Automatic electrostatic chuck bias compensation during plasma processing

Abstract

Embodiments of the present disclosure relate to a system for pulsed direct-current (DC) biasing and clamping a substrate. In one embodiment, the system includes a plasma chamber having an electrostatic chuck (ESC) for supporting a substrate. An electrode is embedded in the ESC and is electrically coupled to a biasing and clamping network. The biasing and clamping network includes at least a shaped DC pulse voltage source and a clamping network. The clamping network includes a DC source and a diode, and a resistor. The shaped DC pulse voltage source and the clamping network are connected in parallel. The biasing and clamping network automatically maintains a substantially constant clamping voltage, which is a voltage drop across the electrode and the substrate when the substrate is biased with pulsed DC voltage, leading to improved clamping of the substrate.

Description

BACKGROUND

Field

Embodiments of the present disclosure generally relate to a system used in semiconductor manufacturing. More specifically, embodiments of the present disclosure relate to a system for biasing and clamping a substrate during plasma processing.

Description of the Related Art

Ion bombardment is often used as a source of activation energy for chemical and physical processes in plasma etch and plasma enhanced chemical vapor deposition (PECVD) processes for processing a semiconductor substrate. High energy ions accelerated by plasma sheath are also highly directional and can be used for etching high aspect ratio features. Conventionally, a substrate may be biased using radio frequency (RF) power from an RF source. The RF source supplies an RF voltage to a first electrode embedded in an electrostatic chuck (ESC) or cathode. The first electrode is capacitively coupled to the plasma of a processing chamber through a layer of ceramic, which is a part of the ESC. Non-linear, diode-like nature of the plasma sheath results in rectification of the applied RF field, such that a direct-current (DC) voltage drop, or self-bias, appears between the substrate and the plasma. This voltage drop determines the average energy of the ions accelerated towards the substrate.

The ESC secures the substrate disposed thereon by applying a fixed DC voltage to a second electrode embedded in the ESC to establish an electric field between the ESC and the substrate. The electric field induces opposite polarity charges to accumulate on the substrate and the second electrode, respectively. The electrostatic attractive force between the oppositely polarized charges pulls the substrate toward the ESC to secure the substrate. However, the electrostatic force can be affected by the RF bias power supplied to the first electrode in the ESC, leading to under or over clamping of the substrate. In addition, as large bias voltage becomes many kilovolts, the fluctuation of the self-bias voltage with respect to the fixed DC voltage can lead to an increased risk of arcing or sudden de-clamping and breaking of the substrate. This is particularly a problem with very high bias power (kilovolts (kV) range) which is used during pulsed voltage type of substrate biasing techniques.

Therefore, an improved system for biasing and clamping a substrate is needed.

SUMMARY

Embodiments of the disclosure may provide a plasma processing chamber, comprising a substrate support assembly, a waveform generator, a first power delivery line, a clamping network, a signal detection module, and a controller. The substrate support assembly comprises a substrate supporting surface, a first biasing electrode, and a first dielectric layer disposed between the first biasing electrode and the substrate supporting surface. The first power delivery line electrically couples the waveform generator to the first biasing electrode, wherein the first power delivery line comprises a blocking capacitor. The clamping network is coupled to the first power delivery line at a first point between the blocking capacitor and the biasing electrode, the clamping network comprising a direct-current (DC) voltage source coupled between the first point and ground, and a blocking resistor coupled between the first point and an output of the direct-current (DC) voltage source. The signal detection module is configured to receive a first electrical signal from a first signal trace that is coupled to the first power delivery line at a point disposed between the blocking capacitor and the biasing electrode. The controller is configured to communicate with the signal detection module and control a magnitude of a voltage supplied to the first power delivery line at the first point by the direct-current (DC) voltage source due to information received within the received electrical signal.

Embodiments of the disclosure may further provide a plasma processing chamber comprising a substrate support assembly, a waveform generator, a first power delivery line, a clamping network, and a signal detection module. The first power delivery line electrically couples the waveform generator to the first electrode, wherein the first power delivery line comprises a blocking capacitor. The clamping network is coupled to the first power delivery line at a first point between the blocking capacitor and the first electrode, the clamping network comprising a direct-current (DC) voltage source coupled between the first point and ground, and a blocking resistor coupled between the first point and the direct-current (DC) voltage source. The signal detection module is configured to receive a first electrical signal from a first signal trace that is coupled to the first power delivery line at a point disposed between the blocking capacitor and the first electrode.

Embodiments of the disclosure may further provide a method for plasma processing a substrate, comprising: generating a plasma within a processing region of a processing chamber, wherein the processing region comprises a substrate support that comprises a substrate supporting surface, a first biasing electrode, and a first dielectric layer disposed between the first biasing electrode and the substrate supporting surface; delivering, from a waveform generator, a plurality of pulsed-voltage waveforms to the first biasing electrode through a first power delivery line during a first time period, wherein the first power delivery line comprises a blocking capacitor that is disposed between the waveform generator and the biasing electrode; halting the delivery of the plurality of pulsed-voltage waveforms to the first biasing electrode during all of a second time period; applying, from a clamping network, a first clamping voltage to the first biasing electrode; detecting at least one characteristic of one or more of the delivered plurality of pulsed-voltage waveforms during the first time period by receiving an electrical signal from a signal trace that is coupled to the first power delivery line at a first point disposed between the blocking capacitor and the biasing electrode; detecting at least one characteristic of an electrical signal received from the signal trace during the second time period; and adjusting the first clamping voltage applied to the first biasing electrode based on the detected characteristic of the one or more of the delivered plurality of pulsed-voltage waveforms and the at least one characteristic of the electrical signal received from the signal trace during the first and second time periods.

Embodiments of the disclosure may further provide a method for plasma processing a substrate, comprising: generating a plasma within a processing region of a processing chamber, wherein the processing region comprises a substrate support that comprises a substrate supporting surface, a first biasing electrode, and a first dielectric layer disposed between the first biasing electrode and the substrate supporting surface; delivering, from a waveform generator, one or more waveforms to the first biasing electrode through a first power delivery line during a first time period; halting the delivery of the one or more waveforms to the first biasing electrode for a second time period; applying, from a clamping network, a first clamping voltage to the first biasing electrode; detecting at least one characteristic of the one or more waveforms during the first time period by receiving an electrical signal from a signal trace that is coupled to the first power delivery line at a first point disposed on the first power delivery line; detecting at least one characteristic of an electrical signal received from the signal trace during the second time period; and adjusting the first clamping voltage applied to the first biasing electrode based on the detected characteristic of the one or more waveforms received from the signal trace during the first time period; and the detected at least one characteristic of the electrical signal received from the signal trace during the second time period.

BRIEF DESCRIPTION OF THE DRAWINGS

So that the manner in which the above recited features of the present disclosure can be understood in detail, a more particular description of the disclosure, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only exemplary embodiments and are therefore not to be considered limiting of its scope, and may admit to other equally effective embodiments.

FIG. 1A is a schematic cross-sectional view of a processing chamber configured to practice methods described herein, according to one embodiment.

FIG. 1B is a close-up schematic cross-sectional view of a portion of the processing chamber illustrated in FIG. 1A, according to one embodiment.

FIG. 1C is a functionally equivalent circuit diagram of a coulombic electrostatic chuck (ESC) that can be used in the process chamber illustrated in FIG. 1A, according to one embodiment.

FIG. 1D is a functionally equivalent circuit diagram of a Johnsen-Rahbek electrostatic chuck (ESC) that can be used in the process chamber illustrated in FIG. 1A, according to one embodiment.

FIG. 1E is a schematic diagram illustrating an example of a processing chamber that includes a feedback loop that is illustrated in FIG. 1A, according to one embodiment.

FIG. 2A is a functionally equivalent circuit diagram of a system that can be used to generate negative pulses in a process chamber, according to one embodiment.

FIG. 2B is a functionally equivalent circuit diagram of a system that can be used to generate positive pulses in a process chamber, according to one embodiment.

FIG. 3A illustrates an example of pulsed voltage (PV) waveforms established at different portions of the functionally equivalent circuit diagram illustrated in FIG. 3B, according to one embodiment.

FIG. 3B is a circuit diagram illustrating a system that can be used to perform one or more methods described herein, according to one embodiment.

FIG. 4A illustrates an example of negative pulsed voltage (PV) waveforms established at the biasing electrode and substrate, according to one embodiment.

FIGS. 4B-4D illustrate examples of a series of pulse voltage (PV) waveform bursts, according to one or more embodiments.

FIG. 5A is a functionally equivalent circuit diagram of a system that can be used to deliver an RF waveform to an electrode within the process chamber, according to one embodiment.

FIG. 5B illustrates an example of RF waveforms established at different portions of the functionally equivalent circuit diagram illustrated in FIG. 5A, according to one embodiment.

FIGS. 6A-6B are process flow diagrams illustrating methods of biasing and clamping a substrate during plasma processing, according to one or more embodiments.

To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements and features of one embodiment may be beneficially incorporated in other embodiments without further recitation.

DETAILED DESCRIPTION

Embodiments of the disclosure provided herein include an apparatus and methods for plasma processing of a substrate in a processing chamber. Aspects of one or more of the embodiments disclosed herein include a system and method of reliably biasing and clamping a substrate during processing to improve the plasma processing results. Embodiments of the disclosure may include an apparatus and method for providing a pulsed-voltage (PV) waveform delivered from one or more pulsed-voltage (PV) generators to one or more electrodes within the processing chamber, while biasing and clamping a substrate during a plasma process. In some embodiments, a radio frequency (RF) generated RF waveform is provided from an RF generator to one or more electrodes within a processing chamber to establish and maintain a plasma within the processing chamber, while PV waveform(s) delivered from a PV generator are configured to establish a nearly constant sheath voltage across the surface of a substrate. The established nearly constant sheath voltage across the surface of a substrate can create a desirable ion energy distribution function (IEDF) at the surface of the substrate during one or more plasma processing steps performed within the processing chamber. In some embodiments, the PV waveform is established by a PV generator that is electrically coupled to a biasing electrode disposed within a substrate support assembly disposed within a plasma processing chamber.

During some of the plasma processes, ions are purposely accelerated towards the substrate by the voltage drop formed in an electron-repelling sheath that forms over a substrate placed on top of a substrate-support assembly. While not intending to be limiting as to the scope of the disclosure provided herein, the substrate support assembly is often referred to herein as the “cathode assembly” or “cathode”. FIG. 1A is a schematic cross-sectional view of a processing chamber 100, in which a plasma 101 is formed during a plasma process that is being performed on a substrate 103. During one or more of the plasma processing methods disclosed herein, an ion-accelerating cathode sheath is generally formed during plasma processing by use of a pulsed-voltage (PV) generator 150 that is configured to establish a pulsed-voltage waveform at a biasing electrode 104 (FIGS. 1A-1B) disposed within a substrate support assembly 136. In some embodiments, the substrate support assembly 136 (FIG. 1A) includes a substrate support 105 and a support base 107. The substrate support 105 can include an electrostatic chuck (ESC) assembly 105D that is configured to “clamp” or “chuck” (e.g., retain) a substrate 103 on a substrate receiving surface 105A. In some embodiments, the biasing electrode 104 forms part of a chucking electrode that is separated from the substrate by a thin layer of a dielectric material 105B (FIG. 1B) formed within the electrostatic chuck (ESC) assembly 105D and optionally an edge control electrode 115 that is disposed within or below an edge ring 114 that surrounds a substrate 103 when the substrate 103 is disposed on the substrate supporting surface 105A of the substrate support assembly 136.

During plasma processing, a vacuum pressure formed in a processing volume 129 of the processing chamber 100 results in poor thermal conduction between surfaces of components disposed therein, such as between the dielectric material of the substrate support 105 and the substrate 103 disposed on the substrate receiving surface 105A, which reduces the substrate support's effectiveness in heating or cooling the substrate 103. Therefore, there is often a need for a thermally conductive inert heat transfer gas, typically helium, to be introduced and maintained at an increased pressure (e.g., backside pressure) within a volume (not shown) disposed between a non-device side surface of the substrate 103 and the substrate receiving surface 105A of the substrate support 105 to improve the heat transfer therebetween. The heat transfer gas, provided by a heat transfer gas source (not shown), flows to the backside volume through a gas communication path (not shown) disposed through the support base 107 and further disposed through the substrate support 105.

In an effort to enable the higher relative pressure to be formed behind the substrate, a clamping voltage is applied to the biasing electrode 104 to “clamp” or “chuck” the substrate 103 to the substrate receiving surface 105A by use of a biasing and clamping network, which is also referred to herein as simply a clamping network 116. In some embodiments, the clamping network 116 includes a DC voltage source P₂ (FIGS. 2A-2B), a blocking resistor R₁ (FIGS. 2A-2B) and a diode D₁ (FIG. 2A), and in some configurations will also include a resistor R₂ (FIGS. 2A-2B) and capacitor C₆ (FIGS. 2A-2B). The presence of the diode D₁ is used to maintain a constant voltage difference between the waveform established at the biasing electrode 104 and the waveform established at the substrate surface. In some embodiments, the PV generator 150 and the clamping network 116 are connected in parallel. The clamping network 116 automatically adjusts a clamping voltage applied to the biasing electrode 104 to maintain a clamping voltage at a desired clamping voltage level to improve the plasma processing process results achieved on the substrate 103 and to assure that a clamped substrate 103 is not damaged during processing, due to the application of too large of a clamping voltage or applying a clamping voltage that is too small. The application of too large of a clamping voltage can increase the “de-chucking” time (e.g., time it takes for charge formed in the substrate to dissipate to reduce the attraction of the substrate 103 to the substrate receiving surface 105A) and/or cause the substrate to break due to the application of too large of a “clamping” or “chucking force” to the substrate 103 and/or cause the thin dielectric between substrate backside and the clamping electrode 104 to breakdown. The application of a clamping voltage that is too small can cause the substrate 103 to lose close contact with the substrate receiving surface 105A during processing. The backside helium can leak into the plasma chamber and plasma species can also leak into a position at the substrate backside, causing abrupt pressure and gas composition change at substrate backside. Such abrupt change can ignite plasma at substrate backside, damaging the substrate and the electrostatic chuck.

The plasma potential of the plasma 101 formed in the processing region 129 varies due to the application of a pulsed voltage (PV) or RF bias to one or more electrodes disposed within a plasma processing chamber. As is discussed further below, to reliably generate a desired clamping voltage V_DCV during a plasma process, the variations in the plasma potential need to be accounted for when controlling a clamping voltage applied to a clamping electrode and substrate 103 during processing. In one example, variations in the plasma potential will occur within each pulse of a multiple pulse PV waveform, and also as the characteristics of PV waveforms delivered by a PV generator 150 change as pulsed voltage biasing parameters applied to a biasing electrode are altered within a substrate-processing-recipe, or from substrate-processing-recipe to substrate-processing-recipe, used to process one or more substrates in the processing chamber. Conventional substrate clamping systems (e.g., electrostatic chucks) that provide a constant clamping voltage and do not take into account and adjust for the fluctuations in the plasma potential often provide poor plasma processing results and/or damage the substrate during processing.

However, the ability to reliably measure or determine in real time the variations in the plasma potential so that they can be accounted for during processing is not a trivial task. The ability to reliably measure the fluctuations or variations in the plasma potential so that the clamping voltage can be desirably adjusted in a production worthy plasma processing chamber that is able to sequentially process multiple substrates in a row is an additional challenge. Conventional methods of measuring the plasma potential and the substrate DC bias requires the use of a probe to directly measure substrate surface potential, are good for non-production laboratory testing, but their presence in the chamber can affect the plasma processing results. The conventional methods of estimating the plasma potential and the substrate DC bias are complicated and require the use of one or more models to correlate the directly measured substrate surface DC bias with the voltage, current and phase data to be measured at an RF match at a few calibration conditions, and use that model to estimate the plasma potential and substrate DC bias when used in production device fabrication processes. The apparatus and methods described herein can be used to reliably determine the plasma potential as a function of time and then provide adjustments to the clamping voltage based on the measured plasma potential.

Plasma Processing Chamber Example

FIG. 1A illustrates the processing chamber 100, in which a complex load 130 (FIGS. 2A-2B) is formed during plasma processing. FIG. 1B is a close-up schematic cross-sectional view of a portion of the substrate support assembly 136 illustrated in FIG. 1A, according to one embodiment. In general, the process chamber 100 is configured to utilize one or more PV generators 150 and/or one or more RF generators 118 to generate, control and maintain a plasma 101 in the processing volume 129 during plasma processing. FIGS. 2A and 2B illustrate different configurations of an electrical circuit, or system, that is configured to deliver a plurality of voltage pulses, provided from the PV generator 150, to the biasing electrode 104 disposed in the plasma processing chamber 100. The PV waveform generator 150 illustrated in FIGS. 2A and 2B is disposed within a first PV source assembly 196 (FIG. 1A) disposed within the processing chamber 100.

The processing chamber 100 is configured to practice one or more of the biasing schemes proposed herein, according to one or more embodiments. In one embodiment, the processing chamber 100 is a plasma processing chamber, such as a reactive ion etch (RIE) plasma chamber. In some other embodiments, the processing chamber 100 is a plasma-enhanced deposition chamber, for example a plasma-enhanced chemical vapor deposition (PECVD) chamber, a plasma enhanced physical vapor deposition (PEPVD) chamber, or a plasma-enhanced atomic layer deposition (PEALD) chamber. In some other embodiments, the processing chamber 100 is a plasma treatment chamber, or a plasma based ion implant chamber, for example a plasma doping (PLAD) chamber. In some embodiments, the plasma source is a capacitively coupled plasma (CCP) source, which includes an electrode (e.g., chamber lid 123) disposed in the processing volume 129 facing the substrate support assembly 136. As illustrated in FIG. 1A, an opposing electrode, such as the chamber lid 123, which is positioned opposite to the substrate support assembly 136, is electrically coupled to ground. However, in other alternate embodiments, the opposing electrode is electrically coupled to an RF generator. In yet other embodiments, the processing chamber 100 may alternately, or additionally, include an inductively coupled plasma (ICP) source electrically coupled to a radio frequency (RF) power supply.

The processing chamber 100 also includes a chamber body 113 that includes the chamber lid 123, one or more sidewalls 122, and a chamber base 124, which define the processing volume 129. The one or more sidewalls 122 and chamber base 124 generally include materials that are sized and shaped to form the structural support for the elements of the processing chamber 100, and are configured to withstand the pressures and added energy applied to them while a plasma 101 is generated within a vacuum environment maintained in the processing volume 129 of the processing chamber 100 during processing. In one example, the one or more sidewalls 122 and chamber base 124 are formed from a metal, such as aluminum, an aluminum alloy, or a stainless steel. A gas inlet 128 disposed through the chamber lid 123 is used to provide one or more processing gases to the processing volume 129 from a processing gas source 119 that is in fluid communication therewith. A substrate 103 is loaded into, and removed from, the processing volume 129 through an opening (not shown) in one of the one or more sidewalls 122, which is sealed with a slit valve (not shown) during plasma processing of the substrate 103. Herein, the substrate 103 is transferred to and from the substrate receiving surface 105A of the substrate support 105 using a lift pin system (not shown).

The processing chamber 100 further includes a system controller 126, which is also referred to herein as a processing chamber controller. The system controller 126 herein includes a central processing unit (CPU) 133, a memory 134, and support circuits 135. The system controller 126 is used to control the process sequence used to process the substrate 103 including the substrate biasing methods described herein. The CPU 133 is a general purpose computer processor configured for use in an industrial setting for controlling processing chamber and sub-processors related thereto. The memory 134 described herein, which is generally non-volatile memory, may include random access memory, read only memory, floppy or hard disk drive, or other suitable forms of digital storage, local or remote. The support circuits 135 are conventionally coupled to the CPU 133 and comprise cache, clock circuits, input/output subsystems, power supplies, and the like, and combinations thereof. Software instructions (program) and data can be coded and stored within the memory 134 for instructing a processor within the CPU 133. A software program (or computer instructions) readable by CPU 133 in the system controller 126 determines which tasks are performable by the components in the processing chamber 100. Preferably, the program, which is readable by CPU 133 in the system controller 126, includes code, which when executed by the processor (CPU 133), perform tasks relating to the monitoring and execution of the electrode biasing scheme described herein. The program will include instructions that are used to control the various hardware and electrical components within the processing chamber 100 to perform the various process tasks and various process sequences used to implement the electrode biasing scheme and method of reliably biasing and clamping a substrate during a plasma process, which are described herein. In one embodiment, the program includes instructions that are used to perform one or more of the operations described below in relation to FIGS. 6A-6B.

In some embodiments, an RF source assembly 163, which includes an RF generator 118 and an RF generator assembly 160, is generally configured to deliver a desired amount of a continuous wave (CW) or pulsed RF power at a desired substantially fixed sinusoidal waveform frequency to the support base 107 based on control signals provided from the controller 126. During processing, the RF source assembly 163 is configured to deliver RF power to the support base 107 disposed proximate to the substrate support 105, and within the substrate support assembly 136. The RF power delivered to the support base 107 is configured to ignite and maintain a processing plasma 101 formed by use of processing gases disposed within the processing volume 129. In some embodiments, the support base 107 is an RF electrode that is electrically coupled to the RF generator 118 via an RF matching circuit 162 and a first filter assembly 161, which are both disposed within the RF generator assembly 160. The first filter assembly 161 includes one or more electrical elements that are configured to substantially prevent a current generated by the output of the PV waveform generator 150 from flowing through an RF power delivery line 167 and damaging the RF generator 118. The first filter assembly 161 acts as a high impedance (e.g., high Z) to the PV signal generated from a PV pulse generator P₁ within the PV waveform generator 150, and thus inhibits the flow of current to the RF matching circuit 162 and RF generator 118.

In some embodiments, the plasma generator assembly 160 and RF generator 118 are used to ignite and maintain a processing plasma 101 using the processing gases disposed in the processing volume 129 and fields generated by the RF power provided to the support base 107 by the RF generator 118. The processing volume 129 is fluidly coupled to one or more dedicated vacuum pumps, through a vacuum outlet 120, which maintain the processing volume 129 at sub-atmospheric pressure conditions and evacuate processing and/or other gases, therefrom. The substrate support assembly 136, disposed in the processing volume 129, is disposed on a support shaft 138 that is grounded and extends through the chamber base 124. However, in some embodiments, the RF generator assembly 160 is configured to deliver RF power to the biasing electrode 104 disposed in the substrate support 105 versus the support base 107.

The substrate support assembly 136, as briefly discussed above, generally includes the substrate support 105 (e.g., ESC substrate support) and support base 107. In some embodiments, the substrate support assembly 136 can additionally include an insulator plate 111 and a ground plate 112, as is discussed further below. The support base 107 is electrically isolated from the chamber base 124 by the insulator plate 111, and the ground plate 112 is interposed between the insulator plate 111 and the chamber base 124. The substrate support 105 is thermally coupled to and disposed on the support base 107. In some embodiments, the support base 107 is configured to regulate the temperature of the substrate support 105, and the substrate 103 disposed on the substrate support 105, during substrate processing. In some embodiments, the support base 107 includes one or more cooling channels (not shown) disposed therein that are fluidly coupled to, and in fluid communication with, a coolant source (not shown), such as a refrigerant source or water source having a relatively high electrical resistance. In some embodiments, the substrate support 105 includes a heater (not shown), such as a resistive heating element embedded in the dielectric material thereof. Herein, the support base 107 is formed of a corrosion resistant thermally conductive material, such as a corrosion resistant metal, for example aluminum, an aluminum alloy, or a stainless steel and is coupled to the substrate support with an adhesive or by mechanical means.

Typically, the substrate support 105 is formed of a dielectric material, such as a bulk sintered ceramic material, such as a corrosion resistant metal oxide or metal nitride material, for example aluminum oxide (Al₂O₃), aluminum nitride (AlN), titanium oxide (TiO), titanium nitride (TiN), yttrium oxide (Y₂O₃), mixtures thereof, or combinations thereof. In embodiments herein, the substrate support 105 further includes the biasing electrode 104 embedded in the dielectric material thereof. In one configuration, the biasing electrode 104 is a chucking pole used to secure (i.e., chuck) the substrate 103 to the substrate receiving surface 105A of the substrate support 105 and to bias the substrate 103 with respect to the processing plasma 101 using one or more of the pulsed-voltage biasing schemes described herein. Typically, the biasing electrode 104 is formed of one or more electrically conductive parts, such as one or more metal meshes, foils, plates, or combinations thereof.

The biasing electrode 104 within the electrostatic chuck (ESC) is electrically coupled to the biasing and clamping network 116, which are illustrated in FIGS. 1A and 2A-2B. The biasing and clamping network 116 includes the DC voltage source P₂. The clamping network 116 automatically maintains a clamping voltage, which is a constant voltage drop across the biasing electrode 104 and the substrate 103 when a plurality PV waveforms are applied to the biasing electrode 104 by the pulsed-voltage waveform generator (PVWG) 150 during plasma processing, leading to improved clamping of the substrate 103. The clamping network 116 is described further below in conjunction with FIGS. 2A-4C. In some embodiments, the clamping network 116 is configured to provide a chucking voltage to the biasing electrode 104, such as static DC voltage between about −10,000 Volts (V) and about 10,000 V, such as −3,000 Volts (V) and about 3,000 V during processing.

Referring to FIG. 1A, the substrate support assembly 136 may further include the edge control electrode 115 that is positioned below the edge ring 114 and surrounds the biasing electrode 104 so that when biased, due to its position relative to the substrate 103, it can affect or alter a portion of the generated plasma 101 that is at or outside of the edge of the substrate 103. The edge control electrode 115 can be biased by use of a pulsed-voltage waveform generator (PVWG) 150 that is different from the pulsed-voltage waveform generator (PVWG) 150 that is used to bias the biasing electrode 104. In some embodiments, the edge control electrode 115 can be biased by use of a pulsed-voltage waveform generator (PVWG) 150 that is also used to bias the biasing electrode 104 by splitting part of the power to the edge control electrode 115. In one configuration, a first PV waveform generator 150 of the first PV source assembly 196 is configured to bias the biasing electrode 104, and a second PV waveform generator 150 of a second PV source assembly 197 is configured to bias the edge control electrode 115. In one embodiment, the edge control electrode 115 is positioned within a region of the substrate support 105, as shown in FIG. 1A. In general, for a processing chamber 100 that is configured to process circular substrates, the edge control electrode 115 is annular in shape, is made from a conductive material and configured to surround at least a portion of the biasing electrode 104, as shown in FIG. 1A. In some embodiments, as illustrated in FIG. 1A, the edge control electrode 115 includes a conductive mesh, foil or plate that is disposed a similar distance (i.e., Z-direction) from the surface 105A of the substrate support 105 as the biasing electrode 104. In some other embodiments, the edge control electrode 115 includes a conductive mesh, foil or plate that is positioned on or within a region of a quartz pipe 110, which surrounds at least a portion of the biasing electrode 104 and/or the substrate support 105. Alternately, in some other embodiments, the edge control electrode 115 is positioned within or is coupled to the edge ring 114, which is disposed adjacent to the substrate support 105. In this configuration, the edge ring 114 is formed from a semiconductor or dielectric material (e.g., AlN, etc.).

Referring to FIG. 1A, the second PV source assembly 197 includes a clamping network 116 so that a bias applied the edge control electrode 115 can be similarly configured to the bias applied to the biasing electrode 104 by the clamping network 116 coupled within the first PV source assembly 196. Applying similarly configured PV waveforms and clamping voltages to the biasing electrode 104 and edge control electrode 115 can help improve the plasma uniformity across the surface of the substrate during processing and thus improve the plasma processing process results. While, for simplicity of discussion, the various methods described herein primarily discuss methods used to determine a desirable clamping voltage V_DCV or DC bias voltage (e.g., equation (15) and/or (16)) that is to be applied to the biasing electrode 104, this discussion is not intended to be limiting as to the scope of the disclosure provided herein since one or more of the operations or methods described herein could also be used to determine and control the bias that is to be applied by the clamping network 116 of the second PV source assembly 197 to the edge control electrode 115. In one example, the operations disclosed in relation to FIGS. 6A-6B could be simultaneously applied to the biasing electrode 104 and edge control electrode 115 during plasma processing.

In some embodiments, the processing chamber 100 further includes the quartz pipe 110, or collar, that at least partially circumscribes portions of the substrate support assembly 136 to prevent the substrate support 105, and/or the support base 107, from contact with corrosive processing gases or plasma, cleaning gases or plasma, or byproducts thereof. Typically, the quartz pipe 110, the insulator plate 111, and the ground plate 112 are circumscribed by a liner 108. In some embodiments, a plasma screen 109 is positioned between the cathode liner 108 and the sidewalls 122 to prevent plasma from forming in a volume underneath the plasma screen 109 between the liner 108 and the one or more sidewalls 122.

FIG. 1B is a close-up view of the substrate support assembly 136 that includes a simplified electrical schematic representation of the electrical characteristics of the various structural elements within one or more embodiments of the substrate support assembly 136. The substrate support assembly 136 includes the substrate support 105, the support base 107, the insulator plate 111 and the ground plate 112, which will each be discussed in turn.

Structurally, in an electrostatic chuck (ESC) 191 version of the substrate support 105, the biasing electrode 104 is spaced apart from the substrate receiving surface 105A of the substrate support 105 by the layer of dielectric material 105B. Typically, electrostatic chucks (ESC) 191 can be categorized into two main classes of electrostatic chucks, which are known as a coulombic ESC or a Johnsen-Rahbek ESC. Depending on the type of electrostatic chuck 191, such as the coulombic ESC or the Johnsen-Rahbek ESC, the effective circuit elements used to describe the electrical coupling of the biasing electrode 104 to the plasma 101 will have some differences. FIG. 1C is a functionally equivalent circuit diagram of a coulombic ESC that can be used in the process chamber illustrated in FIG. 1A, according to one embodiment. FIG. 1D is a functionally equivalent circuit diagram of a Johnsen-Rahbek ESC that can be used in the process chamber illustrated in FIG. 1A, according to one embodiment.

In the simplest case, such as the coulombic ESC case, the dielectric layer 105B will include a capacitance C₁ as shown in FIGS. 1B-1C, 2A and 3B. Typically, the layer of dielectric material 105B (e.g., aluminum oxide (Al₂O₃), etc.) has a thickness between about 0.1 mm and about 1 mm, such as between about 0.1 mm and about 0.5 mm, for example about 0.3 mm. In some embodiments, the dielectric material and layer thickness can be selected so that the capacitance C₁ of the layer of dielectric material is between about 5 nF and about 100 nF, such as between about 7 and about 20 nF, for example.

In a more complex case, such as the Johnsen-Rahbek ESC case, the circuit model includes a capacitance C₁ that is coupled in parallel with a dielectric material resistance R_JR and gap capacitance C_JR as shown in FIG. 1D. In the case of a “Johnsen-Rahbek ESC”, the ESC dielectric layer is “leaky”, in that it is not a perfect insulator and has some conductivity, since, for example, the dielectric material may be a doped aluminum nitride (AlN) having a permittivity (ε) of about 9. Same as for the Coulombic chuck, there is a direct capacitance C₁ between the electrode 104 and the substrate 103 through the thin dielectric 105B and the gap filled with helium. The volume resistivity of the dielectric layer within a Johnsen-Rahbek ESC is less than about 10¹² ohms-cm (Ω-cm), or less than about 10¹⁰ Ω-cm, or even in a range between 10⁸ Ω-cm and 10¹² Ω-cm, and thus the layer of dielectric material 105B can have a dielectric material resistance R_JR in a range between 10⁶-10¹¹ Ωs. Since a gap is typically formed between the substrate support surface 105A and a surface of the substrate 103 a gap capacitance C_JR is used to account for the gas containing spaces between the substrate 103 and substrate support surface 105A. It is expected that the gap capacitance C_JR has a capacitance a bit larger than the capacitance C₁.

For ease of discussion, since the substrate 103 is typically made out of a semiconductor material and/or dielectric material with a thin layer of intrinsic dielectric layer on the bottom and top surfaces, the bottom dielectric layer of the substrate 103 can be considered to be electrically a part of the dielectric layer disposed between the biasing electrode 104 and the substrate receiving surface 105A. Thus, in some applications, the effective capacitance C_E (not shown) formed between the biasing electrode 104 and the top surface of the substrate 103 can be approximated by the combined series capacitance of the dielectric material 105B and the substrate bottom layer (i.e., substrate capacitance C_sub (FIG. 1B)). In the coulombic chuck case, since the substrate capacitance C_sub is typically very large (>100 nF), or the substrate may be conductive (infinite capacitance), the series capacitance is dominated by the capacitance C₁. In the Johnsen-Rahbek ESC case, assuming the substrate capacitance C_sub is typically very large, the effective capacitance for clamping the substrate, C_E will be dominated by the gap capacitance C_JR for DC clamping voltage (FIG. 1D). The finite resistance of the top dielectric 105B results in a resistance R_JR in series with the gap capacitance C_JR, the two of which are in parallel with the direct coupling C₁ between the electrode 104 and the substrate 103. C₁ is the capacitance that couples the RF frequency voltage from the electrode 104 to the substrate 103 during plasma processing.

Referring back to FIG. 1B, the electrical schematic representation of the circuit formed within the substrate support assembly 136 includes a support base dielectric layer capacitance C₂, which represents the capacitance of the dielectric layer positioned between the support base 107 and the biasing electrode 104. In some embodiments, the thickness of the portion of the dielectric material 105C disposed between the support base 107 and the biasing electrode 104 is greater than the thickness of the dielectric material 105B disposed between the biasing electrode 104 and the substrate 103. In some embodiments, the dielectric materials used to form the dielectric layers on either side of the biasing electrode are the same material and form the structural body of the substrate support 105. In one example, the thickness of the dielectric material 105C (e.g., Al₂O₃ or AlN), as measured in a direction extending between the support base 107 and the biasing electrode 104, is greater than 1 mm, such as having a thickness between about 1.5 mm and about 100 mm. The support base dielectric layer capacitance C₂ will typically have a capacitance between about 0.5 and about 10 nanofarads (nF).

The electrical schematic representation of the circuit formed within the substrate support assembly 136 also includes a support base resistance RP, an insulator plate capacitance C₃, and ground plate resistance R_G that is coupled to ground on one end. Since the support base 107 and ground plate 111 are typically formed from a metal material the support base resistance RP and ground plate resistance R_G are quite low, such as less than a few milliohms. The insulator plate capacitance C₃ represents the capacitance of the dielectric layer positioned between the bottom surface of the support base 107 and the top surface of the ground plate 112. In one example, the insulator plate capacitance C₃ has a capacitance between about 0.1 and about 1 nF.

Referring back to FIG. 1A, a PV waveform is provided to the biasing electrode 104 by the PV waveform generator 150 of the first PV source assembly 196 and a PV waveform is provided to the edge control electrode 115 by the PV waveform generator 150 of the second PV source assembly 197. The pulsed voltage waveforms provided to the load (e.g., the complex load 130) disposed within the processing chamber 100. The PV waveform generators 150 include a PV generator P1, such as PV generator P1_A in FIG. 2A or PV generator P1_B in FIG. 2B, that is coupled to the biasing electrode 104 through a power delivery line 157. While not intending to be limiting as to the scope of the disclosure provided herein, and to simplify the discussion, the components within the RF assembly (e.g., RF generator assembly 160 and RF generator 118) and the second PV source assembly 197 are not schematically shown in FIGS. 2A and 2B. The overall control of the delivery of the PV waveform from each of the PV waveform generators 150 is controlled by use of signals provided from the controller 126. In one embodiment, as illustrated in FIGS. 2A and 2B, a PV waveform generator 150 is configured to output a periodic voltage function at time intervals of a predetermined length by use of a signal from a transistor-transistor logic (TTL) source. The periodic voltage function can be two-states DC pulses between a predetermined negative or positive voltage and zero. In one embodiment, a PV waveform generator 150 is configured to maintain a predetermined, substantially constant negative voltage across its output (i.e., to ground) during regularly recurring time intervals of a predetermined length, by repeatedly closing and opening one or more switches at a predetermined rate. In one example, during a first phase of a pulse interval a first switch is used to connect a high voltage supply to the biasing electrode 104, and during a second phase of the pulse interval a second switch is used to connect the biasing electrode 104 to ground. In another embodiment, as illustrated in FIG. 2B, the PV waveform generator 150 is configured to maintain a predetermined, substantially constant positive voltage across its output (i.e., to ground) during regularly recurring time intervals of a predetermined length, by repeatedly closing and opening its internal switch (not shown) at a predetermined rate. In one configuration of the embodiment shown in FIG. 2B, during a first phase of a pulse interval a first switch is used to connect the biasing electrode 104 to ground, and during a second phase of the pulse interval a second switch is used to connect a high voltage supply to the biasing electrode 104. In an alternate configuration of the embodiment shown in FIG. 2B, during a first phase of a pulse interval a first switch is positioned in an open state, such that the biasing electrode 104 is disconnected from the high voltage supply and the biasing electrode 104 is coupled to ground through an impedance network (e.g., inductor and resistor connected in series). Then, during a second phase of the pulse interval the first switch is positioned in a closed state to connect the high voltage supply to the biasing electrode 104, while the biasing electrode 104 remains coupled to ground through the impedance network.

In FIGS. 2A-2B, the PV waveform generators 150 have been reduced to a minimal combination of the components that are important for understanding of its role in establishing a desired pulsed voltage waveform at the biasing electrode 104. Each PV waveform generator 150 will include a PV generator P1_A or P1_B and one or more electrical components, such as but not limited to high repetition rate switches (not shown), capacitors (not shown), inductors (not shown), fly back diodes (not shown), power transistors (not shown) and/or resistors (not shown), that are configured to provide a PV waveform to an output. An actual PV waveform generator 150, which can be configured as a nanosecond pulse generator, may include any number of internal components and may be based on a more complex electrical circuit than what is illustrated in FIGS. 2A-2B. The schematic diagrams of FIGS. 2A-2B are each provided to help explain the fundamental principle of operation of the PV waveform generator, its interaction with the plasma in the processing volume, and its role in establishing a pulsed voltage waveform, such as the pulsed voltage waveform 301 in FIG. 3A or pulsed waveform 401 in FIG. 4A, at the biasing electrode 104.

The power delivery line 157 (FIGS. 1A-1B), electrically connects the output of the PV waveform generator 150 to an optional filter assembly 151 and the biasing electrode 104. While the discussion below primarily discusses the power delivery line 157 of the first PV source assembly 196, which is used to couple a PV waveform generator 150 to the biasing electrode 104, the power delivery line 158 of the second PV source assembly 197, which couples a PV waveform generator 150 to the edge control electrode 115, will include the same or similar components. In FIGS. 2A and 2B, the output of the PV waveform generator 150 is provided to a node N₃. The electrical conductor(s) within the various parts of the power delivery line 157 may include: (a) a coaxial transmission line (e.g., coaxial line 106), which may include a flexible coaxial cable that is connected in series with a rigid coaxial transmission line, (b) an insulated high-voltage corona-resistant hookup wire, (c) a bare wire, (d) a metal rod, (e) an electrical connector, or (f) any combination of electrical elements in (a)-(e). The power delivery line 157, such as the portion of power delivery line 157 within the support shaft 138 (FIG. 1A) and the biasing electrode 104, will have some combined stray capacitance C_stray (not shown) to ground. The optional filter assembly 151 includes one or more electrical elements that are configured to substantially prevent a current generated by the output of the RF generator 118 from flowing through the power delivery line 157 and damaging the PV waveform generator 150. The optional filter assembly 151 acts as a high impedance (e.g., high Z) to RF signal generated by the RF generator 118, and thus inhibits the flow of current to the PV waveform generator 150.

In some embodiments, as shown in FIGS. 1A and 2A-2B, the PV waveform generator 150 of the first PV source assembly 196 is configured to provide a pulsed voltage waveform signal to the biasing electrode 104, and eventually the complex load 130, by delivering the generated pulsed voltage waveforms through node N₃ and the blocking capacitor C₅, the filter assembly 151, high-voltage line inductance Li, and capacitance C₁. The PV waveform generator 150 is connected between a ground node N_G and node N₃. The capacitor C₅ is further connected between the node N₃ and a node N₁ at which the clamping network 116 is attached. The clamping network 116 is connected between the node N₁ and a ground node N_G. In one embodiment, as shown in FIG. 2A, the clamping network 116 includes at least a diode D₁, a capacitor C₆, a DC voltage source P₂, a current-limiting resistor R2, and a blocking resistor R₁. In this configuration, the diode D₁ and the blocking resistor R₁ are connected between the node N₁ and a node N₂, and the capacitor C₆ and the DC voltage source P₂, which is in series with the current-limiting resistor R₂, are connected between the node N₂ and a ground node N_G. In another embodiment, as shown in FIG. 2B, the clamping network 116 includes a capacitor C₆, a DC voltage source P₂, a resistor R₂, and a blocking resistor R₁. In this configuration, the blocking resistor R₁ is connected between the node N₁ and a node N₂, and the capacitor C₆ and the DC voltage source P₂, which is in series with the current-limiting resistor R₂, are connected between the node N₂ and a ground node N_G. In general, the DC voltage source P₂ is used to establish an output voltage of the clamping network 116, which is a voltage difference across the capacitor C₆.

The clamping network 116, when used in combination, as shown in FIGS. 2A and 2B, forms a current suppressing/filtering circuit for the PV waveforms from the PV generator, so that PV waveforms do not induce a significant current through the clamping network 116 to ground. The effect of the clamping network 116 on the operation of the PV generator P1_A (FIG. 2A) or the PV generator P1_B (FIG. 2B) can be made negligible by selecting an appropriately large blocking capacitor C₅ and blocking resistor R₁. The blocking resistor R₁ schematically illustrates a resistor positioned within the components connecting the clamping network 116 to a point within the power delivery line 157, such as node N₁. The main function of the blocking capacitor C₅ is to protect the PV pulse generator P1_A from the DC voltage produced by the DC voltage source P₂, which thus drops across blocking capacitor C₅ and does not perturb the PV waveform generator's output. The value of blocking capacitor C₅ is selected such that while blocking only the DC voltage generated by the DC voltage source P₂, it creates a negligible impedance to the pulsed bias generator's pulsed voltage output provided to node N₃ so that most of the pulsed voltage is delivered to the complex load 130. By selecting a sufficiently large blocking capacitor C₅ capacitance (e.g., 10-80 nF), the blocking capacitor C₅ is nearly transparent for a 400 kHz PV waveform signal, which is generated by the PV waveform generator 150 for example, in that it is much bigger than any other relevant capacitance in the system and the voltage drop across this element is very small compared to that across other relevant capacitors, such as sheath capacitance C_SH and C_WALL (FIG. 2A-2B).

Referring to FIGS. 2A-2B, the purpose of the blocking resistor R₁ in the clamping network 116 is to block the generated pulsed voltage by the PV waveform generator 150 enough to minimize the current it induces in the DC voltage supply P₂. This blocking resistor R₁ is sized to be large enough to efficiently minimize the current (i₁) through it. For example, a resistance of ≥200 kOhm, such as greater than 1 MOhm, or greater than 10 MOhms, or even in a range between 200 kOhms and 50 MOhms is used to make a current generated from the delivery of 400 kHz pulsed voltage signal to node N₁ by the PV waveform generator 150 into the clamping network 116 negligible. The average induced DC current is desirably less than about 40 mA, such as less than 30 mA, or less than 20 mA, or less than 10 mA, or less than 5 mA, or even between 1 μA-20 mA.

In some configurations, the blocking resistor R₁ provides a charging/discharging path that is useful to reset the clamping voltage formed across capacitor C₁, when the diode D₁ is not in the forward bias mode. For example, at the beginning of a plasma process, the substrate is clamped to the electrostatic chuck surface 105A by charging the capacitor C₁ to a predetermined voltage. Such charging current supplied to the capacitor C₁ can be provided by the clamping network 116 through the resistor R₁ (FIGS. 2A and 2B). Similarly, the discharging current from capacitor C₁ in a dechucking step of the substrate can flow through R₁. The charging or discharging current of the capacitor C₁ determines the speed to reach a steady state of either clamping (e.g., chucking) or dechucking of the substrate. Therefore, in some embodiments, the blocking resistor R₁ is selected so that its resistance is not too large, such as less than about 50 MOhms.

In one embodiment of the processing chamber 100, as illustrated in FIG. 5A, an RF waveform is provided to the support base 107, which is positioned at node N₅, by an RF source assembly 163. In some embodiments, the RF source assembly 163 can be a multi-frequency RF source. In this configuration, the RF source assembly 163 is coupled to an electrode, such as the support base 107, via an effective capacitance C₈ that is part of a RF match 162 and a first filter assembly 161, and the clamping network 116 is coupled to the biasing electrode 104 through the power delivery line 157. The RF waveform is provided to the load (e.g., the complex load 130) disposed within the processing chamber 100. The RF source assembly 163 in FIG. 5A, is capacitively coupled to the load 130 via the delivery of RF power to the support base 107. While not intending to be limiting as to the scope of the disclosure provided herein, and to simplify the discussion, one or more PV source assemblies, which are optional in this example, are not schematically shown in FIG. 5A. The overall control of the delivery of the RF waveform is controlled by use of signals provided from the controller 126. As illustrated in FIG. 5B, a sinusoidal RF waveform provided to the processing region 129 from the RF source assembly 163 is provided during a burst period 510, and is halted during a burst-off period 514. In FIG. 5A, the RF source assembly 163 has been reduced to a minimal combination of the components that are important for understanding of its role in establishing a desired RF waveform to the support base 107. As discussed above, the RF source assembly 163 can include components within an RF matching circuit 162 and a first filter assembly 161.

Process Monitoring and Control Examples

In some embodiments, as illustrated in FIG. 1A, the processing chamber 100 further includes a signal detection module 188 that is electrically coupled to one or more electrical components found within a processing chamber 100 by use of a plurality of signal lines 187, which are illustrated in FIG. 1E. FIG. 1E illustrates a schematic view of the processing chamber 100 that includes multiple signal traces 192 that are coupled to various electrical components within the processing chamber 100, and are configured to deliver electrical signals to signal detection elements found within the signal detection module 188. In general, the signal detection module 188 includes one or more input channels 172 and a fast data acquisition module 120. The one or more input channels 172 are each configured to receive electrical signals from a signal trace 192, and are electrically coupled to the fast data acquisition module 120. The received electrical signals can include one or more characteristics of waveforms generated by the PV waveform generator 150 and/or the RF generator 118. In some embodiments, the fast data acquisition module 120 is configured to generate a control signal that is used to automatically control and maintain a substantially constant clamping voltage during processing, leading to improved clamping of the substrate during plasma processing. Further, the fast data acquisition module 120 includes one or more acquisition channels 122. The controller 126 is configured to generate a control signal that is used to automatically control and maintain the clamping voltage based on the signal information provided to the signal detection module 188 through the one or more signal lines 187, processed by components to the fast data acquisition module 120 and then received by the controller 126. The signal information received by the controller 126 can then be analyzed so that a desired real-time adjustment of the voltage applied by the DC voltage supply P₂ of the clamping network 116 can be provided based on the analyzed characteristics of the received signal information.

FIG. 1E schematically illustrates an example of the signal detection module 188 that includes multiple input channels 172 that are each electrically coupled to a corresponding acquisition channel 122 of the fast data acquisition module 120. The multiple input channels 172, such as input channels 172₁-172₃ are coupled to connection points that are positioned in various parts of the first PV source assembly 196, to measure and collect electrical data from these connection points during processing. Additionally, multiple input channels 172, such as input channels 172₄-172_N, are coupled to connection points that are positioned in various parts of the RF source assembly 163 (FIG. 1A), to measure and collect electrical data from one or more points or nodes within the RF source assembly 163 during processing. In one example, input channels 172₄-172_N are configured to detect an RF waveform 181 that is established at different points within the RF source assembly 163 during plasma processing. The multiple input channels 172 may also be coupled to various electrical sensing elements, such as a current sensor, which are configured to measure and collect electrical data at various points within the processing chamber 100. While FIG. 1E schematically illustrates a configuration in which only a few input channels 172 are coupled to points within the first PV source assembly 196 and RF source assembly 163, this configuration is not intended to be limiting as to the scope of the disclosure provided herein since the number of input channels 172 can be increased or decreased as required to control desired chamber processing applications. In some embodiments, one or more input channels 172 are connected to different portions of the first PV source assembly 196, second PV source assembly 197 and RF source assembly 163.

One or more of the input channels 172 can include a conditioning circuit 171, such as, for example, a conditioning circuit 171₁ in input channel 172₁ and a conditioning circuit 171₂ in input channel 172₂. Further, the one or more input channels 172 are configured to generate output waveforms that are conditioned. In some embodiments, the conditioning circuits 171 may each include a voltage divider, a low pass filters, both a voltage divider and a low pass filters, or even in some cases neither a voltage divider nor a low pass filter which is referred to herein as an unattenuated conditioning circuit. Examples of various conditioning circuit elements, such as voltage dividers and filters, and their integration with the input channels are further described in the U.S. Pat. No. 10,916,408, which is herein incorporated by reference in its entirety.

FIG. 1E illustrates a configuration in which the signal detection module 188 includes a plurality of input channels, such as the input channels 172₁-172_N, where N is generally a number greater than one. Each of the input channels 172₁-172_N may be connected to different points within the plasma processing chamber 100. For example, the input channel 172₁ may be connected to an electrical conductor that is used to couple the PV waveform generator 150 to a blocking capacitor C₅ (FIG. 1E). In embodiments where the input channel 172₁ is coupled between the output of the PV waveform generator 150 and the blocking capacitor C₅, the input channel 172₁ receives an input pulsed voltage waveform (e.g., first input pulsed voltage waveform 182 (FIG. 1E)) and the conditioning circuit 171₁ generates an output waveform (e.g., a conditioned waveform). In one example, a received or measured input pulsed voltage waveform includes voltage pulses that include positive and negative voltage levels within different phases of each of the voltage pulses and high frequency oscillations within various phases of a pulse within the input pulsed voltage waveform, which when conditioned by the components, such as a voltage divider, within the conditioning circuit 171₁ forms an output waveform that is at least provided at a lower voltage level due to the use of a voltage divider. In one example, an input pulsed voltage waveform, which includes voltage pulses that include positive and negative voltage levels within different phases of each of the voltage pulses and high frequency oscillations within at least one of the phases of each pulse within the input pulsed voltage waveform, is received by the input channel 172₁ and is then conditioned by the components, such as a voltage divider and a low pass filter, within the conditioning circuit 171₁ to form an output waveform that is a filtered waveform that is at a reduced voltage level. In embodiments where the input channel 172₂ is coupled between the blocking capacitor C₅ and the biasing electrode 104, the input channel 172₂ receives an input pulsed voltage waveform (e.g., second input pulsed voltage waveform) and the conditioning circuit 171₂ generates an output waveform (e.g., a conditioned waveform). In general, the first input pulsed voltage waveform received the input channel 172₁, will have different waveform characteristics from the second input pulsed voltage waveform received the input channel 172₂, due to the position of their respective connection points along the power delivery line 157 within the plasma processing chamber 100.

The fast data acquisition module 120 is generally configured to receive analog voltage waveforms (e.g., conditioned waveforms) and transmit digitized voltage waveforms. The fast data acquisition module 120 includes one or more acquisition channels 122 that are each electrically coupled to a respective conditioning circuit 171 of a first input channel 172, and the fast data acquisition module 120 is configured to generate a digitized voltage waveform from a received conditioned voltage waveform (e.g., output waveform), wherein a data acquisition controller 123 of the fast data acquisition module 120 is configured to determine one or more waveform characteristics of the conditioned voltage waveform by analyzing the first digitized voltage waveform.

As illustrated in FIG. 1E, the fast data acquisition module 120 comprises a plurality of acquisition channels 122₁-122_N, the data acquisition controller 123 and memory 124 (e.g., non-volatile memory). Each of the acquisition channels 122 is electrically coupled to the output of a corresponding one of the input channels 172 such that an acquisition channel 122 receives an output waveform from the corresponding one of the input channels 172. For example, the acquisition channel 122₁ is electrically coupled to the output end of the input channel 172₁ and receives either a first output waveform, depending on the position of the connection point of input end of the input channel 172₁. Further, the acquisition channel 122₂ is electrically coupled to the output end of the input channel 172₂ and receives a second output waveform. Additionally, or alternatively, the acquisition channel 122₃ is electrically coupled to the output end of the input channel 172₃ and receives a third output waveform. The acquisition channel 122_N is electrically coupled to the output end of the input channel 172_N and receives the Nth output waveform.

The data acquisition controller 123 is electrically coupled to an output of each of the acquisition channels 122 and is configured to receive the digitized voltage waveform from each of the acquisition channels 122. Further, the algorithms stored within the memory 124 of the data acquisition controller 123 are adapted to determine one or more waveform characteristics of each of the conditioned waveforms by analyzing each of the digitized voltage waveforms. The analysis may include a comparison of information received in the digitized voltage waveform with information relating to one or more stored waveform characteristics that is stored in memory 124 and is discussed further below.

The data acquisition controller 123 can include one or more of an analog-to-digital converter (ADC) (not shown), a processor 121 (FIG. 1E), communication interface (not shown), a clock (not shown) and an optional driver (not shown). The processor may be any general computing processor. Further, the processor may be a Field Programmable Gate Array (FPGA). The ADC converts the signal within the output waveforms from the analog domain to the digital domain and the output digital signal of the ADC is provided to the processor 121 for processing. The processor 121 of the data acquisition controller 123 determines the one or more waveform characteristics of the output waveform by analyzing the output digital signal provided from the ADC.

In various embodiments, the fast data acquisition module 120 additionally includes memory 124. The memory 124 may be any non-volatile memory. Further, the data acquisition controller 123 is electrically coupled with the memory 124 and is configured to cause waveform characteristics to be stored within the memory 124. In various embodiments, the memory 124 includes instructions executable by the data acquisition controller 123 to cause the data acquisition controller 123 to analyze the received output waveforms and/or transmit information corresponding to determined waveform characteristics based on the analysis of the received output waveforms. A waveform analyzer stored in memory 124 includes instructions executable by the data acquisition controller 123 and when executed cause the data acquisition controller 123 to analyze the output waveforms to determine the waveform characteristics. Information relating to the analyzed waveform characteristics can then be transmitted to one or more of a feedback processor 125 and/or the controller 126. The analysis performed by the data acquisition controller 123 can include a comparison of the waveform characteristics and one or more waveform characteristic threshold values stored in memory.

In some embodiments, the fast data acquisition module 120 is coupled to the feedback processor 125 via a data communication interface 125A, wherein the feedback processor 125 is configured to generate one or more control parameters using one or more waveform characteristics determined by one or more algorithms that are executed by the processor disposed within the data acquisition controller 123. In general, the feedback processor 125 may be any general computing processor. In some embodiments, the feedback processor 125 is generally one of: an external processor connected to the fast data acquisition module 120 via a data communication interface; an internal processor integrated within the fast data acquisition module 120; or is a portion of a substrate processing chamber controller (e.g., controller 126) connected to the fast data acquisition module via a data communication interface. The data acquisition module 120 may communicate information corresponding to one or more of the received output waveforms to the feedback processor 125. For example, the data acquisition module 120 may communicate information related to one or more detected and/or processed waveform characteristics of the one or more of the received output waveforms to the feedback processor 125. Further, the feedback processor 125 may be communicatively coupled with the plasma processing system 100 via a communication link 350 (FIG. 3B).

In various embodiments, the feedback processor 125 includes memory that further contains a software algorithm for instructing a processor within the feedback processor 125 to perform one or more portions of the methods described herein. The one or more algorithms include instructions, which when executed by the processor 121 in the fast data acquisition module cause the fast data acquisition module to process the one or more output waveforms (e.g., conditioned voltage waveforms) to determine one or more waveform characteristics of the received output waveforms. The controller 126, or feedback processor 125 disposed within the controller 126, includes memory that includes instructions, which when executed by a processor (CPU) causes the controller 126 or the feedback processor 125 to generate one or more control parameters using the determined one or more waveform characteristics provided from the fast data acquisition module 120. The instructions executed by the controller 126 or feedback processor 125 may also be further configured to cause the transmission of information, along the communication link 350 (FIG. 3B), that relates to the generated one or more control parameters to the clamping network 116. The clamping network 116 and/or controller 126 may also include memory that includes instructions, which when executed by a processor in the clamping network 116 and/or controller 126 causes the clamping network 116 to establish a desired chucking voltage level at the biasing electrode 104 based on the one or more control parameters generated by the feedback processor 125.

In one or more embodiments, the fast data acquisition module 120 may be electrically (wired or wirelessly) coupled with the controller 126 of the processing chamber 100. For example, the fast data acquisition module 120 transmits data to and/or receives data from the controller 126. For example, the fast data acquisition module 120 communicates information related to one or more waveform characteristics to the controller 126. Further, the processing chamber controller 126 may be communicatively coupled with the clamping network 116 of the processing chamber 100 via the communication link 350. In various embodiments, the processing chamber controller 126 is omitted. An algorithm stored within the memory of the processing chamber controller 126 can include instructions, which when executed by the controller CPU cause various process chamber set points to be adjusted, such as a chucking voltage set point on a chucking power supply, based on the information related to one or more waveform characteristics determined by the data acquisition controller 123.

Clamping Module Control Methods and Hardware Examples

As discussed above, the ability to provide real time control to the clamping voltage level applied to the clamping electrode (e.g., biasing electrode 104) during plasma processing is useful to improve and achieve repeatable plasma processing results and to assure that clamped substrates are not damaged during processing. FIG. 3A illustrates a burst 316 of pulsed voltage waveforms that includes a plurality of waveforms generated and delivered from one or more sources to one or more electrodes disposed in the process chamber 100. For example, waveforms 301-304 are each established at different points within the system 300 (FIG. 3B) by the delivery of a pulsed voltage waveform (not shown) that is generated by the PV waveform generator 150. In some embodiments, a series of individual bursts 316, which are each separated by a burst-off period 314, are provided to the biasing electrode 104. A burst cycle 317, which includes a burst 316 of pulsed voltage waveforms and a serially performed burst-off period 314, can be repeated multiple times during the processing of a substrate, as discussed further below in relation to FIGS. 3A and 4B-4D.

The system 300 is a simplified schematic which is generally represents a portion of the process chamber 100 that includes, for example, the PV waveform generator 150 of the first PV source assembly 196 (FIG. 1A) to the biasing electrode 104 disposed within the substrate support assembly 136. The components within system 300 are used to detect and determine waveform characteristics of one or more PV waveforms delivered from a PV waveform generator 150 by detecting the characteristics of electrical signals detected at different points within the system 300 at different times. Signal lines 321-325 are similar to the plurality of signal lines 187 illustrated in FIG. 1E, and thus are intended to illustrate the connections between various points within the processing system and the input channels 172 (not shown in FIG. 3B) of the signal detection module 188.

As illustrated in FIG. 3A, a plurality of measured PV waveforms 301-304 include a series of pulses that are provided during a PV waveform burst 316. In this example, the last three of a series of pulses are shown within the burst 316. Each of the three pulses within each of the PV waveforms 301-304 have a waveform period T. After the delivery of a burst 316 of pulses, which has a burst-on period 310, the output of the PV waveform generator 150 is stopped so that the system 300 experiences a period of time when no PV waveforms are being generated by the PV waveform generator 150. The time when no PV waveforms are being formed is referred to herein as a non-burst period 314, or “burst off” period 314. Between the burst 316 and a steady state portion of the non-burst period 314 is a transition region, which is referred to herein as a plasma relaxation period 312. At the end of the non-burst period 314, a second burst (not shown) that includes a plurality of pulses is generated and delivered from the PV waveform generator 150. During processing of a substrate it is typical for each burst 316, within a series of bursts, to be separated by the burst-off periods 314, such that a burst 316 and a burst-off period 314 (i.e., burst cycles 317) are serially formed multiple times. Thus, a single burst cycle 317, which includes a burst 316 and a burst-off period 314, has a length that is equal to the sum of a burst-on period 310 (i.e., T_ON) plus the burst-off period 314 (i.e., T_OFF), which is also referred to herein as the burst period T_BD (FIG. 4B). In one example, the burst-on period 310 is between about 100 microseconds (μs) and about 10 milliseconds (ms), such as between about 200 μs and about 5 ms. In one example, the waveform period T_p is between about 1 μs and about 10 μs, such as about 2.5 μs. The burst duty cycle can be between about 5%-100%, such as between about 50% and about 95%, wherein the duty cycle is the ratio of the burst-on period 310 divided by the burst-on period 310 plus the non-burst period 314.

The PV waveform 301 is measured at a point between the blocking capacitor C₅ and the biasing electrode 104, such as illustrated in FIG. 3B as node N₁. The measured voltage is thus related to actual voltage measured at the biasing electrode 104 during the different phases of a processing sequence performed in a processing chamber. The measured voltage of this PV waveform, which is also referred to herein as being electrode voltage V_E, varies over time as a series of bursts 316 and non-burst periods 314 are provided from a PV waveform generator 150 to the biasing electrode 104 during processing. In one embodiment, the PV waveform 301 is measured by an electrical coupling assembly (not shown) that is positioned at node N₁. The electrical coupling assembly is coupled to signal trace 324 that is in communication with an input channel 172 within the signal detection module 188.

In some embodiments, a PV waveform (not shown) generated at the output of the PV waveform generator 150 is measured and utilized in one or more of the processes described herein by measuring the voltage formed at an electrical coupling assembly (not shown) that is positioned at node N₃. The PV waveform measured at the PV waveform generator 150 will closely track the PV waveform 301, and will have a measured voltage that is offset from the PV waveform 301 by an amount that is at least related to the set point of DC voltage supply P₂. In this configuration, as shown in FIG. 3B, the electrical coupling assembly is coupled to signal trace 321 that is in communication with an input channel 172 within the signal detection module 188.

The PV waveform 302 is intended to represent the voltage established on a substrate 103 during processing due to the delivery of the PV waveforms provided from the PV waveform generator 150. As shown in FIG. 3A, the PV waveform 302 tracks the measured PV waveform 301 very closely, such that the PV waveform 302 is typically considered to be offset a fixed amount from the PV waveform 301. The offset voltage formed during processing between the electrode 104 and substrate 103 is referred to herein as clamping voltage, and is primarily set by the set point of DC voltage supply P₂. In some configurations, PV waveform 302 can be measured by a voltage probe in good contact with the frontside or backside of the substrate 103 and attached to signal trace 322 that is in communication with an input channel 172 within the signal detection module 188. In most process chamber hardware configurations the substrate voltage is not easy to directly measure due to ESC hardware limitations, measurement signal integrity issues and capacitive coupling between chamber component related issues. By use of the methods described herein the need for the direct measurement of the substrate voltage can be avoided by use of the various measurement techniques and processes described herein.

In some embodiment, the PV waveform 303 is measured at a node directly coupled to a second conductor plate positioned within the processing chamber 100. In one embodiment, the second conductor plate is the support base 107, which is positioned at node N₅ in FIG. 3B. As shown in FIG. 3B, the second conductor plate is positioned between the capacitance C₃ and the capacitance C₂, which are intended to represent the capacitances formed by the presence of the insulator plate 111 and the dielectric layer 105C, respectively. The measured voltage is thus related to actual voltage measured at the support base 107 during the different phases of a processing sequence performed in a processing chamber. The measured voltage of the PV waveform, which is referred to herein as being voltage V_c, varies over time as a series of bursts 316 and non-burst periods 314 are provided from a PV waveform generator 150 to the biasing electrode 104 during the processing. In some embodiments, the PV waveform 303 is formed by use of a source, such as the RF source assembly 163 which is coupled to the chamber through the conductor plate 107. The PV waveform 303 can be measured by use of an electrical coupling assembly, which is positioned at node N₅, and is coupled to signal trace 323 that is in communication with an input channel 172 within the signal detection module 188.

In some embodiments, the PV waveform 304 is measured at a node directly coupled to a PV source 150. The measured voltage of the PV waveform 304, which is referred to herein as being voltage VR, varies over time as a series of bursts 316 and non-burst periods 314 are provided from a PV waveform generator 150. In some embodiments, the PV waveform 304 is configured to achieve a desired voltage V4 during the non-burst periods 314, and thus does not electrically float during the non-burst periods 314. In some embodiments, the PV waveform 304 is configured to electrically float during the non-burst periods 314. The PV waveform 304 can be measured by use of an electrical coupling assembly that is coupled to the signal trace 321 that is coupled to node N₃ and is in communication with an input channel 172 within the signal detection module 188.

FIG. 4A illustrates an example of a PV waveform 401 to the biasing electrode 104 by use of a PV waveform generator 150 within a PV source assembly during a portion of the waveform period T_p and a DC voltage source P₂ in the assembly 116. The PV waveform 402, as shown in FIG. 4A, includes a series of PV waveforms established at the substrate (e.g., V_w) due to the PV waveform 401 being established at the biasing electrode 104 by a PV waveform generator 150 and a DC voltage source P₂. The PV waveforms 401 and 402 are intended to illustrate more detailed examples of portions of the PV waveforms 301 and 302 that are illustrated in FIG. 3A.

The output of the PV waveform generator 150, which can be controlled by a setting in a plasma processing recipe stored in the memory of the controller 126, forms the PV waveform 401 that includes a peak-to-peak voltage, which is also referred to herein as the pulse voltage level V_pp. The PV waveform 402, which is the waveform seen by the substrate 103 due to the delivery of the PV waveform 401, is characterized as including a sheath collapse and recharging phase 450 (or for simplicity of discussion the sheath collapse phase 450) that extends between point 420 and point 421, a sheath formation phase 451 that extends between point 421 and point 422, and an ion current phase 452 that extends between point 422 and back to the start at point 420 of the next sequentially established pulse voltage waveform. Depending on the desired plasma processing conditions, it is desirable to control and set at least the PV waveform characteristics, such as PV waveform frequency (1/T_P), pulse voltage level V_pp, pulse voltage on-time, and/or other parameters of the PV waveforms within a burst 316 to achieve desirable plasma processing results on a substrate. In one example, pulse voltage (PV) on-time, which is defined as the ratio of the ion current time period (e.g., time between point 422 and the subsequent point 420 in FIG. 4A) to the waveform period T_p, is varied from one plasma processing recipe to another to adjust the etch rate. In some embodiments, the PV on-time is greater than 50%, or greater than 70%, such as between 80% and 95%.

FIG. 4B illustrates PV waveforms in which a series of bursts 462 of pulsed voltage waveforms are established at the biasing electrode 104 and the substrate surface. In the example illustrated in FIG. 4B, a plurality of pulses 461 within each burst 462 include a series of PV waveforms 401 that are established at the biasing electrode 104. In this example, each of the bursts 462 includes pulses 461 that have a PV waveform that has a consistent pulsed voltage shape (e.g., constant voltage magnitude is provided during a portion of each PV waveform 401), a burst delivery length T_ON that does not vary from one burst 462 to another over time and a burst rest length T_OFF that does not have a varying length over time. The burst rest length T_OFF is formed by halting the delivery of the PV waveforms provided during the burst delivery length T_ON time for a period of time. The duty cycle of the bursts 462, which is the ratio of the length of time the plurality of pulses are delivered during the burst (i.e., burst delivery length T_ON) divided by the duration of a burst period (i.e., T_BD=T_ON+T_OFF), is also constant in this example. For clarity of discussion, the total of the burst on period 310 and the burst off period 314, referenced in FIG. 3A, is intended to be equivalent to the burst period T_BD referenced in FIG. 4B. One will appreciate that in other processing methods, the plurality of pulses 461 could include negative pulse waveforms, shaped pulse waveforms or positive pulse waveforms, or combinations thereof. As illustrated in FIG. 4B, during the burst rest length T_OFF the biasing electrode potential curve 436 is primarily controlled by the chucking voltage that is applied and controlled by the bias compensation module 116, and thus may be at a different voltage level than the plasma potential.

FIG. 4C illustrates a PV waveform in which a multilevel series of bursts 490 of pulsed voltage waveforms are established at an electrode, such as the biasing electrode 104. During processing the series of bursts 490, which include a plurality of bursts 491 and 492 and a burst off period 493, are provided to the biasing electrode 104. The series of bursts 490, which include a series of bursts 491 and 492 that are followed by a burst off period 493, can be sequentially repeated one or more times. In one example, each of the plurality of bursts 491 and 492 include a plurality of PV waveforms 401 that are supplied at different voltage levels as illustrated by the difference in the levels of each of the peaks of the each of the bursts 491 and 492. In some embodiments, the transition from burst 492 to a burst 491 is separated by the burst off period 493, while the transition from burst 491 to burst 492 is not separated by the burst off period 493. The series of bursts 490 are often referred to as being a “high-low” series of bursts due to the pulse waveforms provided to the biasing electrode 104 during the burst 491 having a higher pulse voltage level V_pp (FIG. 4A) than the pulse waveforms provided to the biasing electrode 104 during the burst 492. The bursts 491 often referred to herein as including a “high” pulse voltage level V_pp and busts 492 often referred to herein as including a “low” pulse voltage level V.

FIG. 4D illustrates a PV waveform in which a multilevel series of bursts 494 of pulsed voltage waveforms are established at an electrode, such as the biasing electrode 104. The series of bursts 494, which also includes the plurality of bursts 491 and 492 and the burst off period 493 that are oriented in a different “low-high” configuration, can be sequentially repeated one or more times. In one example, each of the plurality of bursts 491 and 492 include a plurality of PV waveforms 401 that are supplied at different voltage levels as illustrated by the difference in the levels of each of the peaks of the each of the bursts 491 and 492. In some embodiments, the transition from burst 492 to a burst 491 is not separated by the burst off period 493, while the transition from burst 491 to the next burst 492 is separated by the burst off period 493.

FIG. 5B illustrates a burst 516 of RF waveforms that includes a plurality of waveforms generated and delivered from an RF source assembly 163 to an electrode disposed in the process chamber 100. For example, waveforms 502-504 are each established at different points within the system 500 (FIG. 5A) by the delivery of an RF waveform 501 that is generated by the RF generator 118. Waveforms 502-504 include the waveform V_w formed at the substrate, waveform V_S formed at the surface of the electrostatic chuck, and the waveform V_E formed at the biasing electrode 104, respectively. FIG. 5A illustrates an example of a configuration of a system 500 that is used to detect and determine waveform characteristics of one or more RF waveforms delivered from a RF generator 118 by detecting the characteristics of electrical signals detected at different points within the system 500 at different times. The signal traces 322-325 are similar to the plurality of signal traces 192 of the plurality of signal lines 187 illustrated in FIG. 1E, and thus are intended to illustrate the connections between various points within the processing system and the input channels 172 (not shown in FIG. 3B) of the signal detection module 188.

As illustrated in FIG. 5B, a plurality of measured RF waveforms 501-504 include a series of pulses that are provided during an RF burst 516. In this example, two cycles of the RF waveform are shown within the burst 516. The measured RF waveforms, such as RF waveforms 501-504, have a waveform frequency that is controlled by the RF generator 118, which can be between 100 kHz and 120 MHz. After the delivery of the RF burst 516, which has a burst period 510, the output of the RF generator 118 is stopped so that the system 500 experiences a period of time when no RF waveforms are being generated by the RF generator 118. The time when no RF waveforms are being formed is referred to herein as a non-burst period 514, or “burst off” period 514. Between the burst 516 and a steady state portion of the non-burst period 514 is a transition region, which is referred to herein as a plasma relaxation period 512. At the end of the non-burst period 514, a second burst (not shown) that includes a plurality of RF waveforms is generated and delivered from the RF generator 118. During processing of a substrate it is typical for each burst 516, within a series of bursts, to be separated by the non-burst periods 514, such that the series of bursts 516 and non-burst periods 514 are serially formed multiple times. In one example, the burst period 510 is between about 20 microseconds (μs) and about 100 milliseconds (ms), such as between about 200 μs and about 5 ms. The burst duty cycle can be between about 5%-100%, such as between about 50% and about 95%, wherein the duty cycle is the ratio of the burst period 510 divided by the burst period 510 plus the non-burst period 514.

Plasma Potential Analysis

To reliably generate a desired clamping voltage V_DCV during a plasma process, the variations in the plasma potential need to be accounted for when delivering a clamping voltage to a clamping electrode during processing. As discussed above, the ability to reliably measure and monitor the plasma potential in a processing chamber that is configured to serially process multiple substrates in a production environment is a non-trivial task. In one or more of the embodiments of the disclosure provided herein, the plasma potential is determined based on measurements made at different points within the plasma processing system during different portions of a substrate processing sequence. FIG. 6A illustrates a processing method that can be used to measure, monitor and control attributes of a plasma formed in a processing chamber so that the desirable clamping voltage can be reliably controlled and applied to a clamping electrode disposed within a substrate support. It is assumed that the capacitance C₁ is a net series capacitance of the dielectric layer 105B, and the gap between dielectric surface 105A and substrate backside, and a possible thin dielectric layer on substrate backside surface.

A plasma potential curve 433, which is shown in FIG. 4A, illustrates the local plasma potential during the delivery of the PV waveform 401 that is established at the biasing electrode 104 by use of a PV waveform generator 150. During processing the plasma potential generally remains at or close to zero volts throughout most of the burst-on period 310 and during the burst-off period 314. The plasma potential will achieve its peak value (V_PL) during the sheath collapse phase 450, which coincides with time T₁ in FIGS. 3A and 4A. Additionally, at time T₁, when the multiphase PV waveform 401 has reached its peak value, the voltage at the biased electrode (e.g., biasing electrode 104) will be equal to the output voltage (V_BCM) supplied by the DC voltage source P₂. Therefore, the fluctuation in the plasma potential can be on the order of 1 kV or larger and thus substrate clamping systems that do not take into account the fluctuations in the plasma potential due to the delivery of a bias to one or more electrodes within the processing chamber 100 can lead to poor plasma processing results and/or damage to the substrate. Referring to FIG. 4A, the times T₂ and T₃ illustrate the start of the burst-off period and the end of the transition period 312, respectively. The period of time between times T₂ and T₃ is referred to herein as the plasma relaxation time, which is generally the time it takes the plasma to extinguish once the PV waveforms and RF power delivery has been halted during the burst-off period 314. The time T₄ is intended to represent a measurement time that is positioned after the transition period 312 has ended and before the next burst 316 (not shown) has started.

In an effort to provide the desired clamping voltage (V_DCV) at node N₁, the set point of the DC voltage source P₂, V_BCM at node N₂, is adjusted by use of computer implemented instructions that are configured to determine and thus account for the variation in the plasma potential. The desired clamping voltage (V_DCV) set point is generally equal to the peak plasma potential (V_PL), which is affected by the plasma processing conditions and the substrate surface material, plus the clamp voltage set point (V_clamp) for the type of electrostatic chuck that is being used during processing. The desired clamping voltage set point (V_DCV) can thus be written as shown in equation (1).V_DCV=V_PL+V_Clamp  (1)

The clamp voltage set point (V_clamp) is a constant voltage value that has been determined through prior testing and evaluation of the electrostatic chucking characteristics of the actual electrostatic chuck or type of electrostatic chuck (e.g., columbic electrostatic chuck). The prior testing and evaluation results are used to determine a minimum substrate clamping force voltage to assure that the substrate will have a good thermal contact to the dielectric surface 105A and negligible helium will leak through the outer sealing band of the substrate support 105 when the substrate is clamped to the surface of the substrate support 105 during plasma processing. The clamp voltage set point (V_clamp) value will vary due to the type of electrostatic chuck that is being used (e.g., columbic or Johnsen-Rahbek electrostatic chuck), backside gas pressure being used during processing, and temperature of the dielectric 105A during plasma processing.

In some embodiments of the clamping network 116, a diode D₁ electrically connects nodes N₁ and N₂ (See FIGS. 2A and 3B), and is configured to only allow current to flow in a direction from node N₁ to N₂ (i.e., anode side of the diode D₁ is coupled to node N₁ and cathode side of diode D₁ is coupled to node N₂). Due to the configuration of the diode D₁, the voltage at node N₁ is restricted to a value that is no higher than the voltage at node N₂ (V_BCM) at all times. Thus, during each pulse period T_p (FIG. 3A) of a PV waveform, the peak voltage at node N₁ is reset to the voltage of node N₂ (V_BCM), which is the output voltage of the DC voltage source P₂ at steady state when a large capacitance C₆ (e.g., between 0.5 and 10 μF) is used. The peak voltage at node N₁ is the sum of peak plasma potential V_PL and the actual clamp voltage across capacitor C₁. To achieve the clamp voltage set point, the set point of DC voltage source P₂ (V_BCM) should be equal to the desired clamping voltage set point (V_DCV), as shown in below in a rewritten version of equation (1).V_BCM=V_DCV=V_PL+V_Clamp

However, in some embodiments of the clamping network 116, no diode D₁ is used to connect nodes N₁ and N₂ (see FIGS. 2B and 5A). In this configuration, the voltage of node N₂ (V_BCM) will still be equal to the voltage of the DC voltage source P₂ at steady state given large C₆ (e.g. between 0.5 and 10 μF). In some embodiments, the resistor R₁ and capacitor C₆ values in the clamping network 116 are selected so that the time constant of R₁*C₆ is much larger than the burst period T_BD (FIG. 4B), so that the voltage at node N₂ is substantially constant within one burst period T_BD. Since nodes N₁ and N₂ are connected through a high resistance value resistor, resistor R₁, the time-averaged (in a burst period T_BD) voltage at node N₁ will equal the time-averaged (in a burst period T_BD) voltage at node N₂, which is equal to the clamping voltage V_BCM. To achieve the clamp voltage set point V_Clamp across capacitor C₁, the time-averaged (in a burst period T_BD) voltage at node N₁ should be the time-averaged substrate (in a burst period T_BD) voltage plus the clamp voltage set point V_Clamp. As is discussed further below, the time-averaged substrate voltage can be approximated by use of the PV waveform generated by the PV waveform generator 150 and the peak plasma potential V_PL. Thus, the set point of DC voltage source P₂ (DC voltage source output voltage V_BCM) can be determined by the pulser voltage waveform, the peak plasma potential V_PL, and the clamp voltage set point V_Clamp.

As illustrated in FIG. 4A, the plasma potential V_Plasma (i.e., curve 433) is equal to or near zero for most of the time during processing and reaches a peak level at time T₁. Therefore, to determine the peak plasma potential (V_PL) at the surface of the substrate, which is formed at time T₁, measurements that account for all of the various factors that will affect the plasma potential are measured and used to adjust the output voltage (V_BCM) of the DC voltage source P₂ to achieve the desired clamping voltage (V_DCV). To determine the peak plasma potential (V_PL) it is first assumed that, and the system is configured such that, charge conservation is maintained at one or more nodes within an electrode biasing circuit at a time scale of a burst period T_BD. In some embodiments, as shown in FIGS. 2A-2B, 3B and 5A, charge conservation is maintained at node N₁ of the electrode biasing circuit. In one embodiment, capacitors C₁, C₂, and C₅ are directly coupled to the node N₁, and it is assumed that the inductor Li (e.g., line inductance illustrated in FIGS. 2A-2B) is small enough to induce negligible voltage oscillation compared to the PV voltages generated by the PV waveform generator 150. As illustrated in FIGS. 2A-2B, 3B and 5A, node N₁ is also coupled to the resistor R₁ and then to the capacitor C₆. Thus, the total charge (Q_T) flowing through resistor R₁ at the time scale of a burst period T_BD is on the order of Q_T≈T_BD*V_BCM/R₁. In some embodiments, the resistance of the resistor R₁ is selected such that it is large enough, so that the charge flowing through resistor at the time scale of a burst period T_BD, is negligible compared to the charges stored in capacitors C₁, C₂, and C₅, for example. Therefore, in this configuration, the presence of a large blocking resistor R₁ will cause the capacitor C₆ to functionally appear to not be directly coupled to node N₁, and the electrostatic charge associated to node N₁ will be the sum of electrostatic charge stored in the capacitors C₁, C₂, and C₅, which are directly coupled to the node N₁.

Equation (2) below is used to describe charge conservation at a node in the electrode biasing circuit, meaning the sum of electrostatic charge Q_Burst measured during a portion of the burst-on period 310 is equal to the amount of stored charge Q_off measured during the burst-off period 314 right after burst-on period 310.ΣQ_Burst=ΣQ_Off  (2)

FIGS. 2A-2B and 3A, provide system configuration examples in which the charges stored within the capacitances C₁, C₂, and C₅ can be assumed to be conserved, and thus allow the peak plasma potential V_PL to be determined by use of one or more of the methods described herein. The electrical signals that are detected within the one or more methods described herein can include one or more characteristics of waveforms generated by the PV waveform generator 150 and/or the RF generator 118. The detected one or more waveform characteristics can include but are not limited to voltage at one or more times within a pulse, slope at one or more times within a pulse, a pulse period, and pulse repetition frequency. However, the assumption that charge is conserved in a region surrounding a node, such as node N₁ in FIGS. 2A-2B and 3B, is limited by or depends on the amount of stored charge lost due to a magnitude of the current that flows to ground, such as the current (FIG. 3B) that flows through the blocking resistor R₁ to ground. As is further discussed below, the ability to accurately determine the peak plasma potential V_PL is dependent on the blocking resistor's ability to assure that the amount of charge lost is negligible prior to the signal detection module 188 measuring the generated electrical signals during one or more of the phases of a burst sequence, which includes at least one burst-on period 310 and burst-off period 314. As noted above, it is desirable that the resistance of the resistor R₁ be >100 kOhm, for example.

Example 1

In one example, based on the system 300 configuration illustrated in FIGS. 3A-3B, for node N₁, equation (2) can be rewritten as shown in equation (3).C₁(ΔV₁)_Burst+C₂(ΔV₂)Burst+C₅(ΔV₅)_Burst=C₁(ΔV₁)_Off+C₂(ΔV₂)_Off+C₅(ΔV₅)_Off  (3)

In equation (3), C₁, C₂, and C₅ are the capacitances that are known, and ΔV₁, ΔV₂, and ΔV₅ are the voltages of the capacitor plates directly coupled to node N₁ minus the voltages of the opposing capacitor plates for the capacitances C₁, C₂, and C₅, which are measured during either the burst-on period 310 or burst-off period 314. Therefore, if the measurement made during the burst-on period 310 is made at one of the time T₁ instants in time, and the measurement made during the burst-off period 314 is made at time T₄, equation (3) can be rewritten as equation (4).C₁(V₁−V_PL)+C₂(V₁−V₅)+C₅(V₁−V₃)=C₁(V₂−0)+C₂(V₂−V₆)+C₅(V₂−V₄)  (4)

In equation (4), the voltage V₁ is the voltage of the electrode 104 at time T₁ during the burst-on period 310, the peak plasma potential V_PL is the plasma potential at time T₁ during the burst-on period 310, the voltage V₅ is the voltage measured at node N₅ at time T₁ during the burst-on period 310, the voltage V₃ is the voltage measured at node N₃ at time T₁ during the burst-on period 310, the voltage V₂ is the voltage measured at node N₁ during the burst-off period 314, the voltage V₆ is the voltage measured at node N₅ during the burst-off period 314, and the voltage V₄ is the voltage measured at node N₃ during the burst-off period 314. As noted above, the plasma potential is effectively zero during the burst-off period and thus the charge stored in the capacitor C₁ during the burst-off period 314 is effectively equal to the voltage V₂ times the capacitance C₁. The actual clamping voltage during burst off period is V₂. Therefore, after reorganizing equation (4), which is shown in equation (5), the peak plasma potential V_PL can be found by solving the equation (5) for the system configuration illustrated in FIG. 3B.

$\begin{array}{cc}{V}_{\mathrm{PL}}=\left(V_1-V_2\right)\left(1+\frac{C2+C5}{C1}\right)+\frac{C2}{C1}\left(V_6-V_5\right)+\frac{C5}{C1}\left(V_4-V_3\right)& \left(5\right)\end{array}$

For simplicity of discussion, each of the capacitance terms that are multiplied by the voltage difference terms in equation (5), and any of the equations provided below, are generally referred to herein as a “combined circuit capacitance” that has a combined circuit capacitance value, which is determined by the arithmetic combination of the capacitances (e.g., capacitances C₁, C₂, and C₅ in equation (5)) based on the configuration of the various connected circuit elements (e.g., electrostatic chuck 191, RF generator 118, and PV waveform generator 150) relative to a desired node (e.g., node N₁).

However, in configurations where the biasing element (e.g., PV source 150) connected at node N₃ floats during the burst-off period, or is disconnected from ground during the burst-off period, the capacitor C₅ which is directly coupled to node N₃ will not have current through it during the burst-on to burst-off transition. In other words, the charge stored in capacitor C₅ is the same during the burst-on to burst-off transition, so its effect can be removed from the charge conservation equations (2), (3) and (4). The equation used to find the voltage V_PL can be simplified to equation (6).

$\begin{array}{cc}{V}_{\mathrm{PL}}=\left(V_1-V_2\right)\left(1+\frac{C2}{C1}\right)+\frac{C2}{C1}\left(V_6-V_5\right)& \left(6\right)\end{array}$

In some embodiments, there is negligible current flowing through the RF source assembly 163 to the capacitor C₂ coupled at node N₅ during the burst-on to burst-off transition, so that the majority of current flowing through C₂ also flows through C₃. Thus, the series of C₂ and C₃ can be treated as one capacitor of value (C₂C₃)/(C₂+C₃) and grounded. Thus, in equation (6), V₅=V₆=0 and C₂ is replaced by (C₂C₃)/(C₂+C₃).

$\begin{array}{cc}{V}_{\mathrm{PL}}=\left(V_1-V_2\right)\left(1+\frac{C2C3}{C1\left(C2+C3\right)}\right)& \left(7\right)\end{array}$

Therefore, since the capacitance C₁ is typically much larger that the capacitances C₂ and C₃ in most systems, equation (7) can be reduced to the simple equation, in this example, of a floating biasing element, shown in equation (8).V_PL≈V₁−V₂  (8)

In any case, using either equations (5), (6), (7) or (8), the knowledge of the capacitance values of C₁, C₂, C₃, and/or C₅, and the measured voltages detected during the burst-on period 310 and burst-off period 314 by use of the signal detection module 188, the peak plasma potential V_PL can be calculated so that the desired clamping voltage V_DCV can be determined.

Example 2

In another example, the biasing element (e.g., PV source 150) connected at node N₃ is controlled at a constant voltage V₄ (such as zero) during the burst-off period. In some embodiments, there is negligible current flowing through the RF source assembly 163 to the capacitor C₂ coupled at node N₅ during the burst-on to burst-off transition, so that the majority of current flowing through C₂ also flows through C₃. Then the voltage V_PL can be found by solving equation (9) for the system configuration illustrated in FIG. 3B.

$\begin{array}{cc}{V}_{\mathrm{PL}}=\left(V_1-V_2\right)\left(1+\frac{C2C3+C2C5+C3C5}{C1\left(C2+C3\right)}\right)+\frac{C5}{C1}\left(V_4-V_3\right)& \left(9\right)\end{array}$

In this case, using equation (9), the knowledge of the capacitance values of C₁, C₂, C₃, and/or C₅, and the measured voltages during the burst-on period 310 and burst-off period 314 by use of the signal detection module 188, the peak plasma potential V_PL can be calculated so that the desired clamping voltage V_DCV can be determined.

Example 3

In another example, based on the system 500 configuration illustrated in FIG. 5A, equation (2) can be rewritten as shown in equation (10). In this example, as schematically shown in FIG. 5A, the RF source assembly 163 is connected to node N₅ and is utilized to generate substrate bias voltage during plasma processing. In this example, a PV waveform generator 150 is not connected to the system 500. Equation (2) can thus be rewritten as shown in equation (10).C₁(ΔV₁)_Burst+C₂(ΔV₂)_Burst=C₁(ΔV₁)_Off+C₂(ΔV₂)_Off  (10)

Therefore, voltage V_PL can be found by use of equation (11).

$\begin{array}{cc}{V}_{\mathrm{PL}}=\left(V_1-V_2\right)\left(1+\frac{C2}{C1}\right)+\frac{C2}{C1}\left(V_6-V_5\right)& \left(11\right)\end{array}$

In this case, using equation (11), the knowledge of the capacitance values of C₁ and C₂, and the measured voltages during the burst period 510 and burst-off period 514 by use of the signal detection module 188, the peak plasma potential V_PL can be calculated so that the desired clamping voltage V_DCV can be determined.

Plasma Processing Method Examples

FIG. 6A is a process flow diagram of a method 600 for determining a desired clamping voltage based on the application of a process recipe that is used during plasma processing of a substrate in a processing chamber. In addition to FIG. 6A, the method 600 is described in reference to FIGS. 1A-5B. In one embodiment, the method 600 can be performed by executing, by the CPU 133, computer implemented instructions that are stored within the memory 134 of the controller 126. In one embodiment, the method 600 at least includes a clamping voltage determination process 605, which includes operations 606-614.

At operation 602, a processing recipe is initiated in a processing chamber 100, which causes a plasma 101 to form in the processing region 129 of the processing chamber 100. In some embodiments, during this operation, the RF source assembly 163 delivers enough RF power at an RF frequency to an electrode within the processing chamber to form the plasma 101. In one example, the RF source assembly 163 delivers RF power at an RF frequency of between 400 kHz and 100 MHz, such as 40 MHz to the support base 107 disposed within the substrate support assembly 136. The RF power delivered to the support base 107 is configured to ignite and maintain a processing plasma 101 formed by use of processing gases disposed within the processing volume 129.

At operation 604, the controller 126 sends a command signal to the DC voltage source P₂ to initiate and establish a first clamping voltage at the biasing electrode 104. The magnitude of the first clamping voltage is set to the clamping voltage in the recipe which is stored in the memory of the controller 126. The recipe set point is generally set to a level that through initial testing or by general knowledge has a magnitude that is low enough to not cause breakdown of top dielectric layer within the substrate support, but has a magnitude high enough to achieve good thermal contact with the substrate receiving surface 105A in order to seal substrate backside gas (e.g. helium) sufficiently.

At operation 606, in one embodiment, the PV waveform generator 150 begins to generate a series of PV waveforms that establishes a PV waveform at the biasing electrode 104. During operation 606, the PV generator 150 can be configured to generate and provide bursts 316 of PV waveforms to the biasing electrode 104 within the processing chamber 100. In an alternate embodiment, the RF source assembly 163 begins to generate bursts of RF waveforms, as discussed in relation to FIG. 5B, at an electrode (e.g., support base 107) within the processing chamber 100.

In some embodiments, during operation 606, it is desirable for the pulse voltage level (e.g., V_pp) applied to the electrode, such as the biasing electrode 104, to be controlled at a desired ramp rate that is not greater than the rate to charge or discharge the capacitors C₅ and C₆ through resistors R₁ and R₂ respectively, (FIG. 3B), so that the actual clamp voltage across capacitor C₁ remains constant while V_DCV and VPS are ramped together with the pulse voltage V. If such ramp rate relation is satisfied, according to equation (1), the actual clamp voltage across capacitor C₁ will be kept close to the clamp voltage set point V_clamp during the pulse voltage ramping. The charge or discharge rate of capacitor C₅ through resistor R₁ is determined by the RC time constantτ₁=R₁(C₅)  (12).

The charge or discharge rate of capacitor C₆ through resistor R₂ is determined by the RC time constantτ₂=R₂(C₆)  (13).

Therefore, the ramp time for the pulse voltage level V_pp change should be larger than the RC time constants τ₁ and τ₂. In some embodiments, the ramp time for the pulse voltage level V_pp is set to be at least three times of the larger of the RC time constants τ₁ and τ₂.

At operation 608, while ramping pulse voltage level (e.g., V_pp) applied to the biasing electrode 104, the signal detection module 188 is used to monitor the waveforms established within different portions of the processing chamber 100 during the execution of the plasma processing recipe. In one example, the signal detection module 188 is configured to monitor the waveforms established at the biasing electrode 104 and the support base 107 over time, while the pulsed voltage level is ramped. In one example, the waveforms established at the biasing electrode 104 and the support base 107 can be detected by measuring waveform signals established at nodes N₁ and N₅ within the system 300 or 500 illustrated in FIG. 3B or 5A, respectively. In general, during operation 608 the signal detection module 188 is used to continuously monitor or repetitively sample the waveform signals established at the various nodes within a system over time, such as detecting the waveform signals at one or more of the times T₁-T₅ illustrated in FIG. 3A, 4A or 5B.

At operation 610, the information collected during operation 608 is used to calculate the plasma potential during the plasma process by use of at least one equation that is derived from equation (2), such as equations (5), (6), (7), (8), (9) or (11). The desired equation that is to be used to determine the peak plasma potential V_PL is based on a knowledge of the system configuration that is being used during plasma processing and/or a setting found in the software instructions stored within memory. Typically, when the pulse voltage level, RF power, or other plasma relevant parameters (e.g. pressure, gas composition, etc.) are changing during plasma processing, one or more of the relevant equations, which are incorporated within the instructions stored in the memory of the controller 126, can be used during the execution of the stored instructions by the CPU 133 to determine the peak plasma potential V_PL at any time during processing.

At operation 612, the desired clamping voltage V_DCV that is to be used during a subsequent portion of the current plasma process is determined by use of equation (1) and the results of operation 610. As discussed above, the clamp voltage set point (V_Clamp) found in equation (1) is the clamp voltage set point in the recipe, typically a predetermined value that is stored within the memory of the controller 126.

At operation 614, a command signal is then sent by the controller 126, or the feedback processor 125, to the DC voltage source P₂ so that a desired clamping voltage V_DCV can be applied to the biasing electrode 104 by setting the DC voltage source P₂ voltage properly, as discussed above. In some embodiments, operations 606-614 of the clamping voltage determination process 605 are repeated during the pulse voltage ramping phase at least one more time, or until a desired pulse voltage level (e.g., V_pp) is achieved during a burst-on period 310 during plasma processing. In some other embodiments, only operations 608-614 of the clamping voltage determination process 605 are repeated one or more times during plasma processing. In one example, operations 608-614 are repeated one or more times once the desired pulse voltage level (e.g., V_pp) has been achieved during a burst-on period 310.

After a steady state value for the pulse voltage level (e.g., V_pp) has been achieved after performing operation 608-614 one or more times, operation 616 is performed in which the set point of the DC voltage source P₂, or DC voltage source output voltage V_BCM, is stored in memory. In some embodiments, it is desirable to store intermediate set point of DC voltage source output voltage V_BCM values (e.g., non-final values determined during the pulse voltage ramping phase) in memory so that they can be used as a baseline in a future plasma processing sequence. The set point of DC voltage source output voltage V_BCM that is stored in memory can be used in future plasma processes performed on additional substrates that are processed using the same or similar plasma processing recipe. As briefly discussed above, plasma processing recipes generally include one or more processing steps that are adapted to control one or more plasma processing parameters performed on a substrate disposed within a processing chamber. The one or more plasma processing parameters can include PV waveform characteristics (e.g., duty cycle, pulse voltage level V_pp, burst period, burst off period, pulse voltage on-time, etc.), chamber pressure, substrate temperature, gas flow rates, gas composition, and other useful parameters. For example, the PV waveform generator 150 is set to provide pulses having a pulse voltage level (e.g., V_pp) from 0.01 kV to 10 kV and the DC voltage source output voltage V_BCM of the clamping network 116 is set to a constant DC voltage between −3 kV to +3 kV, such as +2.5 kV.

Referring to FIG. 4C, in some embodiments, the generated series of PV waveforms formed during operation 606 include establishing a series of PV waveforms within a burst 490. The “low” pulse voltage level V_pp found in the PV waveforms formed during burst 492 have a magnitude that is significantly less than the “high” pulse voltage level V_pp found in burst 491. The “high” pulse voltage level V_pp found in burst 491 will have the greatest effect on the desired clamping voltage V_DCV set point, due to larger peak-to-peak pulse voltage. Therefore, in some embodiments, since the system is configured such that charge conservation is maintained within a region of an electrode biasing circuit that is coupled to an electrode, the peak plasma potential VPS achieved during the “high” pulse voltage level V_pp can be used to determine the set point of DC voltage source output voltage V_BCM even though the “low” pulse voltage level V_pp containing burst 492 is positioned between the “high” pulse voltage level V_pp containing burst 491 and the burst off period 493. In one example, one of the equations (5), (6), (7), (8), (9) or (11) can be used to determine the plasma potential during the plasma processing.

Referring to FIG. 4D, in some embodiments, the generated series of PV waveforms formed during operation 606 include establishing a series of PV waveforms within a burst 494. In some embodiments, since the system is configured such that charge conservation is maintained within a region of an electrode biasing circuit, the peak plasma potential VPS achieved during the “high” pulse voltage level V_pp can be determined and used to determine the set point of DC voltage source output voltage V_BCM even though the “low” pulse voltage level V_pp containing burst 492 is positioned between the “high” pulse voltage level V_pp containing burst 491 and the burst off period 493. Therefore, equations (5), (6), (7), (8), (9) or (11) can be used to determine the plasma potential during the plasma processing.

FIG. 6B is a process flow diagram of a method 650 that is used to deliver a desired clamping voltage V_DCV based on the determination of the set point of DC voltage source output voltage V_BCM in a prior plasma processing sequence, such as after performing method 600 at least once. The method 650 can be performed by executing, by the CPU 133, computer implemented instructions that are stored with the memory 134 of the controller 126.

At operation 652, a processing recipe is initiated in a processing chamber by forming a plasma 101 in the processing region 129 of a processing chamber. Operation 652 can be performed in a manner similar to methods described above in relation to operation 602.

At operation 654, the controller 126 sends a command signal to the DC voltage source P₂ to initiate and establish a first clamping voltage at the biasing electrode 104. The magnitude of the first clamping voltage is set based on a set point in the processing recipe or stored in the memory of the controller 126. In one embodiment, the stored set point is based on a DC voltage source output voltage V_BCM value used during a prior performed process, such as a result from the performance of one of the operations found in method 600.

At operation 656, in one embodiment, the PV waveform generator 150 begins to generate a series of PV waveforms that establish a PV waveform at the biasing electrode 104. In an alternate embodiment, the RF source assembly 163 begins to generate an RF waveform that establishes a RF waveform at an electrode, such as the support base 107, within the processing chamber 100. As discussed above in relation to operation 606, the pulse voltage level (e.g., V_pp) applied to the electrode is ramped within a time period that is larger (e.g. twice or three times larger) than the RC time constants to charge or discharge C₅ through R₁ and C₆ through R₂. Typically, operation 656 is performed in a manner that is similar to methods described above in relation to operation 606.

At operation 659, at the same time of operation 656, a command signal is sent by the controller 126, or the feedback processor 125, to the DC voltage source P₂ to reach a set point of the DC voltage source output voltage V_BCM so that a desired clamping voltage V_DCV is applied to and maintained at the biasing electrode 104 by the DC voltage source P₂ during at least a portion of the processing steps performed on the substrate. The method 650 can additionally be performed on all of the substrates that subsequently processed in the processing chamber. However, if one or more plasma processing recipe parameters are altered in any subsequent plasma processes it may be desirable to perform method 600 and then method 650 on all of the subsequent processes performed using these altered plasma processing recipe parameters.

In some embodiments, steps 608-614 in method 600 can be used repeatedly within a processing step to adjust for plasma property and peak plasma potential V_PL drift which results in different DC voltage source output voltage V_BCM of the DC voltage source P₂ in order to maintain the clamp voltage set point V_Clamp.

DC Bias Analysis Example

In some embodiments, the amount of DC bias (V_{DC Bias}) applied to a substrate during processing is calculated and then used to adjust one or more of the processing parameters during one or more portions of a plasma processing recipe. The DC bias at any time during plasma processing, in which a symmetric waveform (e.g., sinusoidal waveform (RF waveform) or sigmoidal waveform) is delivered, can be calculated by use of equation (14).V_{DC Bias}=(V_PL−V_pp)/2  (14)

During one or more of the operations described herein, the signal detection module 188 and controller 126 are used to detect and monitor the waveform signals established at the various nodes within a system over time, so that one or more computer implemented instructions can be used to determine the DC bias and/or peak DC bias.

Aspects of one or more of the embodiments disclosed herein include a system and method of reliably biasing and clamping a substrate during processing to improve the plasma processing results performed on a plurality of substrates.

While the foregoing is directed to embodiments of the present disclosure, other and further embodiments of the disclosure may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.

Claims

1. A method for plasma processing a substrate, comprising: generating a plasma within a processing region of a processing chamber, wherein the processing region comprises a substrate support that comprises a substrate supporting surface, a first biasing electrode, and a first dielectric layer disposed between the first biasing electrode and the substrate supporting surface;delivering, from a waveform generator, one or more waveforms to the first biasing electrode through a first power delivery line during a first time period;halting the delivery of the one or more waveforms to the first biasing electrode for a second time period;applying, from a clamping network, a first clamping voltage to the first biasing electrode;detecting at least one characteristic of the one or more waveforms during the first time period by receiving an electrical signal from a signal trace that is coupled to the first power delivery line at a first point disposed on the first power delivery line;detecting at least one characteristic of an electrical signal received from the signal trace during the second time period; andadjusting the first clamping voltage applied to the first biasing electrode based on: the detected characteristic of the one or more waveforms received from the signal trace during the first time period; andthe detected at least one characteristic of the electrical signal received from the signal trace during the second time period.

2. The method of claim 1, wherein a plurality of pulses are provided from the waveform generator during the first time period, each of the plurality of pulses have a pulse voltage level, and during a second portion of the first time period the pulse voltage level of one or more pulses of the plurality of pulses is increased relative to one or more pulses provided within a first portion of the first time period.

3. The method of claim 2, wherein applying the first clamping voltage during the portion of the first time period comprises increasing the voltage supplied by the clamping network to the biasing electrode.

4. The method of claim 1, wherein the clamping network comprises: a direct-current (DC) voltage source coupled between the first point and ground; anda blocking resistor coupled between the first point and the DC voltage source.

5. The method of claim 4, wherein the one or more waveforms each comprise a pulse voltage level, and during a portion of the first time period the pulse voltage level is increased from a first voltage level to a second voltage level.

6. The method of claim 5, wherein the first power delivery line comprises a blocking capacitor that is disposed between the waveform generator and the biasing electrode, and the voltage of the clamping network ramps at a substantially equivalent rate to the ramp of the voltage across the blocking capacitor.

7. The method of claim 4, wherein the blocking resistor has a resistance greater than 100 kOhms.

8. The method of claim 4, wherein a DC current flowing through the blocking resistor to ground at any instant in time less than about 20 mA.

9. The method of claim 1, wherein applying the first clamping voltage during a portion of the first time period comprises increasing the voltage supplied by the clamping network to the biasing electrode.

10. The method of claim 1, wherein applying the first clamping voltage during a portion of the first time period comprises decreasing the voltage supplied by the clamping network to the biasing electrode.

11. The method of claim 1, further comprising determining a peak plasma potential achieved during the first time period by analyzing: the at least one characteristic of the one or more waveforms detected during the first time period; andthe at least one characteristic of the electrical signal detected during the second time period.

12. The method of claim 11, wherein adjusting the first clamping voltage further comprises: adding the determined peak plasma potential to an clamp voltage set point constant value stored in memory to form a desired clamping voltage; anddelivering a control signal to a direct-current (DC) voltage source of the clamping network, wherein the control signal comprises information relating to the formed desired clamping voltage.

13. The method of claim 1, wherein the detecting the at least one characteristic of the one or more waveforms during the first time period comprises detecting a first voltage at a peak of a pulsed-voltage waveform of the one or more pulsed-voltage waveforms, andand detecting the at least one characteristic of the one or more waveforms during the second time period comprises detecting a second voltage during the second time period.

14. The method of claim 13, wherein adjusting the first clamping voltage applied to the first biasing electrode based on the detected characteristic further comprises: determining a difference between the first voltage and the second voltage; anddetermining a plasma potential value based on the determined difference between the first voltage and the second voltage, andwherein the adjusting the first clamping voltage comprises delivering a substrate biasing voltage to the first biasing electrode, and the substrate biasing voltage comprises a sum of the determined plasma potential value and a previously determined clamp voltage set point value.

15. The method of claim 14, wherein the determining the plasma potential value further comprises multiplying the determined difference between the first voltage and the second voltage by a combined circuit capacitance value, and the combined circuit capacitance value comprises capacitance values of circuit elements directly coupled to the first point.

16. A method for plasma processing a substrate, comprising: generating a plasma within a processing region of a processing chamber, wherein the processing region comprises a substrate support that comprises a substrate supporting surface, a first biasing electrode, and a first dielectric layer disposed between the first biasing electrode and the substrate supporting surface;delivering, from a waveform generator, a plurality of pulsed-voltage waveforms to the first biasing electrode through a first power delivery line during a first time period, wherein the first power delivery line comprises a blocking capacitor that is disposed between the waveform generator and the biasing electrode;halting the delivery of the plurality of pulsed-voltage waveforms to the first biasing electrode during all of a second time period;applying, from a clamping network, a first clamping voltage to the first biasing electrode;detecting at least one characteristic of one or more of the delivered plurality of pulsed-voltage waveforms during the first time period by receiving an electrical signal from a signal trace that is coupled to the first power delivery line at a first point disposed between the blocking capacitor and the biasing electrode;detecting at least one characteristic of an electrical signal received from the signal trace during the second time period; andadjusting the first clamping voltage applied to the first biasing electrode based on the detected characteristic of the one or more of the delivered plurality of pulsed-voltage waveforms and the at least one characteristic of the electrical signal received from the signal trace during the first and second time periods.

17. The method of claim 16, wherein a plurality of pulsed-voltage waveforms are provided from the waveform generator during the first time period, each of the plurality of pulses have a pulse voltage level, and during a second portion of the first time period the pulse voltage level of one or more pulses of the plurality of pulses is increased relative to one or more pulses provided within a first portion of the first time period.

18. The method of claim 17, wherein applying the first clamping voltage during the portion of the first time period comprises increasing the voltage supplied by the clamping network to the biasing electrode.

19. The method of claim 16, wherein the clamping network comprises: a direct-current (DC) voltage source coupled between the first point and ground; anda blocking resistor coupled between the first point and the DC voltage source.

20. The method of claim 19, wherein the one or more waveforms each comprise a pulse voltage level, and during a portion of the first time period the pulse voltage level is increased from a first voltage level to a second voltage level.

21. The method of claim 19, wherein a DC current flowing through the blocking resistor to ground at any instant in time less than about 20 mA.

22. The method of claim 16, further comprising determining a peak plasma potential achieved during the first time period by analyzing: the at least one characteristic of the one or more waveforms detected during the first time period; andthe at least one characteristic of the electrical signal detected during the second time period.

23. The method of claim 22, wherein adjusting the first clamping voltage further comprises: adding the determined peak plasma potential to an clamp voltage set point constant value stored in memory to form a desired clamping voltage; anddelivering a control signal to a direct-current (DC) voltage source of the clamping network, wherein the control signal comprises information relating to the formed desired clamping voltage.

24. The method of claim 16, wherein the detecting the at least one characteristic of the one or more waveforms during the first time period comprises detecting a first voltage at a peak of a pulsed-voltage waveform of the one or more pulsed-voltage waveforms, andand detecting the at least one characteristic of the one or more waveforms during the second time period comprises detecting a second voltage during the second time period.

25. The method of claim 24, wherein adjusting the first clamping voltage applied to the first biasing electrode based on the detected characteristic further comprises: determining a difference between the first voltage and the second voltage; anddetermining a plasma potential value based on the determined difference between the first voltage and the second voltage, andwherein the adjusting the first clamping voltage comprises delivering a substrate biasing voltage to the first biasing electrode, and the substrate biasing voltage comprises a sum of the determined plasma potential value and a previously determined clamp voltage set point value.

26. The method of claim 25, wherein the determining the plasma potential value further comprises multiplying the determined difference between the first voltage and the second voltage by a combined circuit capacitance value, and the combined circuit capacitance value comprises capacitance values of circuit elements directly coupled to the first point.