Model-based purge gas flow

Abstract

An embodiment herein provides a method for processing a semiconductor substrate. The method described herein includes receiving a first input corresponding to a first geometric hardware configuration of a processing chamber, receiving a second input corresponding to a first process recipe of the processing chamber, determining a first purge gas flow rate for the processing chamber based on the first input and the second input, measuring a deposition characteristic of the processing chamber via a first sensor, determining a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic, and flowing a purge gas at the second purge gas flow rate during a deposition process.

Description

[Technical field]

[0001] The present disclosure relates generally to methods for processing semiconductor substrates, and in particular to the deposition, modification, or removal of semiconductor film material on a substrate. In particular, the present disclosure relates to a method and apparatus for determining purge gas flow in a processing chamber. [Background technology]

[0002] Semiconductor substrates are processed for a wide variety of applications, including the fabrication of integrated and microdevices. One method of processing a substrate includes depositing a material, such as a semiconducting or conductive material, on the top surface of the substrate. The deposition, modification, or removal of the semiconducting material on the substrate relies primarily on the flow of a purge gas and a process gas across the surface of the substrate. During a deposition operation, a process gas, such as an inert gas, and a purge gas flow simultaneously into a processing chamber, such as an epitaxial (EPI) chamber. One purpose of the purge gas is to reduce the possibility that the process gas will expand beyond the substrate into the hard to reach areas of the processing chamber and deposit an undesired layer of film on the chamber surface. The undesired layer of film results in an increase in avoidable maintenance and a subsequent decrease in overall throughput. For example, these sporadic, undesired depositions can result in coatings on the backside of the substrate support, coatings on optically transparent surfaces, and / or coatings and / or condensation between liners or around lift and rotate mechanisms inside the processing kit, which can result in particle problems and process variations. If there is no purge gas or the purge gas flow is too low, additional particle problems can also occur when unexhausted precursors or process particles are exposed to air during substrate transfer.

[0003] If the purge gas flow rate is too low, the purge gas will not sufficiently reduce the likelihood of the process gas reaching these hard-to-clean areas of the processing chamber to prevent the problems described above, but if the purge gas flow rate is too high, the purge gas will dilute the process gas and displace it, slowing the deposition process, reducing throughput, and potentially creating non-uniformities across the substrate during the deposition process, resulting in process variations.

[0004] Therefore, a need exists for a method for improving purge gas flow into a processing chamber. Summary of the Invention

[0005] In one embodiment, a method for processing a semiconductor substrate is provided, the method includes receiving a first input corresponding to a first geometric hardware configuration of a processing chamber, receiving a second input corresponding to a first process recipe for the processing chamber, running a first deposition process in the processing chamber using the first process recipe at a first purge gas flow rate, measuring a deposition characteristic of the first deposition process via a first sensor, determining a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic, and running the second deposition process at the second purge gas flow rate.

[0006] In another embodiment, a method is provided that includes receiving a first input corresponding to a first hardware configuration of a processing chamber, receiving a second input corresponding to a first process recipe for the processing chamber, running a first deposition process using the first process recipe at a first purge gas flow rate in the processing chamber, measuring a deposition characteristic of the first deposition process at a first location via a first sensor, determining a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic, running a second deposition process using the second purge gas flow rate in the processing chamber, measuring a deposition characteristic of the second deposition process at the first location via the first sensor, and determining a third purge gas flow rate different from the second purge gas flow rate based on a change in the measured deposition characteristic.

[0007] In yet another embodiment, a non-transitory computer readable medium is provided that includes instructions that, when executed by a processor, cause the processor to receive a first input corresponding to a first geometric hardware configuration of a processing chamber, receive a second input corresponding to a first process recipe for the processing chamber, run a first deposition process in the processing chamber using the first process recipe at a first purge gas flow rate, measure a deposition characteristic of the first deposition process via a first sensor, determine a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic, and run the second deposition process at the second purge gas flow rate.

[0008] In a manner in which the above-mentioned features of the present disclosure may be understood in detail, a more particular description of the present disclosure, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the accompanying drawings. It should be noted, however, that the accompanying drawings are merely illustrative of exemplary embodiments, and that the present disclosure is open to other equally effective embodiments and therefore should not be considered as limiting in scope. [Brief description of the drawings]

[0009] [Figure 1A] 1 is a schematic cross-sectional view of a processing chamber that can be used to practice various embodiments of the present disclosure. [Figure 1B] FIG. 1B is a schematic cross-sectional view of the processing chamber of FIG. 1A rotated 90 degrees clockwise. [Diagram 2] FIG. 1 illustrates a method for adjusting a purge gas flow rate according to an embodiment of the present disclosure. [Figure 3A] FIG. 13 shows a graphical comparison of measured deposition characteristics of a lower window in a processing chamber as a function of both different hardware configurations and different purge gas flow rates. [Figure 3B] FIG. 13 shows a graphical comparison of measured deposition characteristics of a lower window in a processing chamber as a function of both different hardware configurations and different purge gas flow rates. [Figure 3C] FIG. 13 shows a graphical comparison of measured deposition characteristics on a substrate in a processing chamber as a function of both different hardware configurations and different purge gas flow rates. [Figure 3D] FIG. 13 shows a graphical comparison of measured deposition characteristics of the backside of a substrate support in a processing chamber as a function of both different hardware configurations and different purge gas flow rates. [Figure 3E] FIG. 3B is a comparative table of measured deposition characteristics of a portion of the backside of the substrate support, the substrate, and the window in the chamber as a function of both different hardware configurations and different purge gas flow rates according to the graphs of FIGS. 3A-3D. [Figure 4] FIG. 2 is a schematic diagram of a controller according to some embodiments of the present disclosure. [Diagram 5] FIG. 1 illustrates a processing system according to an embodiment of the present disclosure. [Figure 6] 2 is a schematic cross-sectional view of another processing chamber that can be used to practice various embodiments of the present disclosure. [Figure 7] 1 is a schematic cross-sectional view of yet another processing chamber that can be used to practice various embodiments of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] For ease of understanding, wherever possible, like reference numerals will be used to refer to like elements common to the figures, and it is contemplated that elements and features of one embodiment may be beneficially incorporated in other embodiments without further recitation.

[0011] The present disclosure relates to methods for purging a processing chamber. The methods provided herein can increase throughput and deposition uniformity, reduce total production downtime, and shorten cleaning duration.

[0012] Instead of a constant purge gas flow rate, embodiments of the present disclosure provide a variable purge gas flow rate that can be adjusted for changes in geometric hardware configurations and process recipes. As used herein, the phrase "process recipe" refers to a finite number of process conditions for performing a given operation on a substrate, such as a semiconductor wafer. According to embodiments of the present disclosure, adjusting the purge gas flow rate does not substantially dilute the process gas on the substrate, but successfully reduces the likelihood that the process gas will expand beyond the substrate into the hard and reach areas of the processing chamber and deposit an undesirable film layer. In turn, less avoidable maintenance (hereafter, less production downtime) is performed, and increased throughput is achieved. Furthermore, by adjusting the purge gas flow rate during or prior to a deposition operation, the methods described herein can improve overall process uniformity, such as center-to-edge (CE) deposition reaction uniformity.

[0013] FIG. 1A is a schematic cross-sectional view of a processing chamber 100 that can be used to implement various considerations in the present disclosure. FIG. 1B is a schematic plan view of the processing chamber of FIG. 1A. The processing chamber 100 is utilized to grow an epitaxial film on a substrate, such as a substrate 102. The substrate 102 may be, for example, a 300 mm wafer formed from silicon. However, other substrates are contemplated. The processing chamber 100 is an epitaxial (EPI) chamber, although other chambers, such as a rapid thermal processing (RTP) chamber or an etch chamber, are also contemplated (see FIGS. 6 and 7).

[0014] The processing chamber 100 includes an upper body 156, a lower body 148 disposed below the upper body 156, and a flow module 112 disposed between the upper body 156 and the lower body 148. Optionally, cooling channels can be disposed within the upper body 156 and / or the lower body 148. Collectively, the upper body 156, the flow module 112, and the lower body 148 form a chamber body. In some embodiments, the chamber body can be constructed of a process-resistant material, such as aluminum or stainless steel. Disposed within the chamber body are a substrate support 106, an upper window 108, a lower window 110, a plurality of upper lamps 141, and a plurality of lower lamps 143. In some embodiments, the upper window 108 and / or the lower window 110 comprise transparent quartz. The substrate support 106 is disposed between the upper window 108 and the lower window 110. In some embodiments, the substrate support 106 is formed from a material with high thermal resistance (e.g., black quartz or aluminum nitride), or a ceramic or graphite material coated with a silicon material such as silicon carbide. The substrate support 106 can be formed from any material that can withstand a high temperature environment, such as the high temperature environment of a chemical vapor deposition (CVD) process. In some embodiments, the substrate support 106 is an annular ring. In other embodiments, the substrate support 106 is a disk / plate-like shape that protects the bottom surface of the substrate from unwanted deposition during the deposition process. During operation, the processing chamber 100 creates a flow of precursors across the top surface 150 of the substrate 102.

[0015] The plurality of upper lamps 141 are disposed between the upper window 108 and the lid 154. The plurality of upper lamps 141 form an upper lamp assembly 147. The lid 154 includes a plurality of sensors 153 disposed therein for measuring a temperature within the process chamber 100. The plurality of lower lamps 143 are disposed between the lower window 110 and the floor 152. The plurality of lower lamps 143 form a lower lamp assembly 145. In some embodiments, the upper lamp assembly 147 and / or the lower lamp assembly 145 can be disposed about 3 centimeters (cm) or less from the substrate support 106, such as about 2 cm or less above the substrate support 106. In some embodiments, the lamps 141 and / or 143 can be disposed more than 3 cm away from the substrate support 106. The lamps 141 and / or 143 can be disposed above and / or below the substrate support 106. In some embodiments, the cooling channel 146 acts as a heat sink to cool the lower lamp assembly 145 and / or the upper lamp assembly 147 .

[0016] A process volume 136 is formed between the upper window 108 and the lower window 110. The process volume 136 includes a substrate support 106 disposed therein. The substrate 102 is disposed on a top surface of the substrate support 106 during processing. The operating pressure within the process chamber 100 can be reduced to a subatmospheric pressure prior to introducing process gases via the inlet port 180. An optional edge ring 125 surrounds the substrate support 106. The edge ring 125 may be part of the substrate support 106 or may be in contact with the substrate support 106. In some embodiments, the substrate 102 is disposed on a top surface of the edge ring 125. The process chamber 100 further includes a preheat ring 133. The preheat ring 133 is configured to be disposed around the substrate support 106 while the substrate support 106 is in the processing position. In some embodiments, overlapping substrate supports are utilized such that a gap 188 between the preheat ring 133 and the substrate support 106 is minimized. The overlap can be achieved by additional substrate support components coupled to the substrate support 106. The gap 188 can be from about 0 mm to about 5 mm, such as from about 0.5 mm to about 4.5 mm, such as from about 1 mm to about 4 mm, such as from about 2 mm to about 3 mm. In some embodiments, the gap 188 can be non-uniform or asymmetric (e.g., if the substrate support is non-uniform or asymmetric, such as an ellipse). In some embodiments, as shown in FIG. 1A, the preheat ring 133 is coupled to a liner 163 disposed on an inner surface of the flow module 112 such that an outer or lower surface of the preheat ring 133 is coupled to an inner or upper surface of the liner 163. In some embodiments, the preheat ring 133 can be substantially "L" shaped. In some embodiments, a "gap" exists in the annular preheat ring 133 at one or more exhaust outlets 116, as shown in FIG. 3A.

[0017] The substrate support 106 is attached to a shaft 118. The shaft is connected to a motion assembly 121. The motion assembly 121 includes one or more actuators and / or adjustment devices that provide movement and / or adjustment of the shaft 118 and / or the substrate support 106 within the process volume 136. The motion assembly 121 includes a rotary actuator 122 that rotates the shaft 118 and / or the substrate support 106 about a longitudinal axis A of the processing chamber 100. In some embodiments, the rotor actuator 122 can be configured to rotate the substrate 102 at a rotational spin rate of about 2 Hz or greater, such as about 4 Hertz (Hz) or greater. In some embodiments, the rotor actuator 122 can be configured to rotate the substrate 102 at a rotational spin rate of about 1 Hz or less, such as about 0.5 Hz or less. In some embodiments, the motion assembly 121 further includes a vertical actuator 124 for raising and lowering the substrate support 106 in the z-direction. The motion assembly includes a tilt adjustment device 126 that is used to adjust the planar orientation of the substrate support 106, and a lateral adjustment device 128 that is used to adjust the left-right position of the shaft 118 and substrate support 106 within the process volume 136. In some embodiments, the substrate support 106 may be stationary. In other embodiments, the substrate support 106 may be rotatable about a vertical axis (e.g., the x-axis).

[0018] The substrate support 106 may further include lift pin holes 107 disposed therein. The lift pin holes 107 are sized to accommodate lift pins 132 for placing the substrate 102 on the substrate support 106 and / or lifting the substrate 102 off the substrate support 106 either before or after a deposition process is performed, respectively. The lift pins 132 may rest against lift pin stops 134 when the substrate support 106 is lowered from the processing position to the transfer position.

[0019] The flow module 112 includes a plurality of process gas inlets 114, a plurality of flow module purge gas inlets 164, and one or more exhaust gas outlets 116. One or more cooling channels 146 are disposed below the plurality of process gas inlets 114. As shown in FIG. 1A, one cooling channel 146 is shown below the plurality of process gas inlets 114 and above the slit valve 135, and another cooling channel 146 is shown below the plurality of flow module purge gas inlets 164. The cooling channel 146 may be annular. Slit purge gas may flow from a slit inlet (not shown) in the slit valve 135 and exit from either one or more exhaust gas outlets 116 or a slit outlet (not shown) in the slit valve 135. Additionally, holes or slits in the liner 163 may allow exhaust of gas to the gas outlet 116. It is contemplated that the location of the slit inlet and slit outlet may be switched depending on the configuration selected. In some embodiments, the rotating purge gas may flow from a rotating inlet (not shown) at the bottom of the processing chamber 100 around the shaft 118 and exit through one or more exhaust gas outlets 116. In some embodiments, each of the different purge gases may be exhausted through an exhaust vent, such as a slit or hole in the liner 163, before reaching the one or more gas outlets 116.

[0020] The liner 163 is disposed on the inner surface of the flow module 112 to protect the flow module 112 from reactive gases used during the deposition process. A liner purge gas can flow between the liner 163 and the chamber body and exit through one or more exhaust gas outlets 116. The flow module purge gas, slit purge gas, rotary purge gas, and / or liner purge gas can be an inert gas, such as one or more of helium, hydrogen, nitrogen, and argon. The flow module purge gas, slit purge gas, rotary purge gas, and / or liner purge gas can further include an etchant to affect the selectivity of the deposited film or cleaning area being purged. In some embodiments, the liner 163 is divided into an upper liner and a lower liner, and the lower liner has holes formed along its inner surface to form a vented lower liner, which allows gases in the lower portion of the chamber to be vented directly to an exhaust, such as the exhaust pump 157. The vented lower liner can have about 1 to about 30 holes, such as about 5 to about 20 holes, and each hole can be about 1 mm to about 6 mm, such as about 3 mm to about 4 mm. The purpose of the flow module purge gas, slit purge gas, liner purge gas, and rotary purge gas is to create a local positive pressure to prevent an increase in precursor gas and / or process gas concentration in the area where each purge gas flows, and to facilitate maintaining chamber cleaning both during and between processing operations, such as deposition operations. In some embodiments, the flow module purge gas flow rate and the slit purge gas flow rate can have a more dominant effect on the process conditions on the substrate. As used herein, "purge gas" can mean any one or combination of the flow module purge gas, slit purge gas, liner purge gas, and rotary purge gas flowing from the multiple flow module purge gas inlets 164.As used herein, "purge gas flow rate" and changes in "purge gas flow rate" such as "first purge gas flow rate" or "second purge gas flow rate" can refer to any one or combination of the flow module purge gas flow rate, the slit purge gas flow rate, the liner purge gas flow rate, and the rotational purge gas flow rate.

[0021] The process gas inlet 114 is positioned so that gas flows parallel to the top surface 150 of the substrate 102, while the flow module purge gas inlet 164 is positioned so that purge gas flows directly underneath the substrate support 106. The process gas inlet 114 is fluidly connected to a process gas source 151. The flow module purge gas inlet 164 is fluidly connected to a purge gas source 162. In some embodiments, the purge gas source 162 is fluidly connected to an inlet at a base portion of the shaft 118, as shown in phantom in FIG. 1A. The one or more exhaust gas outlets 116 are fluidly connected to an exhaust pump 157. Each of the process gas source 151 and the purge gas source 162 can be configured to deliver one or more precursors or process gases into the process volume 136. The precursor and / or process gases may include one or more of dichlorosilane (DCS), trichlorosilane (TCS), or higher order silanes such as disilane (DS), trisilane and tetrasilane, and metal-organic precursors such as trimethylaluminum (TMA) and tetraethyl orthosilicate (TEOS). However, other gases are contemplated. As shown in FIG. 1A, a controller 120 is in communication with the process chamber 100 and is used to control processes such as those described herein.

[0022] The processing chamber 100 is configured to receive the substrate 102 via a slit valve 135. The slit valve 135 may be coupled to a transfer chamber having a transfer robot therein. The slit valve 135 allows the substrate 102 to be loaded into and removed from the processing volume 136 of the processing chamber 100 via an opening (e.g., using a robot end effector of the transfer robot). A door (not shown) closes and seals the opening (not shown), thereby allowing the environment of the processing volume 136 to be controlled independently of ambient conditions outside the processing chamber 100.

[0023] Various gas inlets, such as the process gas inlets 144 and the flow module purge gas inlets 164, can be configured to provide individual or multiple gas flows with varying gas parameters, such as velocity, density, or composition. In some embodiments, as shown in FIG. 1B, the exhaust outlet(s) 116 are positioned 180 degrees from the process gas inlets 114 and the flow module purge gas inlets 164 in a cross-flow configuration. In the cross-flow configuration, the slit valve 135 can be positioned 90 degrees in a clockwise direction (as shown in FIG. 1B) or can be positioned 90 degrees in a counterclockwise direction from the exhaust outlet(s) 116. In some embodiments, the slit valve 135 can be positioned below the inlet (e.g., the slit valve 135 can be positioned 0 degrees from the inlet location). The processing chamber 100 can further include an additional gas inlet (not shown) to provide an optional cross-flow gas, which facilitates tuning of the deposition uniformity. As shown in FIG. 1B, the additional gas inlet is positioned 180 degrees from the slit valve 135. However, it is contemplated that the additional gas inlet and the slit valve 135 can be aligned with one another (e.g., located on the same side of the processing chamber 100). In some embodiments, the one or more exhaust outlets 116 are positioned 180 degrees from the slit valve 135 and 270 degrees in a counterclockwise direction from the process gas inlets 114 and the flow module purge gas inlets 164 in a rotational configuration. In this rotational configuration, when the substrate 102 is rotated in a counterclockwise direction, the gas flow sweeps at least 270 degrees around the interior of the processing chamber 100 before being exhausted through the one or more exhaust outlets 116. However, the one or more exhaust outlets 116, the additional gas inlet, the process gas inlets 114, the slit valve 135, and the flow module purge gas inlets 164 can be positioned in various optional locations relative to one another. In some examples, the outlet port 190 is at the same height as the inlet port 180 in the XY plane.

[0024] FIG. 2 is a method 200 of adjusting a purge gas flow rate according to some embodiments of the present disclosure. In operation 210, the controller 120 receives a first input. The first input can be entered by a user or selected from a library or database of information stored by the controller 120. The first input corresponds to a first geometric hardware configuration of the process chamber 100. In some embodiments, the first geometric hardware configuration includes one or more of the following: flow type, preheat ring position, substrate support position, inlet location, slit valve and outlet location, non-vented or vented liner, circular or elliptical preheat ring, overlapping substrate support, dome or liner replacement, and any other new or replaced hardware in the process chamber. The first geometric hardware configuration of the first input can take into account the life and / or current aging of one or more hardware components. For example, a newly replaced hardware component may be identical to a previous hardware component, but may have a different level of wear and initial machining tolerances, resulting in a changed required purge gas flow rate. In some examples, the first input may reference a library or database of values ​​related to chamber hardware or hardware characteristics such as dimensions.

[0025] In operation 220, the controller 120 receives a second input. The second input may be entered by a user or may be selected and referenced by the controller 120 from a library or database of information stored in memory. The second input corresponds to a first process recipe for the processing chamber. In some embodiments, the first process recipe includes at least one process condition, such as a chamber temperature, a chamber pressure, a purge gas flow rate, a process gas flow rate, a substrate support rotation speed, a substrate support vertical position, a duration, and / or a gas composition.

[0026] In operation 230, the first input and the second input are used to determine a first purge gas flow rate. In one example, the purge gas flow rate per recipe includes a "default" setting, and the purge gas flow rate, timing, and / or composition are associated with the "default" setting. The default purge gas setting can be determined by experimentation and can be stored in a memory or database and recalled when a desired hardware configuration and / or recipe is selected. In another example, the first purge gas flow rate is determined by a machine learning (ML) model 404 that selects a flow rate based on the first input and the second input from a library or catalog of prior laboratory data. For example, the ML model 404 determines the first purge gas flow rate and displays the determined value to a user, or the ML model 404 determines the first purge gas flow rate and automatically adjusts the purge gas flow rate to match the determined value.

[0027] In operation 240, a first deposition characteristic of the substrate or other chamber part is measured using one or more first sensors. In some embodiments, the first deposition characteristic may be thickness, color, mass fraction, total coverage, CE change, transmission, transparency, reflectance, roughness, and / or emissivity. The one or more first sensors may include a single or multiple color pyrometer, a broadband optical spectrometer, an optical reflectometer, a camera, a thermal camera, another device for physical sensor feedback, or any combination thereof. The one or more first sensors are outside the lower window 110 and the upper window 108 such that the one or more first sensors have a direct line of sight to a desired target area. For example, the one or more optical spectrometers and / or the one or more thermal cameras may be aimed at or through the lower window 110 at multiple locations where thicker deposition is expected, and / or the one or more pyrometers may be aimed through the lower window 110 at the backside of the substrate support 106. In one example, signals from one or more optical spectrometers can be correlated with loss of transmission or increased occlusion of portions of the lower window 110 indicating coating evolution. In situ image analysis of images from one or more cameras can similarly be correlated with coating evolution on the lower window 110. Metrology data (e.g., physical sensor feedback) from one or more first sensors can be used by the ML model 404 or other neural network for supervised or unsupervised self-learning to adjust purge gas flow rates to achieve improved and / or desired processing results, such as deposition uniformity and increased throughput. For example, purge gas flow rates can be adjusted by the ML model after comparing predicted backside coatings to measured backside coatings and correcting for overdeposition on chamber parts or for uneven / undesirable deposition on substrates.

[0028] In operation 250, the first input, the second input, and the measured deposition characteristics are used to determine a second purge gas flow rate. In some embodiments, the second purge gas flow rate is determined by a machine learning (ML) model 404 (e.g., the ML model 404 determines the first purge gas flow rate and displays the determined value to a user, or the ML model 404 determines the first purge gas flow rate and automatically adjusts the purge gas flow rate to match the determined value). The second purge gas flow rate may be different from the first purge gas flow rate, and the difference between the second purge gas flow rate and the first purge gas flow rate may be based in whole or in part on the measured deposition characteristics. As previously discussed, the first purge gas flow rate and the second purge gas flow rate may refer to a flow module purge gas, a slit purge gas, a rotational purge gas, and / or a liner purge gas.

[0029] The ML model 404 (shown in FIG. 4) is particularly advantageous in that the model is configured for multiple purge gas variables and configured to make a single correction based on these variables. In conventional systems, a change to one variable, such as a rotational purge, can have unintended consequences at other chamber locations and create new problems. Such conventional systems tend to force the operator to "chase" the configuration, in that a first change in process conditions can have unintended consequences elsewhere in the chamber, which can necessitate a second change in process conditions, which can necessitate a third change, and so on. However, the ML model 404 can learn to anticipate these changes and can make the necessary adjustments to eliminate this tracking problem. It is further contemplated that the ML model 404 can include or interact with a "digital twin" of the process chamber. The digital twin is configured according to the first input of operation 210. Fluid dynamics simulations can be performed on the digital twin based on the revisions suggested by the ML model 404 to determine the effects of fluid flow within the digital twin, and conversely, within the actual processing chamber.

[0030] The use of the digital twin eliminates "trial and error" or "chasing" on the physical processing chamber to determine improved and / or optimal process conditions, such as purge flow rates. It is contemplated that the digital twin and / or ML model 404 can be updated based on feedback from one or more sensors in the physical processing chamber, thereby improving results. Any such feedback is valuable in ensuring an accurate digital twin model. Additionally, such feedback facilitates updating the digital twin due to deviations in hardware component specifications from input values. For example, the physical dimensions of the hardware components may change due to iterative processing in the physical chamber, and thus the digital twin model may not accurately reflect the physical model. Feedback provided by sensors in the physical chamber may identify deviations from the digital twin (e.g., the digital twin purge gas recipe may indicate that no deposition is occurring at the bottom window while the physical chamber sensor still indicates deposition forming at the bottom window), and thus the ML model 404 may "learn" to account for these deviations and improve the physical chamber processing by providing updated process recipe parameters.

[0031] Optionally, operations 240-250 are repeated so that multiple measurements of the deposition characteristics can be compared. In some embodiments, a contrast curve analysis can be created to compare the measured deposition characteristics as a function of purge gas flow rate. In some embodiments, operations 240-250 are repeated one or more times during a process, such as a deposition process, and the purge gas flow rate is adjusted in real time, either by the ML model 404 or manually.

[0032] Graphical comparisons of measured deposition characteristics for different hardware configurations, process recipes and / or purge gas flow rates can be made, as shown in Figures 3A-3E. In Figures 3A-3E, the process recipe was held constant to compare deposition characteristics as a function of different hardware configurations and purge gas flow rates. As illustrated, small changes in physical components can result in large differences in deposition characteristics. However, the ML model 404 of the present disclosure can account for these changes to reduce unwanted deposition in undesired locations.

[0033] 3A-3B are graphical comparisons of measured deposition on a lower window in a processing chamber as a function of both different hardware configurations (see bars on the graph) and different purge gas flow rates (see x-axis labels). Hardware configuration A (left bar, HWC-A) includes a first preheat ring, and hardware configuration B (right bar, HWC-B) includes a second preheat ring. In some embodiments, the first preheat ring and / or the second preheat ring may be symmetric or asymmetric. As used herein, symmetric and asymmetric can mean uniformity or non-uniformity. As can be seen from FIG. 3A, hardware configuration A shows a thinner measured deposition compared to hardware configuration B for each purge gas flow rate. Thus, in the case of FIG. 3A, adjusting the purge gas flow rate can include increasing the purge gas flow rate to thin the measured deposition on the lower window of the processing chamber where deposition is undesirable. Clearly, different hardware configurations will result in more or less deposition on surface features when using the same process recipe. However, aspects of the present disclosure can take these differences into account to reduce undesirable results.

[0034] In Figures 3B-3D, hardware configurations A-G include first, second, third, fourth, and fifth preheat rings, respectively, and first, second, third, fourth, and fifth liners, respectively. In some embodiments, one or more of the first, second, third, fourth, and fifth preheat rings may be symmetric or asymmetric, and one or more of the first, second, third, fourth, and fifth liners may be vented liners. As can be seen from Figure 3B, hardware configuration B generally results in increased deposition at the bottom window for all purge gas flow rates. Meanwhile, hardware configurations C, D, and E each show a thicker relative deposition than the other two for different purge gas flow rates. Furthermore, as the purge gas flow rate increases, the measured deposition becomes thinner for all hardware configurations. Thus, in the case of FIG. 3B, the adjustment of the purge gas flow rate can include either decreasing or increasing the purge gas flow rate depending on the hardware configuration of the processing chamber utilized to thin the measured deposition on the lower window of the processing chamber where deposition is undesirable. Different hardware configurations coupled with different recipe conditions thus indicate different purge gas flow rates to avoid unacceptably thick deposition characteristics on the lower window in the processing chamber where deposition is undesirable. This particular purge gas flow rate indicated to avoid unacceptably thick deposition characteristics is the lower threshold of purge gas flow rate for the lower window. The unacceptably thick deposition characteristics depend on the desired accuracy of the thin film application process and the application of the thin film itself. For example, a reduction in the transparency of quartz of more than about 5% and / or a change in the emissivity of the substrate of more than about 0.1% can be considered unacceptable.

[0035] FIG. 3C is a graphical comparison of measured deposition on a substrate in a processing chamber as a function of both different hardware configurations (see bars on the graph) as well as different purge gas flow rates (see x-axis labels). As can be seen from FIG. 3C, the different hardware configurations each show thicker or thinner deposition than the other hardware configurations for different purge gas flow rates. Furthermore, as the purge gas flow rate increases, the measured deposition becomes thinner. Thus, in the case of FIG. 3C, adjusting the purge gas flow rate can include decreasing the purge gas flow rate to increase the measured deposition on the substrate in the processing chamber where deposition is desired. Thus, different hardware configurations coupled with different recipe conditions require different amounts of slit purge gas flow to avoid unacceptably thin deposition on the substrate in the processing chamber where deposition is desired. This particular amount of purge gas flow rate required to avoid unacceptably thin deposition is the upper threshold of acceptable purge gas flow rate. The unacceptably thin deposition characteristic depends on the desired accuracy of the thin film application process and the application of the thin film itself. For example, a reduction in deposition on the substrate of more than about 5% can be considered unacceptable.

[0036] An unacceptably thin deposition on a substrate in a processing chamber may also include substantial problems with deposition uniformity. Substantial problems with deposition uniformity on a substrate depend on the desired accuracy of the thin film application process. For example, non-uniformity of more than about 1% on a substrate may be considered unacceptable.

[0037] FIG. 3D is a graphical comparison of the measured deposition on the backside of the substrate support in the processing chamber as a function of both different hardware configurations (see bars on the graph) as well as different purge gas flow rates (see x-axis labels). As can be seen in FIG. 3D, hardware configuration B shows a higher or increased measured deposition compared to the other hardware configurations for each purge gas flow rate, with the increased purge gas flow rate generally resulting in a thinner deposition for all hardware configurations. Meanwhile, hardware configurations C, D, and E each show a thicker relative deposition than the other two for different purge gas flow rates. Thus, in the case of FIG. 3D, adjusting the purge gas flow rate can include increasing the purge gas flow rate to thin the measured deposition to avoid unacceptably thick deposition on the backside of the substrate in the processing chamber where deposition is undesirable. This particular purge gas flow rate required to avoid unacceptably thick deposition is also the lower threshold of the required purge gas flow rate for the backside of the substrate support. In the case where there are multiple lower threshold required purge gas flow rates for different chamber parts or different portions of the chamber parts, the lowest or smallest lower threshold required purge gas flow rate can be selected as the overall lower threshold purge gas flow rate for the process chamber, in other embodiments, each of the different purge gas types and sources described can each have an individually determined purge gas flow rate.

[0038] FIG. 3E is a comparison table of measured deposition characteristics of the backside of the substrate support, the substrate and the lower window in the chamber according to the graphs of FIGS. 3A-3D. The process recipe was kept constant for each hardware configuration. As can be seen from FIG. 3E, an increase in the purge gas flow rate leads to a decrease in the deposition characteristics of the substrate, the backside of the substrate support and the lower window. To shorten the processing time, increase the deposition uniformity on the substrate and increase the overall throughput of the chamber, it is advantageous to improve the deposition characteristics on the substrate. However, to reduce the cleaning time and avoidable maintenance and increase the overall throughput of the chamber, it is also advantageous to reduce the deposition characteristics of the backside of the substrate support and the lower window. This conflict affects the purge gas flow rate (e.g., increasing the purge gas flow rate will increase the deposition on the substrate as well (advantageous) and increase the deposition on other chamber parts (disadvantageous)).

[0039] For example, in FIG. 3E, the deposition characteristic value for the substrate is shown via the mass fraction ratio of DCS to the total amount of gas. In some examples, the lower threshold for the deposition characteristic for the substrate may be 0.50, and the upper threshold deposition characteristic value for the backside and window portion of the substrate support may be 0.05. This would mean that for hardware configurations B and D, the purge flow rate should be at least 18 standard liters per minute (slm) to stay below the upper threshold deposition characteristic value for the backside and lower window of the substrate support to prevent an unacceptable amount of deposition on the backside and lower window of the substrate support. Similarly, for hardware configurations B and D, the purge flow rate should not exceed 18 slm to stay above the lower threshold deposition characteristic value for the substrate and retain a sufficiently large amount of deposition on the substrate. Thus, in this example, a purge flow rate of 18 slm is an acceptable value for hardware configurations B and D, and the process recipe would be utilized with both hardware configurations.

[0040] The deposition characteristics measured by the methods described above can reveal a sensitivity to the probability of coating the process chamber parts, which is a measure of the concentration of precursor or process gas in close proximity to the process chamber parts, and can be indicated by a number of physical sensor feedbacks, such as those disclosed herein.

[0041] As discussed, the controller 120 includes an ML model 404 that is trained on historical and simulated models of the hardware configuration, process recipe, and known or measured deposition characteristics. Upon determining the deposition characteristics by metrology measurements, such as via any of the sensors described herein, the trained ML model determines the simulated model to adjust the purge gas flow rate, and optionally causes the adjusted purge gas flow rate to be implemented. It is contemplated that for any intermediate values ​​or models, the ML model 404 may use or represent regression or interpolation analysis.

[0042] In some embodiments, the ML model 404 is a component of a larger system or process for generating an adjusted purge gas flow rate. In these embodiments, some or all of the simulated model generated by the ML model 404 is utilized. In some embodiments, the modified simulated model is selected by a table lookup based on a known process recipe, a known hardware configuration, and / or known or measured deposition characteristics of the processing chamber. In these embodiments, the table lookup determines the simulated model to select from a library of simulated models configured to adjust the purge gas flow rate of the processing chamber for the known hardware configuration, process recipe, and / or deposition characteristics.

[0043] According to certain embodiments, the ML model 404 may be supervised or unsupervised machine learning. In some embodiments, the ML model 404 is a classifier, such as a neural network, a deep learning neural network, a k-means classifier, a random forest walk, etc., that can receive inputs, such as a hardware configuration, a process recipe, such as an initial process recipe, and / or metrology data (e.g., measured deposition properties) of a substrate, such as the substrate 102, being processed by a tool, such as the processing chamber 100, or other chamber parts, such as the bottom window 110 or the backside of the substrate support 106.

[0044] FIG. 4 depicts a schematic diagram of a controller according to some embodiments of the present disclosure. As shown in FIG. 4, the controller 120 receives design data 408 including hardware configuration data and generates a purge gas flow rate prediction 406. The design data 408 can include or accompany process recipe data 412 including information about process conditions of the processing chamber. For example, the design data 408 can be provided to the ML model during operation 210, and the process recipe data 412 can be provided to the ML model during operation 220. Upon classifying the provided data, the ML model 404 generates a purge gas flow rate prediction 406 that is provided to a tool 432, such as the processing chamber 100. The ML model 404 further receives metrology data (e.g., deposition characteristics) 416 generated by a metrology tool, such as one or more sensors described herein. In some embodiments, the metrology tool can include more than one metrology tool and / or can measure multiple deposition characteristics. In some embodiments, the metrology tool measures one or more deposition characteristics in real time while processing a substrate. Upon classifying the measured deposition property data, the ML model 404 generates a further purge gas flow rate prediction 406 that is provided to a tool 432, such as the processing chamber 100. In some embodiments, the ML model 404 provides the purge gas flow rate prediction 406 to the tool 432 in real time while a substrate is being processed, such as by automatically adjusting the purge gas flow rate to match the further purge gas flow rate prediction 406.

[0045] The controller 120 further includes an ML training component 420 for training the ML model 404. The ML training component 420 can receive historical data 424, simulated data 428, or both to train the ML model 404. The historical data 424 and the simulated data 428 can include hardware configuration data, process recipe data, and / or metrology data, such as deposition characteristic data. In this context, the simulated data 428 can be a simulated model of predicted deposition characteristics based on known or discovered correlations between hardware configurations, process recipe conditions, and metrology data. The historical data 424 can include historical hardware configuration and / or process recipe data from a previous processing chamber, and / or metrology data (e.g., deposition characteristic data) from a substrate and / or other chamber parts, such as a bottom window or substrate support backside, that have already been measured. In some embodiments, the ML model 404 is trained offline or at some point before processing is performed, such as prior to operations 210 and 220. In some embodiments, training may occur during processing, such as between acts 210-250.

[0046] The ML training component 420 trains the ML model 404 using historical data 424 and / or simulated data 428. Once trained, the ML model 404 can receive the data described above and generate a purge gas flow rate prediction 406 that is provided to a tool 432, such as the processing chamber 100. The ML model 404 can continually store and update records and / or information regarding different process recipes and hardware configurations that have been tested, and the data can be saved in the historical data 424. The ML training component 420 constantly increases the accuracy of the ML model 404.

[0047] FIG. 5 depicts a processing system 500 according to some embodiments of the present disclosure. The processing system 500 is an example of the controller 120 according to certain embodiments and may be used in place of the controller 120 described above. FIG. 5 depicts an exemplary processing system 500 that may operate the embodiment systems described herein to perform embodiments according to the flow charts and methods described herein. The processing system 500 includes a central processing unit (CPU) 502 coupled to a data bus 516. The CPU 502 is configured to process computer-executable instructions stored, for example, in a memory 508 or storage 510 to cause the processing system 500 to perform the method embodiments described herein, for example, on the system embodiments described herein with respect to FIGS. 1A-3E. The CPU 502 is included as representative of a single CPU, multiple CPUs, a single CPU with multiple processing cores, and other forms of processing architectures capable of executing computer-executable instructions.

[0048] Processing system 500 further includes input / output (I / O) device(s) 512 and an interface 504 that allows processing system 500 to interface with input / output devices 512, such as, for example, a keyboard, a display, a mouse device, pen input, and other devices that can interact with processing system 500. It should be noted that processing system 500 can connect to external I / O devices (e.g., an external display device) via physical and wireless connections.

[0049] The processing system 500 further includes a network 514 interface that provides the processing system with access to an external network 514, and thereby to external computing devices. The processing system 500 further includes a memory 508, which in this example includes a receiving component 518, a generating component 520, a controlling component 522, a measuring component 524, and a determining component 526 for performing the operations described herein. It should be noted that while shown in FIG. 5 as a single memory 508 for simplicity, the various aspects stored in the memory 508 may be stored in different physical memories, including memories remote from the processing system 500, but all accessible by the CPU 502 via an internal data connection, such as the bus 516.

[0050] Storage 510 further includes hardware configuration design data 528, process recipe data 530, deposition characteristic data 532, machine learning (ML) model data 534, ML training data 536, purge gas flow rate data 538, and adjusted purge gas flow rate data 540 for performing the operations described herein. As will be appreciated by those skilled in the art, storage 510 may include other data and aspects. As with memory 508, while FIG. 5 depicts a single storage 510 for simplicity, the various aspects stored in storage 510 may be stored in different physical storages, all accessible to CPU 502 via an internal data connection, such as bus 516, or an external connection, such as network interface 506. Those skilled in the art will appreciate that one or more elements of processing system 500 may be located remotely and accessed via network 514.

[0051] FIG. 6 is a schematic cross-sectional view of another processing chamber 600 that can be used in some embodiments. The processing chamber 600 can be an RTP chamber. In some embodiments, the processing chamber 600 is set up in a cross-flow configuration as shown in FIG. 6, whereby the outlet 640 is 180 degrees from the inlet 620. A plurality of lamps 650 are located above a transparent top window 645 that is coupled to an upper body 644 of the processing chamber 600. In some embodiments, the upper body is formed of an opaque material (e.g., aluminum or steel). The substrate 602 is located on an annular substrate support 604 above a chill plate 624 in the processing volume 610 of the processing chamber 600. The annular substrate support is located on an edge ring 615 that is located on a support cylinder 625. The support cylinder 625 is configured to rotate the annular substrate support 604. In some embodiments, the purge gas can flow from the floor 601 of the processing chamber 600. A plurality of reflectors 608 are coupled to the substrate side of the chill plate 624. In some embodiments, a stationary member 655 is coupled to the lower body 654 of the processing chamber 600. The stationary member 655 may be an annular member.

[0052] 7 is a schematic cross-sectional view of yet another processing chamber 700 that can be used in some embodiments. The processing chamber 700 includes a showerhead 710 located inside a processing volume 720 of the processing chamber 700. A substrate support pedestal 730 is configured to support a substrate 702 during processing. In some embodiments, an induction heater coil 740 is housed within the substrate support pedestal 730. An annular purge gas inlet 750 radially surrounds a substrate support shaft 735 at the floor 704 of the processing chamber 700. The substrate support shaft 735 is configured to support the substrate support pedestal 730. In some embodiments, the substrate support shaft 735 is configured to rotate the substrate support pedestal and / or the substrate. In some embodiments, the purge gas from the purge gas inlet 750 can be exhausted 360 degrees throughout the processing chamber 700 via an exhaust vent.

[0053] It is contemplated that the method described above can be integrated into an ML platform, whereby a first input corresponding to a first hardware configuration, a second input corresponding to a first process recipe, and one or more measured deposition characteristics can be used to adjust the purge gas flow rate into the chamber. As mentioned above, if the purge gas flow rate is too high, the total precursor gas concentration or the total process gas concentration across the substrate will be reduced due to dilution. This will in turn adversely affect (e.g., slow) the growth rate of the semiconductor material, and excess purge gas will be wasted, adversely affecting the process. Alternatively, if the purge gas flow rate is too low, the process gas or precursor gas may get into places that are difficult to clean, and may deposit undesirable layers on parts of the processing chamber other than the substrate.

[0054] Aspects of the present disclosure contemplate the use of a digital replica of a physical processing chamber, such as the processing chambers described herein. The digital replica is a digital simulation stored in memory that corresponds to a physical processing chamber. The digital replica reflects the physical characteristics, such as the physical dimensions, materials, and so on, of the actual physical chamber. The digital replica facilitates predicting the operation in the physical chamber. The operation of the physical chamber using the digital replica is simulated using one or more rules, algorithms, physics models, fluid flow models, and so on. The digital replica can predict data such as deposition characteristics in the physical chamber based on a mathematical and physical understanding of the process conditions input into the algorithms of the digital replica. For example, the digital replica can predict the deposition characteristics of the backside of the substrate (in the physical chamber) based on the values ​​of the process gas flow rates and the purge gas flow rates into the physical chamber. In some embodiments, the digital replica can create multiple distinct metrology outputs that are predicted. For example, the digital replica receives data indicative of the fluid flow rates (e.g., purge gas flow rates and process gas flow rates) of the physical chamber and mathematically calculates more than a single solution or outputs metrology data. The digital replica processing chamber can automatically reject predicted output metrology data that is outside of acceptable ranges or that otherwise violates the rules of the digital replica software. In some embodiments, the digital replica processing chamber works in conjunction with a machine learning model. This configuration is known as a Physics-Informed Machine Learning Model (PIMLM) or Physics-Informed Neural Network (PINN).

[0055] The machine learning model can be utilized to update algorithms, rules, or other results output by the digital replica to refine and / or increase the accuracy of the digital replica simulation. Further, the machine learning model can additionally or alternatively recommend hardware configurations and / or process conditions to achieve desired results in the physical chamber. The machine learning model can work in conjunction with the digital replica to provide improved processing in the physical chamber. For example, the machine learning model can work in conjunction with the digital replica to provide outputs to achieve desired process conditions (e.g., deposition characteristics) when hardware configurations, process recipes, etc. are changed. Other advantageous effects are also contemplated.

[0056] The above description is provided to enable all skilled in the art to practice the various embodiments described herein. The examples discussed herein are not intended to limit the scope, applicability, or embodiments set forth in the claims. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments. For example, changes may be made in the function and arrangement of the elements discussed without departing from the scope of the disclosure. The various examples may omit, substitute, or add various procedures or components, where appropriate. For example, the methods described may be performed in a different order than that described, and various steps may be added, omitted, or combined. Also, features described in connection with some examples may be combined in some other examples. For example, an apparatus may be implemented, or a method may be practiced using any number of aspects set forth herein. Additionally, the scope of the disclosure is intended to encompass such an apparatus or method implemented using other structure, functionality, or structure and functionality in addition to or other than the various aspects of the disclosure set forth herein. It should be understood that any aspect of the disclosure disclosed herein may be embodied by one or more elements of a claim.

[0057] The methods disclosed herein include one or more operations or actions for achieving the method. Method operations and / or actions may be interchanged with one another without departing from the scope of the claims. In other words, unless a specific order of operations or actions is specified, the order and / or use of specific operations and / or actions may be modified without departing from the scope of the claims. Furthermore, various operations of the methods described above may be performed by any suitable means capable of performing the corresponding functions. The means may include various hardware and / or software component(s) and / or modules including, but not limited to, circuits, application specific integrated circuits (ASICs) or processors. Typically, where there are operations illustrated in a figure, these operations may have corresponding counterpart means-plus-function components numbered similarly.

[0058] While the forgoing is directed to embodiments of the present disclosure, other and further embodiments of the disclosure may be devised without departing from the basic scope of the disclosure, the scope of which is determined by the appended claims.

Claims

1. 1. A method for processing a semiconductor substrate, comprising: receiving a first input corresponding to a first hardware configuration of a processing chamber; receiving a second input corresponding to a first process recipe for the processing chamber; performing a first deposition process using the first process recipe at a first purge gas flow rate in the processing chamber; Measuring a deposition characteristic of the first deposition process via a first sensor; determining a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic; performing a second deposition process at the second purge gas flow rate; A method comprising:

2. The method of claim 1 , wherein the first sensor comprises a pyrometer.

3. The method of claim 1 , wherein the first sensor comprises a light spectrometer or a camera.

4. The method of claim 3 , further comprising one or more second sensors in the processing chamber.

5. The method of claim 1 , wherein the first hardware configuration includes a non-vented or vented liner, a circular or elliptical preheat ring, and / or a substrate support overlapping the preheat ring.

6. 10. The method of claim 1, wherein the determining the second purge gas flow rate is accomplished by a machine learning model, the machine learning model comparing the measured deposition characteristic to a desired deposition characteristic and providing the second purge gas flow rate based on a difference between the measured deposition characteristic and the desired deposition characteristic.

7. The method of claim 6 , wherein said determining said first purge gas flow rate is accomplished by referencing a database based on user input.

8. The method of claim 6 , wherein the first process recipe is updated to include the second purge gas flow rate.

9. The method of claim 6 , wherein the second purge gas flow rate or a difference between the first purge gas flow rate and the second purge gas flow rate is displayed for an operator.

10. The method of claim 1 , wherein the measured deposition property comprises thickness, mass fraction, transmittance, transparency, emissivity, or any combination thereof.

11. The method of claim 1 , wherein the measured deposition property is a measured deposition property of a substrate, a backside of a substrate support, or a bottom window in the processing chamber.

12. The method of claim 1 , wherein measuring the deposition characteristic of the processing chamber via the first sensor occurs more than once throughout the deposition process.

13. 10. The method of claim 1, wherein the purge gas is a flow module purge gas, a slit purge gas, a rotary purge gas, or a liner purge gas.

14. receiving a first input corresponding to a first hardware configuration of a processing chamber; receiving a second input corresponding to a first process recipe for the processing chamber; performing a first deposition process using the first process recipe at a first purge gas flow rate in the processing chamber; Measuring a deposition characteristic of the first deposition process at a first location via a first sensor; determining a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic; performing a second deposition process in the processing chamber using the second purge gas flow rate; measuring the deposition characteristic of the second deposition process at the first location via the first sensor; determining a third purge gas flow rate different from the second purge gas flow rate based on the change in the measured deposition characteristic; A method comprising:

15. The method of claim 14 , wherein the measured deposition property comprises a film thickness.

16. The method of claim 14 , wherein the measured deposition property is a measured deposition property of a substrate, a backside of a substrate support, or a bottom window in the processing chamber.

17. A non-transitory computer-readable medium having instructions that, when executed by a processor, cause the processor to: receiving a first input corresponding to a first hardware configuration of a processing chamber; receiving a second input corresponding to a first process recipe for the processing chamber; performing a first deposition process using the first process recipe at a first purge gas flow rate in the processing chamber; Measuring a deposition characteristic of the first deposition process via a first sensor; determining a second purge gas flow rate different from the first purge gas flow rate based on the first input, the second input, and the measured deposition characteristic; performing a second deposition process at the second purge gas flow rate; A non-transitory computer-readable medium for causing

18. 20. The method of claim 17, wherein the precursor gas comprises at least one of dichlorosilane (DCS) and trichlorosilane (TCS).

19. The method of claim 17 , wherein the precursor gas comprises disilane (DS).

20. The method of claim 17 , wherein the purge gas is a liner purge gas.

Images