FIG. 5A) for recipe development. The table 510 is the same as the table 130 of FIG. 1B-1 or FIG. 4B, except that in the table, each of the values ​​Γ11, Γ21 through Γn1 corresponds to a different bin of a cycle of the RF signal 130 (FIG. 5A). For example, the value Γ11 is determined by the processor 126 from values ​​of complex voltage and complex current measured during bin 1, e.g., the first bin, of a cycle of the RF signal 130, and the value Γ21 is determined by the processor 126 based on values ​​of complex voltage and complex current measured during bin 2, e.g., the second bin, of a cycle of the RF signal 130. The table 510 is stored by the processor 126 in the memory device 128 (FIG. 5A).

[0158] In one embodiment, each value of the voltage reflection coefficient Γ11, Γ21 through Γn1 is an average calculated by the processor 126 for a corresponding bin over multiple cycles of the RF signal 130. For example, the value Γ11 is an average of multiple values ​​of the voltage reflection coefficient, each of the multiple values ​​being determined by the processor 126 for bin 1 of each corresponding cycle of the RF signal 130. By way of example, the value Γ11 is an average of a first value and a second value. The first value is calculated by the processor 126 based on values ​​of complex voltage and complex current measured by the sensor 131 and the high-speed data acquisition device 123 (FIG. 5A) during bin 1 of the first cycle of the RF signal 130. Similarly, the second value is calculated by the processor 126 based on values ​​of complex voltage and complex current measured by the sensor 131 and the high-speed data acquisition device 123 during bin 1 of the second cycle of the RF signal 130. The second cycle is consecutive to the first cycle. Each of the first and second cycles is divided by the processor 126 into n bins numbered from 1 to n.

[0159] 5C is a diagram of one embodiment of RF path model 150 illustrating the generation of load impedance values ​​ZL11-ZLn1 for each bin of a cycle of RF signal 130 (FIG. 5A) from voltage reflection coefficient values ​​Γ11, Γ21-Γn1 during recipe development. For example, the load impedance value ZL11 for bin 1 is determined by processor 126 from value Γ11, the load impedance value ZL21 for bin 2 is determined by processor 126 from value Γ21, and so on, until the load impedance value ZLn1 for bin n is determined by processor 126 from value Γn1. Note that the load impedance values ​​ZL11-ZLn1 for each bin of a cycle of RF signal 130 are determined from the voltage reflection coefficient values ​​Γ11, Γ21-Γn1 in the same manner as described above with respect to FIG. 1C.

[0160] In one embodiment, instead of voltage reflection coefficient values ​​Γ11, Γ21 through Γn1, complex voltage and complex current measurements are applied by the processor 126 to determine load impedance values ​​ZL11 through ZLn1 for each bin of a cycle of the RF signal 130.

[0161] 5D is a diagram of one embodiment of an RF path model 150 illustrating the determination of values ​​of operational parameters at which the value of the power reflection coefficient for the bins of the cycle 130 of the RF signal 130 (FIG. 5A) is minimized during recipe development. There are no predefined waveforms or initial values ​​of the operational parameters provided to the processor 126 (FIG. 5A). For example, the type of periodic waveform of the RF signal 132 output from the Y MHz RF generator 114 is not provided to the RF path model 150.

[0162] The value of the operating parameter for each bin is determined by the processor 126 such that the value of the power reflection coefficient at the input of the RF path model 150 is minimized when the value of the load impedance is applied to the output of the RF path model 150. For example, the RF path model 150 is initialized to have a capacitance Cknown1 and a radio frequency RFknown. The capacitance Cknown1 is the capacitance of a capacitor of a circuit component of a branch of the impedance matching network 110 between the input I2 and the output of the impedance matching network 110, and the radio frequency RFknown is a value at the Y megahertz RF generator 114 operating during recipe development. The value of the load impedance ZL11 at the output of the RF path model 150 is back-propagated through the circuit elements of the RF path model 150 to determine a value ZL1a of the load impedance at the input of the RF path model 150. The processor 126 determines the value ZL1a by calculating a directional sum of the value ZL11 and the value of the impedance of the circuit elements of the RF path model 150.

[0163] Continuing with the example, the processor 126 derives from the value of the load impedance ZL1a at the input of the RF path model 150 the value of the power reflection coefficient Γ1a at the input of the RF path model 150 2 For example, the processor 126 accesses a correspondence, such as a mapping or link, between the value of the voltage reflection coefficient Γ1a and the value of the load impedance ZL1a from the memory device 126 to identify or determine the value of the voltage reflection coefficient Γ1a.

[0164] Continuing with the example, the processor 126 calculates the square of the value of the voltage reflection coefficient Γ at the input of the RF path model 150 and the value of the power reflection coefficient Γ at the input of the RF path model 150. 2 The processor 126 determines values ​​such as yMHzavfreq1A and yMHzfreqvariation1A, as well as the value of the power reflection coefficient Γ 2is calculated, where yMHzavfreq1A is the average operating frequency at which the Y megahertz RF generator 454 (FIG. 5E) operates during bin 1 of a cycle of the RF signal output from the X KHz RF generator 452 (FIG. 5E), yMHzfreqvariation1A is the variation of the average frequency yMHzavfreq1A during bin 1, and C1A is the total capacitance applied to the branches between the input I21 and the output of the impedance matching network 456 (FIG. 5E).

[0165] To illustrate the determination of values ​​such as yMHzavfreq1A, yMHzfreqvariation1A, and C1A of the operational parameters, processor 126 provides the values ​​of the operational parameters to RF path model 150. By providing the values ​​to RF path model 150, processor 126 makes the values ​​available to RF path model 150. Once RF path model 150 receives the values ​​of the operational parameters from processor 126 and is characterized by the values, processor 126 back-propagates the value of the load impedance ZL11 at the output of RF path model 150 to determine the value of the voltage reflection coefficient Γ1a at the input of RF path model 150. Processor 126 accesses the values ​​of the operational parameters from memory device 126 and makes the values ​​available, e.g., provides the value to RF path model 150 where the value of the voltage reflection coefficient Γ1a is determined.

[0166] Continuing with the example, the value of the load impedance at the output of RF path model 150, ZL11, is back-propagated by processor 126 through the circuit elements of RF path model 150 to determine a value of the load impedance at the input of RF path model 150, ZL1x. Processor 126 calculates a directional sum of value ZL11 and the values ​​of the impedances of the circuit elements of the RF path model to determine value ZL1x.

[0167] Continuing with the example, the processor 126 may derive from the value of the load impedance ZL1x at the input of the RF path model 150 the value of the power reflection coefficient Γ11 at the input of the RF path model 150 2For example, the processor 126 accesses a correspondence, such as a mapping or link, between the value of the voltage reflection coefficient Γ11 and the value of the load impedance ZL1x from the memory device 126 to identify or determine the value of the voltage reflection coefficient Γ11.

[0168] Continuing with the example, the processor 126 calculates the square of the value of the voltage reflection coefficient Γ at the input of the RF path model 150 and the value of the power reflection coefficient Γ at the input of the RF path model 150. 2 The processor 126 determines the value of the power reflection coefficient Γ 2 Determine values ​​for the operating parameters, such as yMHzavfreqn1, yMHzfreqvariationn1, and Cn1, for which yMHzavfreqn1 is calculated, where yMHzavfreqn1 is the average operating frequency at which the Y megahertz RF generator 454 (FIG. 5E) operates, yMHzfreqvariationn1 is the variation of the average frequency, and Cn1 is the total capacitance applied to the branches between the input I21 and the output of the impedance matching network 456 (FIG. 5E).

[0169] To illustrate the determination of values ​​for operational parameters such as yMHzavfreqn1, yMHzfreqvariationn1, and Cn1, processor 126 provides values ​​of the operational parameters to RF path model 150. By providing the values ​​to RF path model 150, processor 126 makes the values ​​available to RF path model 150. Once RF path model 150 receives the values ​​of the operational parameters from processor 126 and is characterized by the values, processor 126 back-propagates the value of the load impedance ZL11 at the output of RF path model 150 to determine the value of the voltage reflection coefficient Γ11 at the input of RF path model 150. Processor 126 accesses the values ​​of the operational parameters from memory device 126 and makes the values ​​available, e.g., provides the value to RF path model 150 where the value of the voltage reflection coefficient Γ11 is determined.

[0170] The processor 126 calculates the value of the power reflection coefficient Γ at the input of the RF path model 150. 2is the value of the power reflection coefficient Γ1a at the input of the RF path model 150 for bin 1. 2 In response to the determination, processor 126 determines to apply frequency parameters yMHzavfreqn1 and yMHzfreqvariationn1 instead of applying frequency parameters yMHzavfreq1A and yMHzfreqvariation1A to Y megahertz RF generator 454 (FIG. 5E), and determines to apply match network parameter Cn1 instead of applying match network parameter C1A to impedance matching network 456 (FIG. 5E).

[0171] In a similar manner, values ​​of operational parameters yMHzfreqvariationn2 and Cn1 are determined by processor 126 for bin 2 of RF signal 484 (FIG. 5E) generated by X kilohertz RF generator 452 (FIG. 5E), except that for bin 2, processor 126 determines a value yMHzavfreqn2 to be the directional sum of the value yMHzavfreqn1 calculated for bin 1 and the value yMHzfreqvariationn1. Also for bin 2, processor 126 initiates RF path model 150 to become the value yMHzavfreqn2. Similarly, values ​​yMHzavfreqnn, yMHzfreqvariationnn, and Cn1 are also determined by processor 126 for bin n of RF signal 484 generated by X kilohertz RF generator 452 (FIG. 5E). The processor 126 stores in the memory device 128 values ​​of the operating parameters of bins 1 to n, such as yMHzavfreqn1, yMHzfreqvariationn1, yMHzfreqvariationn2, yMHzfreqvariationnn, and Cn1.

[0172] In one embodiment, the value Cn1 of the match network parameter of RF path model 150 is determined after determining the values ​​of the frequency parameters of RF path model 150 for bins 1 through n. For example, processor 126 determines values ​​yMHzavfreqn1, yMHzfreqvariationn1, and yMHzfreqvariationn2, etc., up to a value yMHzfreqvariationnn, in a manner similar to that described above with reference to FIG. 5D, except that processor 126 does not determine values ​​Cn1 and C1A. After values ​​yMHzavfreqn1, yMHzfreqvariationn1, yMHzfreqvariationn2, etc., up to a value yMHzfreqvariationnn are determined, processor 126 determines the value Cn1 that minimizes the average of the values ​​of the power reflection coefficient at the input of RF path model 150. Illustratively, processor 126 determines the value Γ1a of the power reflection coefficient at the input of RF path model 150. 2 , Γ2a 2 , and Γ 2 to determine that the RF path model 150 has a value C1A to achieve the first average, and a value Γ11 of the power reflection coefficient at the input of the RF path model 150. 2 , Γ21 2 , and Γ 2 The second average value of Γ1avmin 2 The RF path model 150 calculates the second average value Γ1avmin 2 The processor 126 determines the second average value Γavmin 2 The second average value Γavmin is further determined to be lower than the first average. 2 If the processor 126 determines that the second average value Γavmin is lower than the first average, 2is calculated to be applied to the impedance matching network 456 of Figure 5E. The processor 126 stores the values ​​of the operational parameters yMHzavfreqn1, yMHzfreqvariationn1, yMHzfreqvariationn2, and yMHzfreqvariationnn, and the value Cn1 in the memory device 128 (Figure 5A).

[0173] 5E is a diagram of one embodiment of a system 450 illustrating the use of values ​​yMHzavfreqn1, yMHzfreqvariationn1, yMHzfreqvariationn2, and yMHzfreqvariationnn, and value Cn1 during processing of a substrate SU. In the system 450, a computer 463 receives values ​​of the operational parameters from the computer 118 of FIG. 5A. For example, the values ​​of the operational parameters yMHzavfreqn1, yMHzfreqvariationn1, yMHzfreqvariationn2, and yMHzfreqvariationnn, and value Cn1 are received by a processor 464 of the computer 463 from the processor 126 of the computer 118 via a computer network. The processor 464 stores values ​​such as the values ​​of the operational parameters yMHzavfreqn1, yMHzfreqvariationn1, yMHzfreqvariationn2, and yMHzfreqvariationnn, and value Cn1 in a memory device 468 of the computer 463.

[0174] While processing the substrate SU in the plasma chamber 458, the processor 464, in response to determining that the phase of the RF signal 484 is such that the RF signal 484 crosses a zero crossing to achieve a positive voltage or power value from a negative voltage or power value, transmits frequency parameter values ​​yMHzavfreqn1 and yMHzfreqvariationn1 for bin 1 of the RF signal 484 to the RF power supply 476. Upon receiving the values ​​yMHzavfreqn1 and yMHzfreqvariationn1, the RF power supply 476 generates an RF signal 550 having a frequency that is a directional sum of the values ​​yMHzavfreqn1 and yMHzfreqvariationn1 during bin 1 of the first cycle of the RF signal 484.

[0175] The processor 464 determines that the bin 2 period of the first cycle of the RF signal 484 is about to begin. Upon so determining, the processor 464 accesses the value yMHzfreqvariationn2 from the memory device 468 and provides the value yMHzfreqvariationn2 to the RF power supply 476. During the bin 2 period of the first cycle of the RF signal 484, the RF power supply 476 outputs an RF signal 550 having an average value of the frequency parameter and a directional sum of the value yMHzfreqvariationn2. The average value of the frequency parameter for bin 2 of the RF signal 484 is the directional sum of the values ​​yMHzavfreqn1 and yMHzfreqvariationn1. In a similar manner, the processor 464 provides the remaining values ​​of the frequency parameter for the remaining bins 3 through n to the RF power supply 476 during the remaining bins of the first cycle of the RF signal 484 generated by the X kilohertz RF generator 452 to output the RF signal 550.

[0176] During the next cycle of RF signal 484, the frequency parameter for bins 1 through n of each next cycle is the same as the frequency parameter for bins 1 through n of the first cycle. For example, RF signal 550 has values ​​yMHzavfreqn1 and yMHzfreqvariationn1 during bin 1 of the second cycle of RF signal 484. RF signal 550 has an average value of the frequency parameters and a directional sum of value yMHzfreqvariationn2 during bin 2 of the second cycle of RF signal 484. The average value of the frequency parameter for bin 2 of the second cycle of RF signal 484 is the directional sum of values ​​yMHzavfreqn1 and yMHzfreqvariationn1.

[0177] Additionally, the processor 464 controls the combined capacitance of the branch between the input I21 and the output of the impedance matching network 456 by controlling the motor system 472 in the same manner as described above with reference to Figure 4E. For example, the processor 464 controls one or more capacitors of the impedance matching network 456 via the driver system 470 and the motor system 472 to achieve a capacitance Cn1 of the branch between the input I21 and the output of the impedance matching network 456 during processing of the substrate SU.

[0178] The RF power supply 476 supplies an RF signal 550 to an input I21 of an impedance matching network 456 having a combined capacitance Cn1 via an output O21 of the Y megahertz RF generator 454 and an RF cable RFC21. Upon receiving the RF signals 484 and 550, the impedance matching network 456 matches the impedance of a load coupled to the output of the impedance matching network 456 with the impedance of a source coupled to the inputs I11 and I21 of the impedance matching network 456, and outputs a modified RF signal 552 at the output of the impedance matching network 456. When the lower electrode of the plasma chamber 458 receives the modified RF signal 552 and one or more process gases are supplied to a gap between the upper electrode 460 and the lower electrode of the plasma excitation electrode 462 of the plasma chamber 458, a plasma is struck or sustained in the plasma chamber 458 to process a substrate SU in the plasma chamber 458.

[0179] In one embodiment, a directional sum as used herein is a vector sum.

[0180] In one embodiment, the method described herein with reference to FIG. 5E is used to process multiple substrates in plasma chamber 458 or in a different system similar in structure and function to system 450.

[0181] Note that in one embodiment, instead of applying an RF signal to the lower electrode of plasma excitation electrode 462 and coupling the upper electrode 460 to ground potential, an RF signal is applied to the upper electrode 460 and the lower electrode of plasma excitation electrode 462 is coupled to ground potential.

[0182] The embodiments described herein may be practiced with a variety of computer system configurations including hand-held hardware units, microprocessor systems, microprocessor-based or programmable consumer electronics, minicomputers, mainframe computers, etc. The embodiments described herein may also be practiced in distributed computing environments where tasks are performed by remote processing hardware units that are linked through a computer network.

[0183] In some embodiments, the controller is part of a system that may be part of the examples described above. Such systems include semiconductor processing equipment including one or more processing tools, one or more chambers, one or more processing platforms, and / or specific processing components (wafer pedestals, gas flow systems, etc.). These systems are integrated with electronics for controlling pre-, during, and post-processing operations of semiconductor wafers or substrates. Such electronics may be referred to as a "controller" and may control various components or subparts of the system. The controller is programmed to control any of the processes disclosed herein depending on the processing requirements and / or type of system. Such processes include delivery of process gases, temperature settings (e.g., heating and / or cooling), pressure settings, vacuum settings, power settings, RF generator settings, RF matching circuit settings, frequency settings, flow rate settings, fluid delivery settings, position and motion settings, wafer loading and unloading from the tool, and wafer loading and unloading from other transport tools and / or load locks connected or interlocked with the system.

[0184] Broadly, in various embodiments, a controller is defined as an electronic device having various integrated circuits, logic, memory, and / or software that receive instructions, issue instructions, control operations, enable cleaning operations, enable endpoint measurements, etc. Integrated circuits include chips in the form of firmware that store program instructions, digital signal processors (DSPs), chips defined as ASICs, PLDs, and / or one or more microprocessors, i.e., microcontrollers that execute program instructions (e.g., software). Program instructions are instructions communicated to the controller in the form of various individual settings (or program files) that define operational parameters for carrying out a process on or for a semiconductor wafer. The operational parameters, in some embodiments, are part of a recipe defined by a process engineer to accomplish one or more layers, materials, metals, oxides, silicon, silicon dioxide, surfaces, circuits, and / or one or more processing steps in the fabrication of a wafer die.

[0185] The controller, in some embodiments, is part of a computer that is integrated or coupled with the system or otherwise networked to the system, or is coupled to such a computer, or a combination thereof. For example, the controller may be in the "cloud" or all or part of a fab host computer system. This allows for remote access of wafer processing. The computer may allow remote access to the system to monitor the current progress of a fabrication operation, review the history of past fabrication operations, review trends or performance criteria from multiple fabrication operations, change parameters of a current process, set up processing steps following the current process, or initiate a new process.

[0186] In some examples, a remote computer (e.g., a server) provides a process recipe to the system over a network. Such a network includes a local network or the Internet. The remote computer includes a user interface that allows entry or programming of parameters and / or settings, which are then communicated from the remote computer to the system. In some examples, the controller receives instructions in the form of settings for processing the wafer. It should be understood that the settings are specific to the type of process to be performed on the wafer and the type of tool the controller interfaces with or controls. Thus, as described above, the controller is distributed, for example, by including one or more individual controllers that are networked together and cooperate toward a common purpose (such as performing processes described herein). Examples of distributed controllers for such purposes include one or more integrated circuits on the chamber that communicate with one or more integrated circuits that are remotely located (e.g., at the platform level or as part of a remote computer) and coupled to control the process in the chamber.

[0187] In various embodiments, the systems include, but are not limited to, plasma etch chambers, deposition chambers, spin rinse chambers, metal plating chambers, cleaning chambers, bevel edge etch chambers, physical vapor deposition (PVD) chambers, chemical vapor deposition (CVD) chambers, atomic layer deposition (ALD) chambers, atomic layer etch (ALE) chambers, ion implantation chambers, tracking chambers, and any other semiconductor processing systems associated with or used in the fabrication and / or manufacturing of semiconductor wafers.

[0188] It is further noted that while the above operations are described with reference to a parallel plate plasma chamber, e.g., a capacitively coupled plasma chamber, in some embodiments the above operations apply to other types of plasma chambers, e.g., plasma chambers including inductively coupled plasma (ICP) reactors, transformer coupled plasma (TCP) reactors, conductor tools, dielectric tools, electron cyclotron resonance (ECR) reactors, etc. For example, an X MHz RF generator, a Y MHz RF generator, and a Z MHz RF generator are coupled to an inductor in an ICP plasma chamber.

[0189] As described above, depending on the process operations being performed by the tool, the controller communicates with one or more other tool circuits or modules, other tool components, cluster tools, other tool interfaces, adjacent tools, nearby tools, tools located throughout the factory, a main computer, another controller, or tools used for material transport to and from tool locations and / or load ports within a semiconductor manufacturing factory.

[0190] With the above embodiments in mind, it should be understood that some of the embodiments employ various computer-implemented operations involving data stored in computer systems. These computer-implemented operations are operations that manipulate physical quantities.

[0191] Some of the embodiments also relate to hardware units or devices for performing these operations. The device is specially constructed for a special purpose computer. When defined as a special purpose computer, the computer is operable for its dedicated purpose while performing other processes, program execution or routines that are not part of its dedicated purpose.

[0192] In some embodiments, the operations described above are performed by a computer selectively activated or configured by one or more computer programs stored in a computer memory or obtained over a computer network. If data is obtained over a computer network, the data may be processed by other computers on the computer network (e.g., a cloud of computing resources).

[0193] One or more embodiments described above can also be fabricated as computer readable code on a non-transitory computer readable medium. A non-transitory computer readable medium is any data storage hardware unit (e.g., memory device, etc.) that stores data, which is then read by a computer system. Examples of non-transitory computer readable media include hard drives, network attached storage (NAS), ROM, RAM, compact disk ROM (CD-ROM), CD recordable (CD-R), CD rewriteable (CD-RW), magnetic tape, and other optical and non-optical data storage hardware units. In some embodiments, a non-transitory computer readable medium includes a computer readable tangible medium distributed over a network-coupled computer system such that the computer readable code is stored and executed in a distributed manner.

[0194] Although some of the method operations described above have been described in a particular order, it should be understood that in various embodiments, other housekeeping operations are performed between each method operation, or each method operation is coordinated to occur at slightly different times, or distributed across a system that allows each method operation to occur at various intervals, or performed in an order different from that described above.

[0195] It should further be noted that in one embodiment, one or more features of any of the above-described embodiments may be combined with one or more features of any of the other embodiments without departing from the scope described in the various embodiments described in this disclosure.

[0196] Although the foregoing embodiments have been described in some detail for clarity of understanding, it will be apparent that certain changes and modifications can be practiced within the scope of the appended claims. Thus, the present embodiments should be considered as illustrative and not restrictive, and the embodiments should not be limited to the details set forth herein, but may be modified within the scope of the appended claims and their equivalents.

Claims

1. A method of adjusting, comprising: accessing a plurality of reflection parameter values ​​associated with the second radio frequency generator for a first set of one or more cycles of operation of the first radio frequency generator; calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path, the radio frequency path being between the second radio frequency generator and an electrode of a plasma chamber; receiving a plurality of frequency modulation parameters of a radio frequency signal generated by the second radio frequency generator; determining values ​​of the plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model, the values ​​of the plurality of frequency modulation parameters being determined so as to minimize a reflection coefficient parameter at an input of the computer-based model; controlling the second radio frequency generator according to the values ​​of the plurality of frequency modulation parameters during a second set of one or more cycles of operation of the first radio frequency generator; A method comprising:

2. 2. The method of claim 1 , The method, wherein the accessing of the plurality of reflection parameter values, the calculating of the plurality of load impedance parameter values, the determining of the value of the frequency modulation parameter, and the controlling of the second radio frequency generator are performed during processing of a substrate in the plasma chamber.

3. 2. The method of claim 1 , The method of claim 1, wherein the plurality of reflection parameter values ​​comprises a plurality of voltage reflection coefficient values, and the reflection coefficient parameter is an average power reflection coefficient.

4. 2. The method of claim 1 , A method, wherein the second set of one or more cycles of operations follows the first set of one or more cycles of operations.

5. 2. The method of claim 1 , wherein the first set includes two or more cycles of operation of the first radio frequency generator, and each of the plurality of reflection parameter values ​​is an average of a plurality of reflection coefficient values ​​calculated over the two or more cycles of the first set.

6. 2. The method of claim 1 , repeating the accessing of the plurality of reflection parameter values, the calculating of the plurality of load impedance parameter values, and the determining of the values ​​of the plurality of frequency modulation parameters for a third set of one or more cycles of operation of the first radio frequency generator; repeating said controlling said second radio frequency generator for a fourth set of one or more cycles of operation of said first radio frequency generator. Further comprising: the third set follows the second set, and the fourth set follows the third set; method.

7. 2. The method of claim 1 , A method according to claim 1, wherein the plurality of reflectance parameter values ​​comprises a plurality of values ​​of a complex voltage and a complex current.

8. 2. The method of claim 1 , The method of claim 1, wherein the computer-based model includes a plurality of circuit elements, two adjacent circuit elements of the plurality of circuit elements being coupled to each other via a connection, and the computer-based model has an impedance substantially the same as an impedance of the radio frequency path.

9. 2. The method of claim 1 , The method, wherein the computer-based model includes a plurality of circuit elements, two adjacent circuit elements of the plurality of circuit elements being coupled to each other via a connection, the plurality of circuit elements representing a plurality of circuit components of the radio frequency path and being connected in the same manner as the plurality of circuit components of the radio frequency path.

10. 2. The method of claim 1 , The radio frequency path includes: a radio frequency cable coupling the second radio frequency generator to a branch of a match network; The branch portion; a radio frequency transmission line coupling the match network and the plasma chamber; an electrode of the plasma chamber; A method comprising:

11. 2. The method of claim 1 , The method further includes determining a capacitance for a match network by applying the plurality of load impedance parameter values ​​to the computer-based model, the match network being coupled between the second radio frequency generator and the plasma chamber and between the first radio frequency generator and the plasma chamber.

12. 2. The method of claim 1 , The applying of the plurality of load impedance parameter values ​​to the computer-based model comprises: back-propagating the plurality of load impedance parameter values ​​from an output of the computer-based model through the computer-based model to facilitate calculation of a first plurality of reflection parameter input values ​​at the input of the computer-based model, the first plurality of reflection parameter input values ​​being determined when quantities of the plurality of frequency modulation parameters are available to the computer-based model; calculating a first average of the first plurality of reflectance parameter input values; back-propagating the plurality of load impedance parameter values ​​from the output of the computer-based model through the computer-based model to facilitate calculation of a second plurality of reflection parameter input values ​​at the input of the computer-based model, the second plurality of reflection parameter input values ​​being determined when the values ​​of the plurality of frequency modulation parameters are available to the computer-based model; calculating a second average of the second plurality of reflectance parameter input values; determining whether the second average is lower than the first average; A method comprising:

13. 2. The method of claim 1 , The method of claim 1, wherein the received frequency modulation parameters represent a periodic function related to one of the one or more cycles of operation of the first radio frequency generator.

14. 14. The method of claim 13, The method, wherein the periodic function is one of a trapezoidal function, a sine function, a pulse function, and a sawtooth function.

15. A method of adjusting, comprising: Recipe development in progress accessing a plurality of reflection parameter values ​​associated with the second radio frequency generator for a set of one or more cycles of operation of the first radio frequency generator; calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path, the radio frequency path being between the second radio frequency generator and an electrode of a plasma chamber; receiving a plurality of frequency modulation parameters of a radio frequency signal generated by the second radio frequency generator; determining values ​​of the plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model, the values ​​of the plurality of frequency modulation parameters being determined so as to minimize one or more reflection coefficient parameters at an input of the computer-based model; During processing of the substrate in another plasma chamber, controlling a third radio frequency generator according to the values ​​of the plurality of frequency modulation parameters determined during the recipe development, wherein the controlling of the third radio frequency generator is performed during a set of one or more cycles of operation of a fourth radio frequency generator. A method comprising:

16. 16. The method of claim 15, The method, wherein the one or more reflection coefficient parameters minimized during recipe development comprises an average power reflection coefficient, and the plurality of reflection parameter values ​​comprises a plurality of values ​​of a voltage reflection coefficient.

17. 16. The method of claim 15, wherein the third radio frequency generator is designated to have the same operating frequency as the second radio frequency generator and the fourth radio frequency generator is designated to have the same operating frequency as the first radio frequency generator.

18. 16. The method of claim 15, the set of one or more cycles of operation of the first radio frequency generator comprises two or more cycles of operation of the first radio frequency generator, and each of the plurality of reflection parameter values ​​is an average of a plurality of reflection coefficient values ​​determined over the two or more cycles of operation of the first radio frequency generator.

19. 16. The method of claim 15, A method according to claim 1, wherein the plurality of reflectance parameter values ​​comprises a plurality of values ​​of a complex voltage and a complex current.

20. 16. The method of claim 15, A method according to claim 1, wherein the plurality of frequency modulation parameters define a trapezoidal function, or a sinusoidal function, or a rectangular function, or a triangular function.

21. 16. The method of claim 15, determining a capacitance for a match network by applying the plurality of load impedance parameter values ​​to the computer-based model, the match network being coupled between the third radio frequency generator and the other plasma chamber and between the fourth radio frequency generator and the other plasma chamber.

22. 16. The method of claim 15, The applying of the plurality of load impedance parameter values ​​to the computer-based model comprises: back-propagating the plurality of load impedance parameter values ​​from an output of the computer-based model through the computer-based model to calculate a first plurality of reflection parameter input values ​​at the input of the computer-based model, the first plurality of reflection parameter input values ​​being determined when quantities of the plurality of frequency modulation parameters are available to the computer-based model; calculating a first average of the first plurality of reflectance parameter input values; back-propagating the plurality of load impedance parameter values ​​from the output of the computer-based model through the computer-based model to calculate a second plurality of reflection parameter input values ​​at the input of the computer-based model, the second plurality of reflection parameter input values ​​being determined when the values ​​of the plurality of frequency modulation parameters are available to the computer-based model; calculating a second average of the second plurality of parameter input values; determining that the second average is lower than the first average; A method comprising:

23. A method of adjusting, comprising: Recipe development in progress accessing a plurality of reflection parameter values ​​associated with the second radio frequency generator for a set of one or more cycles of operation of the first radio frequency generator; calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path between the second radio frequency generator and an electrode of a plasma chamber; receiving a plurality of frequency modulation parameters of a radio frequency signal generated by the second radio frequency generator, the plurality of frequency modulation parameters including a frequency variation of the radio frequency signal; determining values ​​of the plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model, the values ​​of the plurality of frequency modulation parameters being determined so as to minimize one or more reflection coefficient parameters at an input of the computer-based model; During processing of the substrate in another plasma chamber, controlling a third radio frequency generator according to the values ​​of the plurality of frequency modulation parameters determined during the recipe development, the controlling of the third radio frequency generator including applying the values ​​of the plurality of frequency modulation parameters to a baseline frequency of operation of the third radio frequency generator; A method comprising:

24. 24. The method of claim 23, A method according to claim 1, wherein the plurality of frequency modulation parameters define a trapezoidal function, or a sinusoidal function, or a rectangular function, or a triangular function.

25. 24. The method of claim 23, The method of claim 1, wherein the one or more reflection coefficient parameters minimized during recipe development comprises an average power reflection coefficient, and the plurality of reflection parameter values ​​comprises a plurality of values ​​of a voltage reflection coefficient.

26. 24. The method of claim 23, The method, wherein the baseline frequency is used to minimize a power-based parameter associated with the third radio frequency generator.

27. A method of adjusting, comprising: Recipe development in progress accessing a plurality of reflectance parameter values ​​associated with a second radio frequency generator for a set of one or more cycles of operation of a first radio frequency generator, each of the plurality of reflectance parameter values ​​corresponding to a respective bin of the one or more cycles of operation of the first radio frequency generator; calculating a plurality of load impedance parameter values ​​from the plurality of reflection parameter values ​​by applying the plurality of reflection parameter values ​​to a computer-based model of at least a portion of a radio frequency path, the radio frequency path being between the second radio frequency generator and an electrode of a plasma chamber; determining values ​​of a plurality of frequency modulation parameters by applying the plurality of load impedance parameter values ​​to the computer-based model, the values ​​of the plurality of frequency modulation parameters being determined to minimize a plurality of values ​​of a reflection coefficient parameter at an input of the computer-based model for each of the bins; During processing of the substrate in another plasma chamber, controlling a third radio frequency generator according to the values ​​of the plurality of frequency modulation parameters determined during the recipe development, wherein the controlling of the third frequency generator is performed during a set of one or more cycles of operation of a fourth radio frequency generator. A method comprising:

28. 28. The method of claim 27, A method, wherein the reflection coefficient parameter comprises a power reflection coefficient parameter.

29. 28. The method of claim 27, the set of one or more cycles of operation of the first radio frequency generator comprises two or more cycles of operation of the first radio frequency generator, and each of the plurality of reflection parameter values ​​is an average of a plurality of reflection coefficient values ​​determined over the two or more cycles of operation of the first radio frequency generator.

30. 28. The method of claim 27, The method of claim 1, wherein the plurality of frequency modulation parameters include an average frequency of the radio frequency signal output by the third radio frequency generator and a frequency variation of the radio frequency signal output by the third radio frequency generator, the frequency variation being for each bin of the radio frequency signal generated by the fourth radio frequency generator.

31. 28. The method of claim 27, the plurality of load impedance parameter values ​​are applied to the computer-based model to determine capacitances for match networks, the match networks being coupled between the third radio frequency generator and the other plasma chamber and between the fourth radio frequency generator and the other plasma chamber.

32. 28. The method of claim 27, The applying of the plurality of load impedance parameter values ​​to the computer-based model comprises: back-propagating a first one of the plurality of load impedance parameter values ​​from an output of the computer-based model through the computer-based model to facilitate calculation of a first reflection parameter input value at the input of the computer-based model, the first reflection parameter input value being determined when amounts of the plurality of frequency modulation parameters are available to the computer-based model; back-propagating the first value of the plurality of load impedance parameter values ​​from the output of the computer-based model through the computer-based model and calculating a second reflection parameter input value at the input of the computer-based model, the second reflection parameter input value being determined when the values ​​of the plurality of frequency modulation parameters are available to the computer-based model; determining whether the second reflectance parameter input value is lower than the first reflectance parameter input value; A method comprising:

Citation Information

Patent Citations

  • Plasma processing apparatus

    JP2015090770A

  • Systems and methods for reducing power reflected towards higher frequency RF generator during period of lower RF generator and for using relationship to reduce reflected power

    JP2017188434A

  • Systems and methods for increasing efficiency of delivered power of a megahertz radio frequency generator in the presence of a kilohertz radio frequency generator

    WO2019018078A1