Concentration sensor for precursor delivery system

Abstract

The concentration sensor assembly may include a vaporization chamber having a compound. The concentration sensor assembly may include a first flow path coupled to the vaporization chamber. The first flow path may direct a first gas into the vaporization chamber. The second flow path may direct a second gas out of the vaporization chamber. The second gas may include the compound and the first gas. A first sensor is disposed along the first flow path. The first sensor measures first data indicative of a first mass flow rate of the first gas. A second sensor is disposed along the second flow path. The second sensor measures second data indicative of a second mass flow rate of the second gas. The computing device may determine a concentration of the vaporized substance in the second gas based on the first data and the second data.

Description

[Technical Field]

[0001] FIELD OF THE INVENTION This disclosure relates generally to gas delivery to a processing chamber, and more particularly to a sensor for measuring the concentration of a precursor in a process gas. [Background technology]

[0002] Integrated circuits have evolved into complex devices containing millions of transistors, capacitors, and resistors on a single chip. Advances in chip design have continually required faster circuit speeds and greater circuit density, placing increasing demands on precision manufacturing processes. Precise processing of substrates requires precise control of temperature, flow rate, and pressure in the supply of fluids used during processing.

[0003] Chemical vapor deposition (CVD) and atomic layer deposition (ALD) are gas-phase deposition processes used to form or deposit a variety of materials on a substrate. Generally, CVD and ALD processes involve the delivery of gaseous reactants to a substrate surface where a chemical reaction occurs under temperature and pressure conditions favorable to the thermodynamics of the reaction. The type and composition of layers that can be formed using CVD or ALD processes are limited by the ability to deliver the chemical reactants or precursors to the substrate surface. A variety of solid and / or liquid precursors have been successfully used during CVD and ALD applications by delivering the precursors in a carrier gas.

[0004] In some cases, the carrier gas is passed through a heated vessel or canister, such as an ampoule or bubbler, containing a volatile liquid precursor under conditions that vaporize the precursor. In other cases, the carrier gas is passed through a heated vessel containing a solid precursor under conditions that sublimate the solid precursor. Sublimation processes are typically carried out in a vessel charged or filled with the solid precursor, and the vessel walls are heated to sublimate the solid precursor material while simultaneously producing the gaseous precursor. In either case, the carrier gas combines with the vaporized precursor to form a process gas, which is drawn from the vessel into the reaction chamber through a dedicated conduit or gas line. Summary of the Invention

[0005] In some embodiments, the concentration sensor assembly includes a vaporization chamber having a compound. The vaporization chamber transitions the compound to a gas (e.g., via sublimation or vaporization). The concentration sensor assembly may further include a first flow path connected to the vaporization chamber. The first flow path can direct a first gas to the vaporization chamber. A second flow path is coupled to the vaporization chamber. The second flow path directs a second gas out of the vaporization chamber. The second gas includes a compound gas and the first gas. A first sensor is disposed along the first flow path. The first sensor measures first data indicative of a first mass flow rate of the first gas in the first flow path. A second sensor is disposed along the second flow path. The second sensor measures second data indicative of a second mass flow rate of the second gas in the second flow path. The concentration sensor may further include a computing device coupled to the first sensor and the second sensor. The computing device determines the concentration of the compound in the second gas based on the first data and the second data.

[0006] In some embodiments, the precursor concentration delivery system includes a vaporization vessel. The vaporization vessel contains a precursor. A first flow path is connected to the vaporization vessel. The first flow path conducts a carrier gas to the vaporization vessel. A second flow path is connected to the vaporization vessel. The second flow path conducts a process gas out of the vaporization vessel. The process gas contains a carrier gas and a precursor. A process chamber is connected to the second flow path. The second flow path conducts the process gas to the process chamber. A first flow meter is disposed along the first flow path. The first flow meter measures first data indicative of a first flow rate of the carrier gas in the first flow path. A second flow meter is disposed along the second flow path. The second flow meter measures second data indicative of a second flow rate of the process gas in the second flow path. A computing device is coupled to the first flow meter and the second flow meter. The computing device determines a concentration of the precursor in the process gas based on the first data and the second data.

[0007] In some embodiments, the method includes receiving, by a processing device, first data from a first sensor indicating a first mass flow rate of a carrier gas. The processing device receives, from a second sensor, second data indicating a second mass flow rate of a compound gas including the carrier gas and a vaporized substance. The processing device receives, from a third sensor, third data indicating a temperature of a vaporization vessel associated with vaporization of the vaporized substance. The processing device determines a concentration of the vaporized substance in the compound gas based on the first data, the second data, and the third data. The method may include modifying the flow rate of the carrier gas. The method may include providing the concentration for display by a graphical user interface (GUI). [Brief explanation of the drawings]

[0008] [Figure 1A-1B] 1 illustrates an exemplary embodiment of a concentration sensor assembly, according to some embodiments. [Figure 2] FIG. 1 illustrates a gas supply system, according to some embodiments. [Figure 3] FIG. 1 illustrates a gas supply system, according to some embodiments. [Figure 4] FIG. 1 illustrates a precursor delivery system, according to some embodiments. [Figure 5] 1 is a flowchart of a method for determining the concentration of a precursor, according to some embodiments. [Figure 6] FIG. 1 illustrates a model training workflow and a model application workflow for a concentration sensor, according to one embodiment of the present disclosure. [Figure 7] 1 is a flow chart illustrating an embodiment of a method for training a machine learning model to determine the concentration of a volatile substance disposed in a process gas, in accordance with an embodiment of the present disclosure. [Figure 8] 1 shows a flow diagram of an example method for determining the concentration of a volatile substance using a trained machine learning model, according to some implementations of the present disclosure. [Figure 9]1 illustrates a block diagram of an exemplary computing device capable of calculating the concentration of a gas, operating in accordance with one or more aspects of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0009] As previously mentioned, chemical vapor deposition (CVD) and atomic layer deposition (ALD) are gas-phase deposition processes used to form or deposit various materials on a substrate. Generally, CVD and ALD processes involve the delivery of gaseous reactants to a substrate surface where a chemical reaction occurs under temperature and pressure conditions favorable to the thermodynamics of the reaction. The type and composition of layers that can be formed using CVD or ALD processes are limited by the ability to deliver the chemical reactants or precursors to the substrate surface. A variety of solid and / or liquid precursors have been successfully used in CVD and ALD applications by delivering the precursors in a carrier gas.

[0010] In some cases, the carrier gas is passed through a heated vessel or canister, such as an ampoule or bubbler, containing a volatile liquid precursor under conditions that vaporize the precursor. In other cases, the carrier gas is passed through a heated vessel containing a solid precursor under conditions that sublimate the solid precursor. Sublimation processes are typically carried out in a vessel charged or filled with the solid precursor, and the vessel walls are heated to sublimate the solid precursor material while simultaneously producing the gaseous precursor. In either case, the carrier gas combines with the vaporized precursor to form a process gas, which is drawn from the vessel into the reaction chamber through a dedicated conduit or gas line.

[0011] Traditionally, vapor deposition processes utilizing solid precursors can encounter several challenges. Solid precursors require sufficient heat to sublimate into a gaseous state, but excessive heat can cause the solid precursor to decompose. Metalorganic solid precursors are typically very expensive and are particularly susceptible to thermal decomposition, so they generally must be maintained within narrow temperature and pressure ranges during the sublimation process. If the solid precursor decomposes, it can contaminate the remaining precursor in the vessel, the delivery system of conduits and valves, and / or the process chamber, as well as the substrate. Furthermore, overheating the solid precursor can result in excessively high precursor concentrations in the process gas, resulting in wasted unused precursor or condensation of the precursor in the delivery lines or on the substrate.

[0012] Alternatively, or in addition, a solid precursor may not sublime if it is exposed to too little heat. When a carrier gas flows through the container and impacts the solid precursor, particulates from the solid precursor may be entrained in the carrier gas and transported to the processing chamber. These solid or liquid particulates can become a source of contamination for the delivery system, processing chamber, and / or substrate. The problem of particulate contamination has been addressed in the art by including a liquid carrier material mixed with the solid precursor. However, a mixture of the liquid carrier material and the solid precursor may not be effective outside of a limited temperature and pressure range, as the liquid carrier material may evaporate and become a contaminant in the delivery system, processing chamber, and / or on the substrate.

[0013] Traditionally, precursors (and other liquids) induce chemical reactions by introducing various types of reactive gases into processing (or reactive) chambers during semiconductor manufacturing. There is a growing trend toward using liquid precursors instead of gases in CVD processes. The popularity of liquid precursors is based in part on their physical properties: they are less hazardous, flammable, corrosive, and toxic than their gaseous counterparts. For example, one of the more common liquids used in semiconductor device fabrication is tetraethyl orthosilicate (TEOS), which is often used in place of silane. Using TEOS, conformal silicon dioxide (SiO2) films without detectable defects can be deposited with better step coverage and significantly reduced hazards than when using silane. In metalorganic CVD (MOCVD) processes, liquid precursors for metals such as copper are often used because gaseous precursors are unavailable.

[0014] Liquid precursors (and other process liquids) are initially liquid and are converted to a gaseous state in order to deliver the precursor as a gas through a delivery line. Disturbances in the manufacturing operation or delivery line environment, particularly those affecting temperature and pressure, can cause some of the gas to condense within the delivery line. Such condensation can absorb and carry particles into the processing chamber, depositing particles on the substrate and potentially causing particle defects in the substrate and the resulting fabricated device.

[0015] A method for detecting and recognizing potential problems in a precursor delivery system by determining the amount of precursor or the concentration of precursor used in a process gas is disclosed. Many of the above-referenced problems can be improved or otherwise mitigated by high-resolution control of the flow rate in the precursor delivery system. Efficient and cost-effective concentration sensors that function under robust measurement conditions are needed. For example, some conventional concentration sensors or concentration monitors may include expensive, specialized equipment that cannot measure the concentrations of various precursors flowing with various carrier gases. Conventional concentration sensors often require expensive equipment, such as optical sensors, that are pre-calibrated and depend on specific equipment specifications and / or configurations for various carrier gases and / or precursors.

[0016] Aspects and implementations of the present disclosure address these and other shortcomings of existing technologies by providing assemblies, systems, and / or methods for determining the concentration of a compound (e.g., precursor) within a compound gas (e.g., process gas). In some embodiments, the relationship between pressure, mass flow rate, and vaporization rate is utilized to determine the mass flow rate of the gas before and after the compound is integrated through vaporization. For example, a mass flow meter can be calibrated and configured to measure the flow rate of the compound (e.g., precursor) before and after vaporization to determine the amount (e.g., concentration) of the compound flowing through a point in a gas delivery system. In another example, the change in density (i.e., the product of vapor pressure and molecular weight) is utilized in combination with the inverse relationship between vapor pressure and molecular weight to determine the concentration of the vaporized compound, recognizing that the molecular weight of the carrier is likely unchanged.

[0017] In an exemplary embodiment, the concentration sensor assembly includes a vaporization chamber containing a compound. The vaporization chamber vaporizes the compound. The concentration sensor assembly may further include a first flow path connected to the vaporization chamber. The first flow path can direct a first gas (e.g., a carrier gas) to the vaporization chamber. A second flow path is coupled to the vaporization chamber. The second flow path directs a second gas (e.g., a carrier gas and a precursor gas) out of the vaporization chamber. The second gas includes a compound gas and the first gas. A first sensor is disposed along the first flow path. The first sensor measures first data indicative of a first mass flow rate of the first gas in the first flow path. A second sensor is disposed along the second flow path. The second sensor measures second data indicative of a second mass flow rate of the second gas in the second flow path. The concentration sensor may further include a computing device coupled to the first sensor and the second sensor. The computing device determines the concentration of the compound in the second gas based on the first data and the second data. The computing device may be, for example, a programmable logic controller (PLC), a system on a chip (SoC), a computer, a field programmable gate array (FPGA), or other type of computing device.

[0018] In an exemplary embodiment, the precursor concentration delivery system includes a vaporization vessel. The vaporization vessel contains a precursor. A first flow path is connected to the vaporization vessel. The first flow path conducts a carrier gas to the vaporization vessel. A second flow path is connected to the vaporization vessel. The second flow path conducts a process gas from the vaporization vessel. The process gas contains a carrier gas and a precursor. A process chamber is connected to the second flow path. The second flow path conducts the process gas to the process chamber. A first flow meter is disposed along the first flow path. The first flow meter measures first data indicative of a first flow rate of the carrier gas in the first flow path. A second flow meter is disposed along the second flow path. The second flow meter measures second data indicative of a second flow rate of the process gas in the second flow path. A computing device is coupled to the first flow meter and the second flow meter. The computing device determines a concentration of the precursor in the process gas based on the first data and the second data.

[0019] In an exemplary embodiment, a method includes receiving, by a processing or computing device, first data from a first sensor indicating a first mass flow rate of a carrier gas. The processing device receives second data from a second sensor indicating a second mass flow rate of a compound gas including the carrier gas and a vaporized substance. The processing device receives third data from a third sensor indicating a temperature of a vaporization vessel associated with vaporization of the vaporized substance. The processing device determines a concentration of the vaporized substance in the compound gas based on the first data, the second data, and the third data. The method may include modifying the flow rate of the carrier gas. The method may include providing the concentration for display via a graphical user interface (GUI). For example, the computing device may include a display capable of outputting the determined concentration.

[0020] Aspects of the present disclosure provide various technical advantages and improvements over conventional systems. As previously outlined, methods for measuring concentration can be difficult, unreliable, and / or generally inefficient. In some embodiments, a concentration sensor assembly can measure the concentration of a compound (e.g., precursor) within a gas (e.g., carrier gas) to which the compound and the gas may be modified. For example, a carrier gas can be selectively used for specific purposes (e.g., reducing mass flow, improving chemical interaction on its surface, enhancing capability with precursors, etc.) without requiring physical equipment reconfiguration. The sensor is calibrated with current data for the carrier gas and precursor used, and concentration measurements are performed.

[0021] 1A-1B illustrate exemplary embodiments of concentration sensor assemblies 100A-100B according to some embodiments. FIG. 1A illustrates the concentration sensor assembly 100A operating in a bypass mode or a calibration mode. FIG. 1B illustrates the concentration sensor assembly 100B operating in a gas supply mode. As shown in FIGS. 1A-1B, the concentration sensor assemblies 100A-100B include a first flow path 102. The first flow path 102 directs a first gas to or through a vaporization chamber 108 (e.g., an ampoule, a container, etc.), which may be coupled to the first flow path via a chamber leg 114A. The vaporization chamber 108 can contain a compound (e.g., a solid and / or liquid precursor) that can be transitioned to a gaseous state (e.g., a vapor) (e.g., by evaporation, sublimation, etc.). A second flow path 112 can be coupled to the vaporization chamber via a chamber leg 114B. The second flow path 112 can direct a second gas out of the vaporization chamber. The second gas may include a compound of the first gas and the first gas in a gaseous state.

[0022] As shown in FIGS. 1A-1B , the first sensor 104 may be disposed along the first flow path 102. The first sensor 104 may measure first data indicative of a first mass flow rate of the first gas in the first flow path. For example, the sensor 104 may include a mass flow meter, a mass flow controller (MFC), a volumetric flow meter, a pressure sensor, a temperature sensor, a mass sensor, and / or other sensors for measuring the condition of the first gas in the first flow path 102. In some embodiments, the sensor 104 directly measures the mass flow rate. In other embodiments, the sensor 104 measures one or more of the volumetric flow rate, molecular weight, and / or pressure of the first gas in the first flow path 102. The data measured by the sensor 104 may be communicated to a computing device (not shown) for data processing.

[0023] The second sensor 110 may be disposed along the second flow path 112. The second sensor 110 may measure second data indicative of a second mass flow rate of the second gas (e.g., a process gas or a compound gas) exiting the vaporization chamber 108. For example, the sensor 110 may include a mass flow meter, an MFC, a volumetric flow meter, a pressure sensor, a temperature sensor, a mass sensor, and / or other sensors for measuring the condition of the second gas in the second flow path 112. In some embodiments, the sensor 110 directly measures the mass flow rate. In other embodiments, the sensor 110 measures one or more of the volumetric flow rate, molecular weight, and / or pressure of the first gas in the second flow path 112. The data measured by the second sensor 110 may be communicated to a computing device (not shown) for data processing.

[0024] 1A, the concentration sensor assembly 100A includes a bypass flow path 106. The bypass flow path 106 may be used by a carrier gas to bypass the vaporization chamber 108 and avoid condensation of the vaporized compound. The bypass flow path 106 may be utilized to calibrate the second sensor 110. For example, the second sensor 110 may be calibrated for linearity and / or zero offset, such as to improve the accuracy of the mass flow measurement and / or overall concentration determination.

[0025] As described, the data measured by the first sensor 104 and the second sensor 110 may be received and processed by a computing device. The following gas properties may be utilized to measure mass flow and calculate concentration: TIFF0007796779000001.tif11170

[0026] In some embodiments, the computing device determines the change in density (e.g., the product of vapor pressure and molecular weight) between the measurements of the first sensor 104 and the second sensor 110. For example, a gas with a molecular weight of 28 grams per mole (g / mol) may pass through the first flow path. This may be measured by the sensor 104 determining that the first gas is flowing at 250 standard cubic centimeters per minute (sccm) at a pressure of 100 torr. After passing through the vaporization chamber 108 to remove the compound, a mixture of the first gas (e.g., carrier gas) and the compound (e.g., precursor) flows through the second flow path. The second sensor may measure an average molecular weight of 30.89 g / mol, which is approximately 10% greater than the carrier gas alone. The average molecular weight may be determined by measuring a flow rate of approximately 275 sccm and a pressure of 100 torr. The rate at which the carrier gas increases due to the vaporized compound indicates the concentration of the compound in the compound gas.

[0027] It should be noted that the concentration calculation method is species-independent. Conventional concentration sensors, such as optical sensors, are recalibrated and manufactured to measure a given species specification window, which is limited by density, molecular weight, flow rate, etc. In contrast, embodiments enable species-independent calculations that can work for a variety of gases (e.g., carrier gases and process gases) and a variety of compounds (e.g., precursors).

[0028] In some embodiments, the first sensor 114 further includes a mass flow controller for controlling the flow of the first gas in the first flow path 102. The computing device can send instructions to one or more flow controllers to modify the flow rate of the first gas in the first flow path 203 and / or the second gas in the second flow path 112 based on the determination of the compound concentration.

[0029] In some embodiments, the concentration sensor assemblies 100A-100B may include a first valve disposed along the first flow path. The first valve may be integrated with or otherwise coupled to the first sensor 104. The first valve may be selectively opened and closed to vary a first flow rate of the first gas. For example, the first valve may be controlled by a computing device (e.g., based on a determination of the compound concentration). In some embodiments, the concentration sensor assemblies 100A-100B may include a second valve disposed along the second flow path 112. The second valve may be integrated with or otherwise coupled to the second sensor 110. The second valve may be selectively opened and closed to vary a second flow rate of the second gas. For example, the second valve may be controlled by a computing device (e.g., based on a determination of the compound concentration).

[0030] In some embodiments, the compound may comprise a precursor for processing a substrate, as described broadly in connection with other figures. Additionally or alternatively, as further described in other embodiments, the concentration sensor assemblies 100A-100B may include a third flow path coupled to the vaporization chamber 108. The third flow path may direct the compound into the vaporization chamber 108. A third sensor may be disposed along the third flow path. The third sensor may measure third data indicative of a third mass flow rate (e.g., similar to sensors 104 and / or 110). The computing device may further calculate the concentration of the compound in the second gas based on the third data. The computing device may further determine a rate of depletion of the compound in the vaporization chamber 108 based on the determined concentration in the second gas.

[0031] In some embodiments, as further described in other embodiments, concentration sensor assembly 100A-100B may include a sensor for measuring data indicative of a condition of the vaporization chamber. For example, the sensor may measure the temperature, pressure, etc. within the chamber. The computing device may further calculate the concentration of the compound based on the condition of the vaporization chamber.

[0032] 2 illustrates a gas delivery system 200 in accordance with some embodiments. As shown in FIG. 2, the gas delivery system 200 includes a processing chamber 202. The processing chamber 202 may be any suitable semiconductor processing chamber, such as a chemical vapor deposition (CVD) chamber, an atomic layer deposition (ALD) chamber, or an etch chamber. In some embodiments, the etch chamber may be suitable for performing tantalum nitride (TaN) ALD.

[0033] The gas delivery system 200 transports the precursor 206 from the container 204 to the processing chamber 202 via a process gas 208. Typically, the precursor 206 changes state from a solid or liquid to a gas (or vapor) within the container 204 via a sublimation or vaporization process. The sublimation or vaporization process can be initiated by any suitable known technique. For example, the precursor 206 can be heated to a predetermined temperature or mixed with a foaming liquid within the container 204. In some embodiments, the temperature of the container 204 can be controlled to regulate the sublimation and / or vaporization process.

[0034] The first flow path may be coupled to a container 204 (e.g., a vaporization container). The first flow path may direct a carrier gas into the container 204. The carrier gas 210 flows into the container 204 through one or more of a valve 212, a flow meter 216, and / or a second valve 214 and mixes with the vaporized precursor 206. The second flow path may be coupled to the container 204. The second flow path may direct a process gas 208 out of the container 204. The second flow path may direct the process gas 208 to the processing chamber 202. For example, the process gas 208 may flow out of the container 204 and transport the vaporized precursor 206 to the processing chamber 202 via one or more of a valve 217, a flow meter 220, and / or a valve 218. In some embodiments, the gas delivery system 200 may include a bypass line, for example, from the valve 214 to the valve 217, or an equivalent flow path that bypasses the container. The bypass line may allow the carrier gas 210 to flow directly into the processing chamber, for example, to calibrate the flow meter 220 or to purge the processing chamber 202 .

[0035] The precursor 206 material may be selected based on the particular process being performed in the process chamber 202. For example, the precursor 206 may be an organometallic material such as tungsten carbonyl (W(CO)6) for depositing a metal film (W) on a wafer. As another example, to form a film containing tantalum, the precursor 206 may be pentadimethylaminotantalum (PDMAT). As another example, the precursor 206 may be a precursor for depositing a layer of a dielectric material on a wafer, or xenon difluoride (XeF2), for example, to supply fluoride to an etch chamber. The carrier gas 210 is typically selected based on the precursor 206. For example, if the precursor 206 is tungsten carbonyl, argon may be selected as the carrier gas 210. The carrier gas 210 may be an inert gas, such as argon or helium, and may be reactive or non-reactive with the precursor 206.

[0036] For ease of understanding, gas delivery system 200 is shown as supplying only one gas to processing chamber 202. However, gas delivery system 200 can supply additional gases (i.e., deliver additional precursors) to processing chamber 202, and multiple gas delivery systems are contemplated. Those skilled in the art will also understand that gas delivery system 200 can include additional components not shown, such as bypass valves, purge valves, flow controllers, and / or temperature controllers.

[0037] The container 204 may be, for example, any suitable container that can withstand the pressures and temperatures used to sublimate and / or vaporize the precursor 206. In some embodiments, the container may include a bubbler. As previously discussed, in conventional processing systems, it may be difficult to determine the amount (e.g., concentration) of precursor 206 disposed in the process gas 208 and delivered to the process chamber 202. Additionally or alternatively, in conventional systems, it may be difficult to determine the amount of precursor 206 remaining in the container 204.

[0038] 2, the gas supply system 200 includes a flow meter 216 disposed along the first flow path. The flow meter 216 measures first data indicative of a first flow rate (e.g., mass flow rate or volumetric flow rate) of the carrier gas 210 (e.g., in the first flow path). The gas supply system 200 includes a flow meter 220 disposed along the second flow path. The flow meter 220 measures a flow rate of the process gas 208 (e.g., in the second flow path). One of the flow meter 216 and / or the flow meter 220 can calculate the mass flow rate and / or volumetric flow rate of the carrier gas 210 and the process gas 208, respectively. For example, density, pressure, and volume measurements can be performed to determine the mass flow rate and / or volumetric flow rate. As used herein, mass flow rate refers to the mass of gas within a predetermined volume (e.g., through the area of ​​a flow meter), volumetric flow rate refers to the volume of gas (such as a carrier gas or process gas) flowing within a defined area, and density of a material or substance refers to the mass of a material or substance within a given volume of the gas delivery system 200. Data obtained from the flow meters 216, 220 may be transmitted to a computing device 222. In some embodiments, the flow meters 216, 220 may include mass flow controllers (MFCs) or operate in a similar manner. For example, the flow meters 216, 220 may control the flow of carrier gas and / or process gas within their respective flow paths.

[0039] 2, the container 204 may include a sensor 223. The sensor 223 may perform measurements of the container 204, such as pressure, temperature, density, etc. The measurements may be related to the sublimation and / or vaporization process. For example, the sensor may measure the current temperature of the container 204. Data obtained by the sensor 223 may be transmitted to the computing device 222.

[0040] 2, the gas delivery system 200 may include a computing device 222. The computing device may receive data from one or more of the flow meters 216, 220 and / or the sensor 223. The computing device 222 may be coupled to the flow meters 216, 220, the sensor 223, and / or the process chamber 202. In some embodiments, the computing device 222 may be coupled to one or more valves 212, 214, 217, 218 of the gas delivery system. The computing device 222 may include a rack-mounted server, a router computer, a personal computer, a mainframe computer, a laptop computer, a tablet computer, a desktop computer, an FPGA, an SoC, a PLC, etc. The computing device 222 may include computing hardware, firmware, and / or software (e.g., one or more features described in connection with FIG. 6).

[0041] The computing device 222 may include a concentration tool 224, a system control module 226, and / or a precursor monitor 228. The concentration tool 224 receives first data indicative of a first flow rate of the carrier gas 210 and second data indicative of a second flow rate of the process gas 208. The concentration tool can calculate the concentration of the precursor 206 in the process gas 208 based on these measurements. As previously described, the relationship between flow rate, density, and vaporization can be utilized to determine the relative concentration of the precursor in the process gas 208. As further described in other embodiments, the computing device can receive data indicative of the flow rates of additional flow paths as needed, and this data can be aggregated with the first and second data to perform updated concentration calculations and further determinations. In some embodiments, as further discussed, sensor data indicative of the condition of the vessel 204 can also be received and incorporated into the precursor 206 concentration determination.

[0042] The system control module 226 monitors and controls system methods such as controlling the flow rate of a gas (e.g., a carrier gas or a process gas), adjusting vessel conditions (e.g., changing temperature, pressure, etc.), and / or running a process chamber procedure (e.g., a CVD process or an ALD process). The precursor 206 concentration determination (e.g., by concentration tool 224) may be incorporated into decisions made by the system control module 226. For example, if the precursor concentration is higher than a threshold level, the system control module 226 can increase the flow rate of the carrier gas (e.g., by opening valve 212 or 214 to increase the flow rate) or decrease the flow rate of the precursor into the vessel (e.g., by partially closing valves 230, 232).

[0043] In some embodiments, the system control module 226 can determine whether the precursor concentration meets a threshold criterion and alter processing operations associated with the processing chamber. For example, processing operations in the chamber may be stopped until the precursor concentration no longer meets the threshold condition. In another example, a processing step may be performed under different parameters (e.g., higher temperature, longer etch duration, deposition duration, etc.) based on the determined precursor concentration.

[0044] The precursor monitor 228 may receive precursor concentration calculations (e.g., from the concentration tool 224) and make precursor-related determinations. In some embodiments, the precursor monitor 228 may determine a precursor 206 depletion rate in the vessel 204 based on the precursor 206 concentration in the process gas 208. For example, the flow rate and concentration may be used to determine the precursor 206 depletion rate. In another example, a flow rate indicating a precursor input rate, such as the flow of precursor 206 from a precursor reservoir into the vessel 204, may be considered and contribute to the calculated precursor 206 depletion rate.

[0045] In some embodiments, as previously described, precursor 206 comprises a precursor for processing a substrate in processing chamber 202. For example, precursor 206 may be used as part of a CVD process or an ALD process.

[0046] In some embodiments, first valves 212, 214 are disposed along the first flow path and second valves 217, 218 are disposed along the second flow path. The first valves 212, 214 can be selectively opened and / or closed to vary a first flow rate of the carrier gas 210. The second valves 217, 218 can be selectively opened and / or closed to vary a second flow rate of the process gas 208.

[0047] In some embodiments, the gas delivery system 200 can include a third flow path (indicated by dashed features). The third flow path can include a path including valves 230, 232, a flow meter 234, and a push gas 236. The third flow path can direct the precursor 206 into the vessel 204 via the push gas 236. The flow meter 234 measures third data indicative of a third flow rate of the precursor in the third flow path. The computing device 222 can further use the third data to calculate the concentration of the precursor 206. For example, the velocities of the precursor and the push gas can account for a portion of the mass flow rate measured by the flow meter 220.

[0048] In another embodiment, a third flow path can direct a second carrier gas into the vessel 204. The flow meter 234 can measure third data indicative of a third flow rate of the second carrier gas in the third flow path. The computing device 222 can further use the third data (e.g., the flow rate of the second carrier gas) to calculate the concentration of the precursor. For example, the flow rate of the second carrier gas can account for a portion of the mass flow rate measured by the flow meter 220.

[0049] In some embodiments, as described above, the vessel 204 may include a sensor 223 for measuring a condition of the vessel 204. For example, the sensor 223 may be a temperature sensor capable of measuring data indicative of the temperature of the vessel 204. The computing device 222 may use the third data to calculate the concentration of the precursor (e.g., using the concentration tool 224).

[0050] 3 illustrates a gas delivery system 300 according to some embodiments. The gas delivery system 300 may be a further embodiment of the gas delivery system 200 described and illustrated in connection with FIG. 2. The gas delivery system 300 may include multiple concentration sensor assemblies that cooperate to determine sub-concentrations and overall concentrations of one or more precursors within the gas delivery system 300.

[0051] As shown in FIG. 3, the first branch may include a first carrier gas source 310A, a flow meter 316A, a vessel 304A, a precursor 306A, and a flow meter 320A. The function and / or location of each of these elements may include features and descriptions described in connection with other figures (see FIG. 3). The flow meter 316A measures data indicative of the mass flow rate of the first carrier gas. The flow meter 320A measures data indicative of the mass flow rate of a process gas disposed along the first branch. The computing device 322 can determine the concentration of precursor 306A in the first process gas at the junction 308A.

[0052] As shown in FIG. 3, the second branch may include a second carrier gas source 310B, a flow meter 316B, a container 304B, a precursor 306B, and a flow meter 320B. The function and / or location of each of these elements may include features and descriptions described in connection with other figures (see FIG. 3). The flow meter 316B measures data indicative of the mass flow rate of the second carrier gas. The flow meter 320B measures data indicative of the mass flow rate of the process gas disposed along the first branch. The computing device 322 can determine the concentration of precursor 306B in the second process gas at the junction 308A.

[0053] The first process gas interacts with the second process gas and mixes at junction 340. The gas supply system 300 includes a mass flow meter 330 that measures the mass flow rate of a compound gas including the first process gas and the second process gas. The mass flow rate of the compound gas can be implemented as described herein.

[0054] As shown in FIG. 3 , the computing device 322 may be coupled to flow meters 316A, 320A, 316B, 320B, and 330. The computing device may determine the relative concentrations of one or more of the first precursor 306A and / or the second precursor 306B within the compound gas. The computing device may determine the concentrations using methodologies described herein (e.g., method 500 of FIG. 5 ). The concentration of each precursor may be determined relative to the overall compound gas. Note that FIG. 3 illustrates a system with two branches with identical elements. However, each branch may include more or fewer elements, such as valves 312A-B, 314A-B, 317A-B, 318A-B, and temperature sensors 323A-B, depending on the embodiment employed. Furthermore, combinations of three or more carrier gases and precursors may be used to determine the concentrations of precursors within the compound gas, such as having three or more carrier gas and / or precursor combinations.

[0055] As described in other embodiments, the computing device 322 may include a concentration tool 324, a system control module 326, and / or a precursor monitor 328. The concentration tool 324 receives first data indicating one or more of a first flow rate of the first carrier gas and / or a second carrier gas, and second data indicating a second flow rate of the first process gas and / or a second process gas. The concentration tool may calculate the concentration of the first precursor 206A in the first process gas and the concentration of the second precursor 206B in the second process gas, where the first precursor 206A and the second precursor 206B in the compound gas comprise the first process gas and the second process gas, respectively. As described above, the relationship between flow rate, density, and vaporization may be utilized to determine the relative concentration of each precursor. As further described in other embodiments, the computing device may receive data indicating the flow rates of additional flow paths as needed, aggregate this data, perform updated concentration calculations, and make further determinations. In some embodiments, as will be discussed further, sensor data indicative of the condition of the vessel 304 may also be received and incorporated into the determined precursor concentration.

[0056] The system control module 326 monitors and controls system methods such as controlling the flow rate of gases (e.g., carrier gases or process gases), adjusting conditions in the vessels 304A-B (e.g., changing temperature, pressure, etc.), and / or running process chamber procedures (e.g., CVD processes, ALD processes, etc.). Concentration determinations of the precursors 306A-B (e.g., by the concentration tool 324) can be incorporated into decisions made by the system control module 326. For example, if the concentration of the precursors 306A-B is higher than a threshold level, the system control module 326 can increase the flow rate of the carrier gas (e.g., by opening valves 312A-B or 314A-B to increase the flow rate) or decrease the flow rate of the precursors 306A-B into the vessels 304A-B (e.g., by partially closing valves 312A-B, 314A-B).

[0057] In some embodiments, the system control module 326 can determine that the precursor concentration meets the threshold criteria and can modify processing operations associated with the processing chamber 302. For example, processing operations in the chamber can be stopped until the precursor concentration no longer meets the threshold condition. In another example, a processing step can be performed under different parameters (e.g., a higher temperature, a longer etch duration, a longer deposition duration, etc.) based on the determined precursor concentration.

[0058] The precursor monitor 328 may receive concentration calculations of the precursors 306A-B (e.g., from the concentration tool 324) and make precursor-related determinations. In some embodiments, the precursor monitor 328 may determine the depletion rates of the precursors 306A-B in the vessels 204A-B based on the concentrations of the precursors 306A-B in each process gas. For example, flow rates and concentrations may be used to determine the depletion rates of the precursors 306A-B. In another example, flow rates, which indicate precursor input rates, such as the flow of the precursors 306A-B from precursor reservoirs into the vessels 304A-B, may be considered and contribute to the calculated depletion rates of the precursors 306A-B.

[0059] FIG. 4 illustrates a precursor delivery system 400 according to some embodiments. The precursor delivery system 400 is suitable for generating process gases including chemical precursors and generally includes a process chamber 406 and a carrier gas source 405 coupled to a gas panel 404, the latter component being controlled by a controller 450 or computing device. The gas panel 404 generally controls the rate and pressure at which various process and carrier gases are supplied to the process chamber 406. The process chamber 406 may be a chamber for performing vapor deposition or thermal processes involving vaporized chemical precursors in liquid, gas, or plasma states. The process chamber 406 is generally a chemical vapor deposition (CVD) chamber, an atomic layer deposition (ALD) chamber, or a derivative thereof.

[0060] In the configuration shown in FIG. 1 , the controller 450 includes a central processing unit (CPU) 452, memory 454, and support circuits 456. The central processing unit 452 can be any type of computer processor that can be used in an industrial environment to control various chambers and sub-processors. The memory 454 is coupled to the CPU 452 and can be one or more of readily available memory, such as random access memory (RAM), read-only memory (ROM), flash memory, compact disk, floppy disk, hard disk, or other form of local or remote digital storage. The support circuits 456 are coupled to the CPU 152 to support the CPU 152 in a conventional manner. These circuits include cache, power supplies, clock circuits, input / output circuits, subsystems, etc. The controller 450 can be and / or include a computing device, such as a personal computer, a server computer, a programmable logic controller (PLC), a microcontroller, etc. The controller 450 can include one or more processing devices, which can be general-purpose processing devices, such as a microprocessor, a central processing unit, etc. More specifically, the processing device may be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, or a processor implementing other instruction sets or combinations of instruction sets. The processing device may also be one or more special-purpose processing devices, such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), a network processor, etc. The controller 450 may include data storage (e.g., one or more disk drives and / or solid state drives), main memory, static memory, a network interface, and / or other components.The controller 450 may execute instructions to perform any one or more of the methodologies and / or embodiments described herein. The instructions may be stored in a computer-readable storage medium, which may include a main memory, a static memory, a secondary storage device, and / or a processing device during execution of the instructions.

[0061] Fluid supply circuit 436 generally fluidly couples carrier gas source 405, ampoule 470, and process chamber 406. Carrier gas source 405 may be a local container, a remote container, or a centralized facility source (e.g., an in-house gas supply) that supplies carrier gas throughout the facility. Carrier gas source 405 typically provides a carrier gas such as nitrogen, hydrogen, argon, helium, or a combination thereof. Fluid supply circuit 436 typically includes a flow controller 420 disposed between carrier gas source 405 and junction 430 and adapted to regulate the flow rate of the carrier gas or other fluid through fluid supply circuit 436. Flow controller 420 may be a proportional valve, an adjustable valve, a needle valve, a regulator, a mass flow controller (MFC), or the like. Junction 430 separates fluid supply circuit 436 into a gas generation line 438 and a bypass line 440. Junction 432 rejoins gas generation line 438 and bypass line 440 before connecting to process chamber 406.

[0062] Gas generation line 438 includes ampoule inlet leg 438A, ampoule outlet leg 438B, valves 408, 410, 412, sensors 426, 428, disconnect fittings 462, 463, and heater 422. Ampoule inlet leg 438A fluidly connected the inlet of ampoule 470 to carrier gas source 405 and bypass line 440. Ampoule outlet leg 438B fluidly connected the outlet of ampoule 470 to process chamber 406 and bypass line 440. Valves 408, 410, 412 are typically remotely controllable shut-off valves that function to divert fluid flow within fluid supply circuit 436 and / or are used to selectively isolate various components within fluid supply circuit 436 to facilitate removal, replacement, and / or maintenance of the isolated components, including sensors 426, 428, heater 422, and ampoule 470. Valves 408, 410, 412, as well as valves 414, 416, 418 (described below in conjunction with bypass line 440), are typically pneumatically or electronically controlled, and their interior wetted surfaces are fabricated from materials compatible with the process fluids and other fluids handled by fluid supply circuit 436. Valves 408, 410, 412, 414, 416, and 418 typically operate in response to signals from controller 450 or a computing device to regulate the supply of gas through fluid supply circuit 436. Sensors 426, 428 are typically adapted to detect the temperature of the process fluid and / or carrier fluid flowing through gas generation line 438, such as thermocouples positioned relative to the conduits of gas generation line 438.

[0063] The bypass line 440 generally includes valves 414, 416 and a heater 424 and functions to fluidly couple the processing chamber 406 and the carrier gas source 405 without the use of a gas generation line 438 or an ampoule 470. A valve 418 is typically coupled between the junction 432 and the processing chamber 406 and may be used to isolate the processing chamber 406 from the fluid supply circuit 436. The heaters 422, 424 are resistive heating elements or other heat sources adapted to heat a fluid flow, such as a carrier gas, flowing through the gas generation line 438 and the bypass line 440, respectively.

[0064] Flow meter 460 is disposed between junction 471 and ampoule 470. Flow meter 461 is disposed between junction 472 and ampoule 470. Flow meters 460 and 470 may include sensors capable of measuring data indicative of mass flow rate. For example, pressure, volumetric flow rate, density, etc., may all be measured. Flow meter 460 measures the flow rate of the carrier gas before passing through ampoule 470. Flow meter 461 measures the flow rate of the process gas (e.g., carrier gas and precursor) after passing through ampoule 470. The flow rate measurements may be received by CPU 452. The CPU can calculate the concentration of the precursor in the process gas based on the measured mass flow rate of the carrier gas and the mass flow rate of the process gas. Furthermore, in some embodiments, ampoule 470 may include sensors for measuring conditions of the ampoule, such as temperature, pressure, etc. The measured conditions of ampoule 470 may be received and further used to calculate the concentration of the precursor in the process gas.

[0065] Ampoule 470 may include terms such as bubble, canister, and other terms known in the art to describe containers designed and used to store, transport, and dispense chemical precursors. Ampoule 470 comes in a variety of sizes and shapes. Ampoule 470 may have a volumetric capacity of chemical precursor ranging from 0.5 L to about 10 L, more typically from about 1.2 L to about 4 L. In one example, ampoule 470 has a volumetric capacity of about 2.5 L. The chemical precursors in ampoule 470 include liquid, solid, and gaseous precursors, and are preferably in a liquid or fluid state at a predetermined temperature and / or pressure. For example, a chemical precursor may exist in a solid state at room temperature, but melt into a liquid state when heated to a predetermined temperature within ampoule 470. In another example, the majority of the chemical precursor may remain in a solid state within ampoule 470, but due to high temperatures during processing, a small amount of the solid precursor may sublimate directly into a vapor. In another example, a chemical precursor may exist in a gaseous state at ambient pressure, but condense into a liquid state when pressurized to a predetermined pressure within ampoule 470 .

[0066] The solid-state chemical precursors are pentakis(dimethylamido)tantalum (PDMAT; Ta(NMe2)5), pentakis(diethylamido)tertiaryamimido-tris(dimethylamido)tantalum (TAIMATA, ( t may be used to form a process gas containing a tantalum precursor such as AmylN)Ta(NMe2)3), t Amyl is a tertiary amyl group (CH 11 - or CH3CH2C(CH3)2-), or a derivative thereof. In one embodiment, the PDMAT has a low halide content (e.g., Cl, F, I, or Br). The halide concentration of the PDMAT may be less than about 100 ppm. For example, the PDMAT may have a chlorine concentration of less than about 100 ppm, preferably less than about 20 ppm, more preferably less than about 5 ppm, more preferably less than about 1 ppm, e.g., about 100 ppb or less.

[0067] Other solid chemical precursors that can be used to form a process gas through a sublimation process include xenon difluoride, nickel carbonyl, tungsten hexacarbonyl, or derivatives thereof. In other embodiments, liquid chemical precursors can be vaporized to form a process gas in the ampoules described herein. Some liquid chemical precursors that can be used to form a process gas include tungsten precursors such as tungsten hexafluoride (WF), tantalum precursors such as tantalum (PDEAT; Ta(NEt)), pentakis(methylethylamido)tantalum (PMEAT; Ta(NMeEt)), tertbutylimino-tris(dimethylamino)tantalum (TBTDMT), t BuNTa(NMe2)3), tertbutylimino-tris(diethylamino)tantalum (TBTDET, t BuNTa(NEt2)3), tertbutylimino-tris(methylethylamino)tantalum (TBTMET, ttitanium precursors such as BuNTa(NMeEt)3) or its derivatives, titanium tetrachloride (TiCl4), tetrakis(dimethylamino)titanium (TDMAT, (Me2N)4Ti)), tetrakis(diethylamino)titanium (TEMAT, (Et2N)4Ti)), or their derivatives; ruthenium precursors such as bis(ethylcyclopentadienyl)ruthenium ((EtCp)2Ru); hafnium precursors such as tetrakis(dimethylamino)hafnium (TDMAH, (Me2N)4Hf)), tetrakis(diethylamino)hafnium (TDEAH, (Et2N)4Hf)), tetrakis(methylethylamino)hafnium (TMEAH, (MeEtN)4Hf)), or their derivatives; Aluminum precursors include aluminum precursors such as pyridine:alane (MPA, MeC4H3N:AlH3), pyridine:alane (C4H4N:AlH3), alkylamine alane complexes (e.g., trimethylamine:alane (Me3N:AlH3), triethylamine:alane (Et3N:AlH3), dimethylethylamine:alane (Me2EtN:AlH3)), trimethylaluminum (TMA, Me3Al), triethylaluminum (TEA, Et3Al), tributylaluminum (BU3Al), dimethylaluminum chloride (Me2AlCl), diethylaluminum chloride (Et2AlCl), dibutylaluminum hydride (Bu2AlH), dibutylaluminum chloride (Bu2AlCl), or derivatives thereof.

[0068] During processing, carrier gas flows from carrier gas source 405 through fluid delivery circuit 436 to ampoule 470. The carrier gas may be heated by heater 422, ampoule 470 may be heated to a target temperature, or in some applications, both the carrier gas and ampoule 470 may be heated. During processing, valves 414 and 416 are closed and all carrier gas flow is directed through gas generation line 438 and ampoule 470 to process chamber 406.

[0069] 5 is a flowchart of a method 500 for determining the concentration of a precursor, according to some embodiments. Method 500 may be performed by processing logic, which may include hardware (e.g., circuitry, dedicated logic, programmable logic, microcode, processing device, etc.), software (e.g., instructions running on a processing device, a general-purpose computer system, or a dedicated machine), firmware, microcode, or any combination thereof. In some embodiments, method 500 may be performed by computing device 222 of FIG. 2.

[0070] Referring to FIG. 5, a method 500 relates to determining the concentration of a compound (eg, a precursor) in a gas delivery system (eg, a precursor delivery system).

[0071] At block 501, processing logic receives first data indicating a first mass flow rate of a carrier gas. The first data may be raw sensor data, such as a carrier gas flow rate, a molecular weight of the carrier gas, and / or a carrier gas pressure at a measurement location in a carrier gas flow path. The first data may be measured by one or more devices (e.g., sensors) described herein. The first data may be measured in a pressure-controlled environment. For example, the system may be designed to maintain a constant carrier gas pressure throughout a measurement region.

[0072] At block 502, the processing logic receives second data indicating a second mass flow rate of a compound gas. The compound gas includes a carrier gas and a vaporized substance (e.g., a precursor). The second data may be raw sensor data, such as a flow rate of the compound gas, a molecular weight of the compound gas, and / or a pressure of the compound gas. The second data may be measured by one or more devices (e.g., sensors) as described herein. The second data may be measured in a pressure-controlled environment. For example, the system may be designed to maintain constant pressure of the carrier gas and the compound gas throughout the measurement region.

[0073] In some embodiments, one or more of the first sensor and / or the second sensor include a mass flow controller. The computing device may send instructions to the first sensor and / or the second sensor to modify the flow rates of the carrier gas and / or the compound gas. As described herein, the flow rate control may be based on concentration measurements and / or determinations made by the processing device.

[0074] At block 503, the processing logic receives third data indicating a temperature of a vaporization vessel associated with the vaporization of the vaporized substance. The temperature may indicate a vaporization rate of the compound in the vaporization vessel. The vaporization vessel may be one or more of the structures disclosed herein (e.g., vaporization chamber 108 of FIG. 1, vessel 204 of FIG. 2, etc.).

[0075] At block 504, processing logic determines the concentration of the vaporized substance in the compound gas based on the first data, the second data, and the third data. The following gas properties may be utilized in performing mass flow measurements and calculating concentrations: TIFF0007796779000002.tif21170

[0076] In some embodiments, the processing device examines the change in density (e.g., the product of vapor pressure and molecular weight). For example, a gas with a molecular weight of 28 grams per mole (g / mol) may pass through a first flow path. This can be measured by determining with sensor 104 that the first gas is flowing at 250 standard cubic centimeters per minute (sccm) at a pressure of 100 torr. After passing through vaporization chamber 108 and removing the vaporized material, the vaporized material gas, including the first gas (e.g., carrier gas) and the vaporized material (e.g., precursor), flows through a second flow path. The second sensor can measure an average molecular weight of 30.89 g / mol, which is approximately 10% greater than the carrier gas alone. The average molecular weight can be determined by measuring a flow rate of approximately 275 sccm and a pressure of 100 torr. The rate at which the carrier gas increases with the vaporized material indicates the concentration of the vaporized material in the compound gas.

[0077] In some embodiments, the processing logic inputs the first data indicative of the first mass flow rate, the second data indicative of the second mass flow rate, and / or the third data indicative of the temperature of the vaporization vessel into a trained machine learning model that outputs a concentration of the vaporized substance.

[0078] It should be noted that the concentration calculation method is species-independent. Traditional concentration sensors, such as optical sensors, must be pre-calibrated and manufactured to measure a given species specification window, such as limits on density, molecular weight, flow rate, etc. The subject matter enabled herein may enable species-independent calculations that can work for a variety of gases (e.g., carrier gases and process gases) and a variety of compounds (e.g., precursors).

[0079] At block 505, the processing logic optionally modifies the flow rate of the carrier gas. In some embodiments, the process monitors and controls gas supply parameters, such as controlling gas flow rates (e.g., carrier gas and process gas), adjusting conditions in the vaporization vessel (e.g., changing temperature, pressure), and / or performing a process chamber procedure (e.g., a CVD process or an ALD process). The concentration determination can be incorporated into decisions made by the processing device. For example, if the precursor concentration is higher than a threshold level, the processing device may increase the flow rate of the carrier gas (e.g., open a valve) or decrease the flow rate of the volatile material to the vessel (e.g., partially close a valve). In some embodiments, the processing device can determine that the concentration of the vaporized material meets a threshold criterion and modify processing operations associated with the processing chamber (e.g., a substrate processing chamber). For example, processing operations in the chamber can be stopped until the concentration of the vaporized material no longer meets the threshold condition. In another example, a processing step can be performed under different parameters (e.g., a higher temperature, a longer etch duration, a longer deposition duration, etc.) based on the determined concentration of the volatile material.

[0080] In some embodiments, the processing logic inputs the first data indicating the first mass flow rate, the second data indicating the second mass flow rate, and / or the third data indicating the temperature of the vaporization vessel into a trained machine learning model that outputs updates to one or more process parameters, such as changes to the flow rates of the carrier gas and / or process gas, updates to the state of the vaporization vessel, and / or other process parameters.

[0081] At block 506, processing logic optionally provides the concentrations for display by a graphical user interface (GUI). The concentrations may be stored and displayed later as part of a post-mortem analysis. In some embodiments, the concentrations may be calculated and / or displayed while the gas delivery process is occurring. For example, the concentration of a precursor may be tracked throughout a portion of a substrate processing procedure.

[0082] In some embodiments, the processing logic receives fourth data from a fourth sensor indicating a third mass flow rate of the second carrier gas. The compound gas may further include a second carrier gas.

[0083] In some embodiments, the processing logic determines an amount of vapor to place in the vaporization container based on the determined concentration. The determined amount of vapor may be provided for display in a GUI.

[0084] 6 illustrates a model training workflow 605 and a model application workflow 617 for performing process control based on a concentration sensor and a detected concentration of a process gas, according to one embodiment of the present disclosure. In an embodiment, the model training workflow 605 can be executed on a server that may or may not include a concentration sensor application, and the trained model is provided to the concentration sensor application (e.g., on computing device 222 of FIG. 2 ), which can execute the model application workflow 617. The model training workflow 605 and the model application workflow 617 can be performed by processing logic executed by a processor of the computing device. One or more of these workflows 605, 617 can be implemented, for example, by one or more machine learning modules implemented in concentration tool 224, system control module 226, precursor monitor 228, and / or other software and / or firmware executing on the processing device of computing device 222 shown in FIG. 2 .

[0085] The model training workflow 605 trains one or more machine learning models (e.g., deep learning models) to perform one or more classification, segmentation, detection, recognition, decision, etc. tasks related to the concentration sensor (e.g., measuring mass flow rates, calculating concentrations, determining precursor data, determining system control data and / or diagnostics, determining process changes, etc.). The model application workflow 617 applies one or more trained machine learning models to perform tasks such as classification, segmentation, detection, recognition, decision, etc. of vaporization data (e.g., carrier gas mass flow rate data, process gas mass flow rate data, precursor quantity and decay rate, vaporization chamber data, etc.). One or more of the machine learning models may receive and process gas delivery data (e.g., molecular weight of one or more associated gases, carrier gas mass flow rate, process gas mass flow rate including vaporized precursor). One or more of the machine learning models may receive and process vaporization chamber data (e.g., temperature, pressure, mass flow rate within the chamber, etc.) and precursor data (molecular weight, chemical composition, quantity within the vaporization chamber, decay rate, etc.).

[0086] Various machine learning outputs are described herein. Specific numbers and arrangements of machine learning models are described and illustrated. However, it should be understood that the number and types of machine learning models used, and the arrangements of such machine learning models, can be varied to achieve the same or similar end results. Therefore, the arrangements of machine learning models described and illustrated are merely examples and should not be construed as limiting.

[0087] In embodiments, one or more machine learning models are trained to perform one or more of the following tasks: Each task may be performed by a separate machine learning model; Alternatively, a single machine learning model may perform each task or a subset of the tasks; Additionally or alternatively, different machine learning models may be trained to perform different combinations of tasks; In one example, one or several machine learning models may be trained, where the trained ML model is a single shared neural network with multiple shared layers and multiple higher-level, individual output layers, each output layer outputting a different prediction, classification, identification, etc. Tasks that the one or more trained machine learning models may be trained to perform include: a. Precursor Concentration Determination—As discussed above, the relationship between flow rate, density, and vaporization can be used to determine the relative concentration of precursors in the process gas. The computing device may optionally receive data indicating the flow rates of additional flow paths, which may be aggregated with the first and second data to perform updated concentration calculations and make further determinations, such as precursor concentration classification, as described herein. The precursor concentration determination may be related to the airflow hardware configuration (e.g., number of flow paths, flow path splits and branches, number of precursors, number of carrier gases, etc.). In some embodiments, the precursor concentration determination provides a calibrated concentration. b. Precursor Concentration Classification—As previously mentioned, the computing device may include a precursor monitor that determines the rate of precursor depletion in the vaporization vessel. The flow rate and precursor concentration may be used to classify the precursor concentration (e.g., within threshold limits, above threshold limits, within threshold limits but depleted above a depletion threshold, etc.). c. System Control Category—As previously described, a computing device may control a gas flow system, such as controlling the flow rate of a gas (e.g., a carrier gas or a process gas), adjusting a vessel condition (e.g., changing the temperature, pressure, etc.), and / or executing a process chamber procedure (e.g., a CVD or ALD process). The machine learning model may output instructions that can be displayed and / or applied to the gas delivery system, such as a corrective or prescribed action for a vaporization system to take. For example, a machine learning module may detect that a precursor concentration is above a threshold level and may instruct the vaporization system to increase the flow rate of the carrier gas (e.g., to dilute or reduce the precursor concentration in the process gas).

[0088] One type of machine learning model that can be used to perform some or all of the above tasks is an artificial neural network, such as a deep neural network. Artificial neural networks typically contain a feature representation component with a classifier or recurrent layer that maps features to a desired output space. For example, a convolutional neural network (CNN) hosts multiple layers of convolutional filters. Pooling is performed to address nonlinearities in lower layers, and a multilayer perceptron is typically added on top to map the top-layer features extracted by the convolutional layers to a decision (e.g., a classification output). Deep learning is a class of machine learning algorithms that uses a cascade of multiple layers of nonlinear processing units for feature extraction and transformation. Each subsequent layer uses the output from the previous layer as input. Deep neural networks can be trained in a supervised (e.g., classification) and / or unsupervised (e.g., pattern analysis) manner. Deep neural networks contain a hierarchy of layers, with different layers learning different levels of representations corresponding to different levels of abstraction. In deep learning, each level learns to transform input data into more abstract and complex representations. For example, in precursor classification, the raw input may be mass flow measurements of the carrier and / or process gases, the second layer may comprise feature data related to vaporization chamber conditions, and the third layer may contain identifying features of the carrier and / or process gases (e.g., molecular weight, density, chemical composition). In particular, deep learning processes can independently learn which features are best placed at which levels. The "deep" in "deep learning" refers to the number of layers through which data is transformed. More precisely, deep learning systems have substantial credit assignment path (CAP) depth. A CAP is a chain of transformations from input to output. A CAP describes the potential causal relationships between inputs and outputs. For feedforward neural networks, the CAP depth can be the network depth, which may be the number of hidden layers plus one. For recurrent neural networks, where signals may propagate through layers more than once, the CAP depth can be unlimited.

[0089] In one embodiment, the one or more machine learning models are recurrent neural networks (RNNs). RNNs are a type of neural network with memory that allows the neural network to capture temporal dependencies. RNNs can learn input-output mappings that depend on both current and past inputs. RNNs account for past and future flow measurements and make predictions based on this continuous measurement information. RNNs can be trained using a training dataset to generate a fixed number of outputs (e.g., to determine the concentration of volatiles at various points along the flow path of a gas distribution system). One type of RNN that can be used is a long short-term memory (LSTM) neural network.

[0090] Training a neural network can be achieved with supervised learning methods, which involve feeding a training dataset consisting of labeled inputs through the network, observing its output, defining an error (by measuring the difference between the output and the label value), and adjusting the network weights across all layers and nodes using techniques such as deep gradient descent or backpropagation so that the error is minimized. In many applications, repeating this process for many labeled inputs in the training dataset produces a network that can generate the correct output when given inputs that differ from those present in the training dataset.

[0091] The model training workflow 605 requires forming a training dataset (e.g., gas delivery data 610) that includes hundreds, thousands, tens of thousands, hundreds of thousands, or more carrier and process gas flow rate measurements. In embodiments, the training dataset can also include associated precursor classifications 612 to form the training dataset, where each data point and / or associated precursor classification can include a different label or classification of one or more types of useful information. Each case can include, for example, data indicating a first flow rate of a carrier gas and a second flow rate of a process gas, as well as a determined precursor classification (e.g., precursor concentration, decay rate, amount stored in the vaporization chamber, etc.). This data can be processed to generate one or more training datasets 636 for training one or more machine learning models. The machine learning models can be trained, for example, to automate one or more processes of the precursor delivery system (e.g., increasing or decreasing the flow rate of the carrier gas and / or process gas, increasing the amount of precursor in the vaporization chamber, and other processes related to the delivery of precursors).

[0092] In some embodiments, a training data set including hundreds, thousands, tens of thousands, hundreds of thousands, or more carrier and process flow rate measurements (e.g., gas delivery data 610) is used to form a training data set. The training data set may also include associated precursor classifications 612 to form the training data set. The precursor classifications 612 may include one or more concentration measurements (e.g., performed using a residual gas analyzer RGA or other concentration sensor).

[0093] In one embodiment, generating the one or more training data sets 636 includes collecting one or more gas flow rate measurements of carrier gases and process gases. The labels used may depend on what the particular machine learning model is trained to do. For example, to train a machine learning model to perform precursor classification, the training data set 636 may include data indicative of gas type (e.g., molecular weight, density, etc.), gas flow rate measurements (e.g., mass flow rates of carrier gases and / or process gases), and semiconductor process specifications. For example, a semiconductor process may be associated with a window of precursor concentrations associated with semiconductor process results that meet threshold criteria.

[0094] To enable training, processing logic inputs the training dataset 636 to one or more untrained machine learning models. Before inputting the first input to the machine learning models, the machine learning models can be initialized. Processing logic trains the untrained machine learning models based on the training dataset to generate one or more trained machine learning models that perform the various operations described above.

[0095] Training may be performed by inputting one or more of the gas delivery data 610 and precursor classifications 612 into the machine learning model one at a time. In some embodiments, training the machine learning model includes tuning the model to receive gas delivery data 610 (e.g., mass flow rates of carrier gas and process gas) and outputting precursor concentration predictions (e.g., precursor classifications 612) within a threshold difference of measured precursor concentrations (e.g., performed using a residual gas analyzer RGA or other concentration sensor).

[0096] Machine learning models process inputs and generate outputs. An artificial neural network includes an input layer, which consists of the values ​​of the data points. The next layer is called the hidden layer, and each node in the hidden layer receives one or more input values. Each node contains parameters (e.g., weights) that it applies to the input values. Thus, each node essentially feeds the input values ​​into a multivariate function (e.g., a nonlinear mathematical transform) to generate an output value. The next layer may be another hidden layer or an output layer. In either case, the nodes in the next layer receive output values ​​from the nodes in the previous layer, and each node applies weights to those values ​​to generate its own output value. This may be done at each layer. The final layer is the output layer, which has one node for each class, prediction, and / or output that the machine learning model can generate.

[0097] Thus, the output may include one or more predictions or inferences. For example, the output prediction or inference may include a determined concentration of a precursor in a precursor delivery system. Processing logic may then compare the predicted or inferred output to the measured or known precursor classification (e.g., measured concentration) included in the training data item. Processing logic determines an error (i.e., classification error) based on the difference between the output of the machine learning model and the known classification (e.g., precursor classification). Processing logic adjusts the weights of one or more nodes in the machine learning model based on the error. An error term, or delta, may be determined for each node in the artificial neural network. Based on this error, the artificial neural network adjusts one or more parameters of one or more nodes (e.g., weights of one or more inputs of the node). Parameters may be updated in a back-propagation manner, with nodes in the top layer updated first, followed by nodes in the next layer, and so on. An artificial neural network includes multiple layers of "neurons," each layer receiving input values ​​from neurons in the previous layer. Each neuron's parameters include a weight associated with the value received from each neuron in the previous layer. Thus, adjusting the parameters may involve adjusting the weights assigned to each of the inputs of one or more neurons in one or more layers of the artificial neural network.

[0098] Once the model parameters are optimized, model validation can be performed to determine whether the model has improved and to determine the current accuracy of the deep learning model. After one or more training rounds, the processing logic may determine whether a stopping criterion has been met. The stopping criterion may be a target level of accuracy, a target number of processed images from the training dataset, a target amount of change in parameters relative to one or more previous data points, a combination thereof, and / or other criteria. In one embodiment, the stopping criterion is met when at least a minimum number of data points have been processed and at least a threshold accuracy has been achieved. The threshold accuracy may be, for example, 70%, 80%, or 90% accuracy. In one embodiment, the stopping criterion is met when the accuracy of the machine learning model stops improving. If the stopping criterion is not met, further training is performed. If the stopping criterion is met, training may be complete. Once the machine learning model is trained, a reserved portion of the training dataset can be used to test the model.

[0099] As an example, in one embodiment, a machine learning model (e.g., precursor classifier 668) is trained to determine precursor concentration and / or precursor state in a vaporization chamber. Similar processes can be performed to train machine learning models to perform other tasks, such as those described above. A large number (e.g., thousands to millions) of sets of gas delivery measurements (e.g., mass flow measurements of carrier gas and / or process gas) can be collected, and concentration data 666 associated with predicted or inferred concentrations associated with input data 662 can be determined.

[0100] Once the one or more trained machine learning models 638 are generated, they may be stored in model storage 645 and added to a concentration sensor application (e.g., concentration tool 224, system control module 226, and / or precursor monitor 228). The concentration sensor application may then use the one or more trained ML models 638 and additional processing logic to implement an automatic mode in which manual input of information by a user is minimized, and in some cases may even be eliminated.

[0101] In one embodiment, the model application workflow 617 includes one or more trained machine learning models that function as the system control classifier 667, the precursor concentration determiner 664, and / or the precursor classifier 668. These logics may be implemented, in embodiments, as separate machine learning models or as a single combined machine learning model. For example, the system control classifier 667, the precursor concentration determiner 664, and the precursor classifier 668 may share one or more layers of a deep neural network. However, each of these logics may include separate, higher-level layers of a deep neural network trained to generate different types of outputs. For convenience, the illustrated example shows only a portion of the functionality described in the list of tasks above. However, it should be understood that any other tasks may also be added to the model application workflow 617.

[0102] For the model application workflow 617, according to one embodiment, input data 662 may be input to a system control classifier 667, which may include a trained neural network. The system control classifier 667 outputs information (e.g., delivery data 669) indicative of the state of the precursor delivery system based on the input data 662. This may include outputting a set of classification probabilities for prescribed actions 672. The prescribed actions 672 may include actions that, when applied to the gas delivery system, change the state of the precursor delivery system and are communicated to a controller (e.g., a user and / or an automated system) (e.g., reduce the concentration of precursor in the process gas, reduce the temperature of the vaporization chamber, stop operation, etc.).

[0103] According to one embodiment, the input data 662 may be input to a precursor concentration determiner 664, which may include a trained neural network. Based on the input data 662, the precursor concentration determiner 664 outputs a determination of the concentration of the precursor in the precursor delivery system associated with the input data 662 (e.g., concentration data 666). For example, a machine learning model may be tuned to receive the gas delivery data 610 (e.g., mass flow rates of carrier gas and process gas), whose mass flow rates are measured, and output a precursor concentration prediction (e.g., precursor classification 612) within a threshold difference of the measured precursor concentration (e.g., performed using a residual gas analyzer RGA or other concentration sensor) of the same carrier gas and process gas used as input to the precursor concentration determiner 664.

[0104] According to one embodiment, input data 662 may be input to a precursor classifier 668, which may include a trained neural network. Based on the input data 662, precursor classifier 668 may classify precursors in the precursor delivery system. For example, the output may indicate a rate of precursor depletion in a vaporization chamber. The output may indicate a quantity of precursor in a vaporization chamber. The output may indicate a chemical composition of the precursor (e.g., concentration percentages of multiple precursors).

[0105] 7 is a flowchart illustrating an embodiment of a method 700 for training a machine learning model to determine the concentration of a volatile substance in a process gas, in accordance with an embodiment of the present disclosure. At block 702 of method 700, processing logic collects a training data set, which may include flow rates (e.g., mass flow rates) of a carrier gas and a process gas in a gas delivery system. Each data item in the training data set (e.g., mass flow rate of the carrier gas and / or mass flow rate of the process gas) may include one or more known concentration levels of a volatile substance disposed in the process gas.

[0106] At block 704, data items from the training dataset are input to an untrained machine learning model. At block 706, the machine learning model is trained based on the training dataset to generate a trained machine learning model that determines the concentration of a volatile substance (e.g., precursor) in a process gas. The machine learning model may also be trained to output one or more other types of predictions, classifications, decisions, etc. For example, a machine learning model may be trained to classify precursors and make a decision (e.g., performed manually or automatically) to modify the operation of a gas delivery system (e.g., gas delivery system 200 of FIG. 2 ).

[0107] In one embodiment, at block 710, training data item inputs are input to a machine learning model. The inputs may include gas flow rate data for a gas delivery system. At block 712, the machine learning model processes the inputs to generate outputs. The outputs may include predictions and / or inferences of concentrations of volatiles (e.g., precursors) in process gases of the gas delivery system and / or one or more process updates for a manufacturing process that uses the gas delivery system (e.g., to perform an atomic layer deposition process, a chemical vapor deposition process, or other process).

[0108] At block 714, processing logic compares the output to known concentration measurements of volatile substances disposed in the associated process gas. At block 716, processing logic determines an error based on the difference between the output and the concentration measurements. At block 718, processing logic adjusts weights of one or more nodes in the machine learning model based on the error.

[0109] Additionally, at block 714, processing logic may compare output probabilities of other predictions, classifications, etc., to one or more other labels associated with the input. For example, one or more process parameter updates may be output by the machine learning model and compared to the label of the appropriate process parameter to be used. At block 716, processing logic may determine an error for each of the comparisons. At block 718, processing logic may adjust the weights of one or more nodes in the machine learning model based on these errors. Thus, the machine learning model may be trained to perform concentration determination, precursor classification, and / or one or more other classification, determination, or prediction operations.

[0110] At block 720, processing logic determines whether a stopping criterion is met. If the stopping criterion is not met, the method returns to block 710, where another training data item is input to the machine learning model. If the stopping criterion is met, the method proceeds to block 725, where training of the machine learning model is complete.

[0111] In one embodiment, multiple different ML models are trained for concentration determination, precursor classification, and / or system control classification. Each ML model may be trained for determination and / or classification of a different type of input data. For example, a first ML model may be trained to perform concentration determination using carrier gas and process gas mass flow rate measurements, a second ML model may be trained to perform concentration determination using data indicative of vaporization chamber conditions (e.g., temperature, pressure, etc.), and a third ML model may be trained to perform concentration determination using a combination of chamber data and gas flow rate data. In one embodiment, a single ML model is trained to perform the operations of the first, second, and third ML models described above.

[0112] 8 shows a flow diagram of an example of a method for determining the concentration of a volatile substance (e.g., a precursor) using a trained machine learning model according to some implementations of the present disclosure. Method 800 is performed by processing logic, which may include hardware (circuitry, dedicated logic, etc.), software (such as running on a general-purpose computer system or a dedicated machine), or any combination thereof. In one implementation, the method is performed using computing device 222 and trained machine learning model 638, although in some other implementations, one or more blocks of FIG. 8 may be performed by one or more other machines not shown.

[0113] Method 800 may include receiving gas delivery data (e.g., gas flow rate data described in connection with FIGS. 2-4 ) and processing the gas delivery data using a trained model, such as trained machine learning model 638. In some embodiments, the trained model may be configured to generate, based on the gas delivery, one or more outputs indicative of (i) a predicted concentration of a volatile substance disposed in an associated process gas and (ii) a confidence level associated with the accuracy of the predicted concentration. In some embodiments, the trained machine learning model may be configured, based on the gas delivery, to generate one or more outputs indicative of (i) an update to a process control parameter and (ii) a confidence level that the update is likely to improve one or more operating conditions of the gas delivery system.

[0114] At block 802, gas supply data associated with a gas supply system (e.g., gas supply system 200 of FIG. 2) is identified. The gas supply data may include data indicative of a first mass flow rate of a carrier gas, a second mass flow rate of a compound gas including the carrier gas and a vaporizer, and / or a temperature of a vaporization vessel associated with vaporization of the vaporizer.

[0115] In some embodiments, the gas delivery data further includes synthetic data or data processed from raw sensor data. For example, as described in previous embodiments, various engineering tools can perform feature extraction and / or create artificial and / or virtual parameter combinations. A feature extractor can create various features by performing variable analysis, such as process control analysis, univariate limit violation analysis, and / or multivariate limit violation analysis, on the raw sensor data.

[0116] At block 804, the gas delivery data is provided as input to a trained machine learning model, and at block 806, one or more outputs are obtained from the trained machine learning model. The one or more outputs may include precursor concentrations, precursor classifications, and / or process control updates. At block 808, confidence data is extracted from the outputs obtained at block 806. In one implementation, the confidence data includes a confidence level associated with the accuracy of the predicted concentrations. In one example, the confidence level is a real number between 0 and 1. Note that the confidence level may not be a probability (e.g., the sum of the confidence levels for a given action may not equal 1). At block 810, processing logic determines whether the confidence level satisfies a threshold condition.

[0117] At block 812, the processing logic optionally modifies the flow rate of the carrier gas. In some embodiments, the process monitors and controls gas delivery parameters, such as controlling the flow rate of a gas (e.g., a carrier gas or a process gas), adjusting conditions in a vaporization vessel (e.g., changing temperature, pressure, etc.), and / or performing a process chamber procedure (e.g., a CVD process or an ALD process). The concentration determination may be incorporated into decisions made by the processing device. For example, in response to an indication of a precursor concentration higher than a threshold level, the processing device may increase the flow rate of the carrier gas (e.g., opening a valve) or decrease the flow rate of volatiles to the vessel (e.g., partially closing a valve). In some embodiments, the processing logic may determine that the concentration of the vaporized material meets a threshold criterion and modify processing operations associated with the processing chamber (e.g., a substrate processing chamber). For example, processing operations in the chamber may be stopped until the concentration of the vaporized material no longer meets the threshold condition. In another example, a processing step may be performed under different parameters (e.g., a higher temperature, a longer etch duration, a longer deposition duration, etc.) based on the determined concentration of the volatile material. In some embodiments, the machine learning outputs modifications and / or adjustments to the gas supply system. For example, the machine learning model may indicate changes in the flow rate of gases (such as carrier gases and / or process gases).

[0118] At block 814, processing logic optionally provides the concentrations for display by a graphical user interface (GUI). The concentrations may be stored and displayed later as part of a post-mortem analysis. In some embodiments, the concentrations may be calculated and / or displayed while the gas delivery process is occurring. For example, the concentration of a precursor may be tracked throughout a portion of a substrate processing procedure.

[0119] 9 shows a block diagram of an exemplary computing device capable of calculating the concentration of a gas, operating in accordance with one or more aspects of the present disclosure. In various illustrative examples, various components of computing device 900 may represent various components of computing device 222 and / or controller 450.

[0120] The exemplary computing device 900 may be connected to other computer devices within a LAN, an intranet, an extranet, and / or the Internet. The computing device 900 may operate in the capacity of a server in a client-server network environment. The computing device 900 may be a personal computer (PC), a set-top box (STB), a server, a network router, switch, or bridge, or any device capable of executing a set of instructions (sequential or otherwise) that specify actions to be performed by that device. Furthermore, while only one example of a computing device is shown, the term "computer" shall also be taken to include any collection of computers that individually or jointly execute a sequence (or multiple sets) of instructions to perform any one or more of the methodologies described herein.

[0121] The exemplary computing device 900 may include a processing device 902 (also referred to as a processor or CPU), a main memory 904 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM)), a static memory 906 (e.g., flash memory, static random access memory (SRAM), etc.), and a secondary memory (e.g., data storage device 918), which may communicate with each other via a bus 930.

[0122] Processing device 902 represents one or more general-purpose processing devices, such as a microprocessor, a central processing unit, or the like. More specifically, processing device 902 may be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor implementing other instruction sets, or a processor implementing a combination of instruction sets. Processing device 902 may also be one or more special-purpose processing devices, such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), a network processor, or the like. According to one or more aspects of the present disclosure, processing device 902 may be configured to execute instructions implementing method 500 shown in FIG. 5.

[0123] The exemplary computing device 900 may further include a network interface device 908 that may be communicatively coupled to a network 920. The exemplary computing device 900 may further include a video display 910 (e.g., a liquid crystal display (LCD), a touch screen, or a cathode ray tube (CRT)), an alphanumeric input device 912 (e.g., a keyboard), a cursor control device 914 (e.g., a mouse), and an audio signal generating device 916 (e.g., a speaker).

[0124] The data storage device 918 may include a machine-readable storage medium (or, more specifically, a non-transitory machine-readable storage medium) 928 on which is stored one or more sets of executable instructions 922. According to one or more aspects of the present disclosure, the executable instructions 922 may include executable instructions associated with performing the methods 500-800 illustrated in FIGS.

[0125] The executable instructions 922 may also reside, completely or at least partially, within the main memory 904 and / or within the processing device 902 during execution by the exemplary computing device 900, the main memory 904 and the processing device 902 constituting computer-readable storage media. The executable instructions 922 may also be transmitted or received over a network via the network interface device 908.

[0126] Although computer-readable storage medium 928 is shown in FIG. 9 as a single medium, the term "computer-readable storage medium" should be interpreted to include a single medium or multiple media (e.g., a centralized or distributed database and / or associated caches and servers) that store one or more sets of operating instructions. The term "computer-readable storage medium" should also be interpreted to include any medium that can store or encode a set of instructions for execution by a machine that cause the machine to perform any one or more of the methodologies described herein. Accordingly, the term "computer-readable storage medium" should be interpreted to include, but is not limited to, solid-state memory, optical media, and magnetic media.

[0127] Some portions of the above detailed descriptions are presented in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. An algorithm is here, and generally, conceived to be a self-consistent sequence of steps leading to a desired result. The steps are steps requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It is sometimes convenient, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.

[0128] It should be borne in mind, however, that all of these and similar terms are associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless otherwise indicated, as will be apparent from the discussion that follows, throughout the description, discussions using terms such as "identify," "determine," "store," "adjust," "cause," "return," "compare," "create," "stop," "load," "copy," "throw," "replace," "execute," and the like refer to operations and processes of a computer system or similar electronic computing device that manipulate and convert data represented as physical (electronic) quantities in the registers or memory of the computer system into other data similarly represented as physical quantities in the memory or registers of the computer system, or other similar information storage, transmission, or display device.

[0129] Examples of the present disclosure also relate to apparatus for performing the methods described herein. This apparatus may be specially constructed for the intended purposes or may be a general-purpose computer system that is selectively programmed by a computer program stored on the computer system. Such computer program may be stored on a computer-readable storage medium, including, but not limited to, any type of disk, such as an optical disk, a compact disk read-only memory (CD-ROM), a magneto-optical disk, a read-only memory (ROM), a random access memory (RAM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic disk storage medium, an optical storage medium, a flash memory device, or any other type of machine-accessible storage medium, or any type of medium suitable for storing electronic instructions, each coupled to a computer system bus.

[0130] The methods and displays presented herein are not inherently related to any particular computer or other apparatus. Various general-purpose systems may be used with programs in accordance with the teachings herein, or it may prove convenient to construct more specialized apparatus to perform the recited method steps. The structure of a variety of these systems will be described below. Moreover, the scope of the disclosure is not limited to any particular programming language. It will be understood that a variety of programming languages ​​can be used to implement the teachings of the disclosure.

[0131] It should be understood that the above description is illustrative, and not restrictive. Many other implementations will be apparent to those skilled in the art upon reading and understanding the above description. While the present disclosure describes particular examples, it will be recognized that the systems and methods of the present disclosure are not limited to the examples described herein, but may be practiced with modification within the scope of the appended claims. Accordingly, the specification and drawings should be regarded in an illustrative, and not a restrictive, sense. The scope of the present disclosure should, therefore, be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

Claims

1. a vaporization chamber containing a compound and for transitioning the compound to a gaseous state; a first flow path coupled to the vaporization chamber for conducting a first gas to the vaporization chamber; a second flow path coupled to the vaporization chamber for directing a second gas containing the compound and the first gas out of the vaporization chamber; a first sensor disposed along the first flow path for measuring first flow data indicative of a first mass flow rate of the first gas within the first flow path; a second sensor disposed along the second flow path for measuring second flow data indicative of a second mass flow rate of the second gas within the second flow path; a third sensor for measuring third flow data indicative of a third mass flow rate of the compound into the vaporization chamber; a controller coupled to the first sensor and the second sensor for determining a concentration of the compound in the second gas based on the first flow data, the second flow data, and the third flow data; A concentration sensor assembly comprising:

2. The concentration sensor assembly of claim 1 , wherein at least one of the first sensor or the second sensor comprises a mass flow controller.

3. a first valve disposed along the first flow path; a second valve disposed along the second flow path; Furthermore, the first valve selectively opens and closes to vary a first flow rate of the first gas, and the second valve selectively opens and closes to vary a second flow rate of the second gas; The concentration sensor assembly of claim 1 .

4. The concentration sensor assembly of claim 1 , wherein the compound comprises a precursor for treating a substrate.

5. The concentration sensor assembly of claim 1 , further comprising a third flow path coupled to the vaporization chamber, the third flow path directing the compound to the vaporization chamber.

6. A concentration sensor assembly as described in claim 5, wherein the third sensor is positioned along the third flow path and the third flow rate data indicates a third mass flow rate of the compound in the third flow path.

7. The concentration sensor assembly of claim 6 , wherein the controller further determines a rate of depletion of the compound in the vaporization chamber based on the concentration.

8. 10. The concentration sensor assembly of claim 1, further comprising a fourth sensor for measuring fourth data indicative of a temperature of the vaporization chamber, wherein the controller calculates the concentration of the compound further based on the fourth data.

9. a vaporization vessel containing a precursor; a first flow path coupled to the vaporization vessel for directing a carrier gas into the vaporization vessel; a second flow path coupled to the vaporization vessel for directing a process gas containing the carrier gas and the precursor out of the vaporization vessel; a processing chamber coupled to the second flow path, the second flow path directing the process gas to the processing chamber; a first flow meter disposed along the first flow path for measuring first flow data indicative of a first flow rate of the carrier gas within the first flow path; a second flow meter disposed along the second flow path for measuring second flow data indicative of a second flow rate of the process gas within the second flow path; a third flow meter for measuring third flow data indicative of a third flow rate of the precursor into the vaporization vessel; a controller coupled to the first flow meter and the second flow meter for determining a concentration of the precursor in the process gas based on the first flow data, the second flow data, and the third flow data; A precursor delivery system comprising:

10. 10. The precursor delivery system of claim 9, wherein the precursor comprises a precursor for the third flow rate data that processes a substrate in the processing chamber.

11. 10. The precursor delivery system of claim 9, further comprising a first valve disposed along the first flow path and a second valve disposed along the second flow path, the first valve selectively opening and closing to vary the first flow rate of the carrier gas, and the second valve selectively opening and closing to vary the second flow rate of the process gas.

12. A precursor supply system as described in claim 9, wherein the third flow meter is positioned along a third flow path, the third flow path directs the precursor to the vaporization vessel, and the third data indicates a third flow rate of the precursor in the third flow path.

13. 10. The precursor delivery system of claim 9, further comprising a fourth flow meter disposed along a third flow path, the third flow path directing a second carrier gas into the vaporization vessel, the fourth flow meter measuring fourth data indicative of a fourth flow rate of the second carrier gas in the third flow path, and the controller further using the fourth data to calculate the concentration of the precursor.

14. 10. The precursor delivery system of claim 9, wherein the controller further determines a depletion rate of the precursor in the vaporization vessel based on the concentration.

15. 10. The precursor delivery system of claim 9, further comprising a temperature sensor for measuring third data indicative of a temperature of the vaporization vessel, wherein the controller further uses the third data to calculate the concentration of the precursor.

16. receiving, by a processing device, first flow rate data from a first sensor indicative of a first mass flow rate of the carrier gas in a first flow path connected to the vaporization chamber; receiving, by the processing device, second flow rate data from a second sensor indicative of a second mass flow rate of a compound gas including the carrier gas and a vaporized material in a second flow path coupled to a vaporization chamber; receiving, by the processing device, third data from a third sensor indicative of a temperature of a vaporization vessel associated with vaporization of the vaporizable material; receiving, by the processing device, fourth data from a fourth sensor indicative of a third mass flow rate of a second carrier gas in a third flow path coupled to a vaporization chamber; determining, by the processing device, a concentration of the vaporized substance in the compound gas based on the first flow rate data, the second flow rate data, the third flow rate data, and the fourth flow rate data; performing at least one of: a) varying the flow rate of the carrier gas; or b) providing the concentration for display by a graphical user interface (GUI); A method comprising:

17. The method of claim 16 , wherein the vaporized material comprises a precursor for treating a substrate.

18. The method of claim 16, wherein the compound gas further comprises the second carrier gas.

19. using the first flow data, the second flow data, and the third flow data as inputs to a machine learning model; obtaining one or more outputs of the machine learning model, the one or more outputs being indicative of the concentration of the vaporized substance in the compound gas; 17. The method of claim 16, further comprising:

20. determining an amount of the vaporized substance disposed in the vaporization container based on the concentration; providing, via the GUI, the amount to be displayed; 17. The method of claim 16, further comprising:

Images