Systems, methods, apparatuses, and media for determining spatial distribution of concentrations of target components

Abstract

The application discloses a kind of target component concentration spatial distribution determination system, method, equipment and medium, it is related to semiconductor detection technical field, the system includes: sampling module is obtained target component sample in process chamber by sampling port, flow limiting element and sampling pipeline;The type and size of flow limiting element are determined according to the pressure of target component in process chamber and preset rule;Component analysis module generates mass spectrum characteristic signal according to target component sample;Spatial distribution determination module generates the concentration spatial distribution information of target component according to mass spectrum characteristic signal, the position coordinates of sampling port and spatial reconstruction model.The system generated between the mass spectrum characteristic signal of different sampling port has comparability, and the concentration spatial distribution information of target component can not only reflect the active ingredient concentration of certain local point in chamber, but also can obtain the concentration spatial distribution of active ingredient in entire chamber in concentration spatial distribution information.

Description

Technical Field

[0001] This invention relates to the field of semiconductor detection technology, and more specifically to a system, method, device, and medium for determining the spatial distribution of the concentration of a target component. Background Technology

[0002] In semiconductor etching processes (such as inductively coupled plasma etching and capacitively coupled plasma etching), the concentration and spatial distribution of active components such as fluorine and chlorine groups (such as free radicals such as F, Cl, and CF2) directly determine the etching rate, anisotropy, selectivity, and wafer uniformity.

[0003] Current gas analysis equipment features multiple inlets, which allow for switching between different process stages to protect the sensor. However, these inlets are often installed in the same or nearby physical area within the process chamber, lacking spatial distribution sensing capabilities. Furthermore, the gas pressure at sampling points corresponding to different inlets varies, and transmission losses differ between the transmission channels corresponding to different inlets, rendering the acquired signals from different sampling points incomparable. Summary of the Invention

[0004] This invention provides a system, method, device, and medium for determining the spatial distribution of target component concentration, in order to solve the problems that component concentration analysis results can only reflect the concentration of active components in a certain area of ​​the chamber, lack spatial information of active components, and that the acquisition signals from different sampling points are not comparable.

[0005] In a first aspect, the present invention provides a system for determining the spatial distribution of the concentration of a target component, the system comprising: a sampling module, a component analysis module, and a spatial distribution determination module; The sampling module includes a preset number of sampling ports, a preset number of current limiting elements, and a preset number of sampling pipelines; the spacing between the sampling ports is greater than a first preset threshold, and the spacing between the sampling ports is determined according to the type of spatial reconstruction model. The first preset threshold is used to control the sampling areas where the sampling ports are located to not overlap. The sampling module is connected to the component analysis module and is used to acquire component samples in the process chamber through the sampling port; The current limiting element is connected to the sampling port; the sampling module is used to adjust the component sample obtained by the sampling port through the current limiting element to obtain the target component sample, and to transmit the target component sample to the sampling pipeline. The type and size of the current limiting element are determined according to the pressure of the target component in the process chamber and the preset rules. The sampling module is used to transfer the target component sample to the component analysis module through the sampling pipeline. The length difference between different sampling pipelines is less than a second preset threshold, which is used to control the length difference between the sampling pipelines. The component analysis module is connected to the spatial distribution determination module. It is used to generate mass spectrometry characteristic signals based on the target component sample and input the mass spectrometry characteristic signals into the spatial distribution determination module. The spatial distribution determination module is used to generate spatial distribution information of the target component's concentration based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model.

[0006] In one optional implementation, the sampling module further includes: a path adjustment component and a path switching component; the sampling module also includes a heating belt; The pathway adjustment component is used to clean residual sample in the sampling tubing before transmitting the target component sample through the sampling tubing; The pathway switching component is connected to the pathway adjustment component through multiple first pathways. It is used to take the first pathway corresponding to the target sampling pipeline as the target pathway and connect the target pathway to the component analysis module so that the target pathway can transmit the target component sample to the component analysis module. The target sampling pipeline is the sampling pipeline that is transmitting the target component sample. Heating bands are wrapped around the outer wall of the sampling pipeline to adjust the temperature of the target component sample in the sampling pipeline.

[0007] The sampling pipeline is cleaned by the pathway adjustment component to eliminate sample cross-interference. The pathway switching component controls the opening or closing of different first pathways, so that the target component samples in different target sampling pipelines are input into the component analysis module separately. Multi-point time-series sampling can be achieved using only a single component analysis module, which can obtain gas samples from multiple locations at low cost and ensure accurate and reliable detection data.

[0008] In one alternative implementation, the passage adjustment assembly includes a three-way valve and a pre-pump. The first port of the three-way valve is connected to the sampling pipeline, the second port of the three-way valve is connected to the pre-pump, and the third port of the three-way valve is connected to the first passage. In the three-way valve, the first port and the second port are connected through the second passage, and the first port and the third port are connected through the third passage. The pathway adjustment component is used to close the third pathway corresponding to the candidate sampling pipeline and open the second pathway corresponding to the candidate sampling pipeline before transmitting the target component sample through the candidate sampling pipeline. The pathway adjustment component is also used to extract residual samples in the candidate sampling pipeline using a pre-pump, and to continue for a preset duration. The pathway adjustment component is also used to close the second pathway corresponding to the candidate sampling pipeline and open the third pathway corresponding to the candidate sampling pipeline after a preset time.

[0009] First, connect the candidate sampling pipeline to the pre-vacuum pump via a three-way valve. Then, use the pre-vacuum pump to perform a pre-vacuum operation to empty the sample remaining in the candidate sampling pipeline, thereby eliminating cross-contamination between different sampling channels and ensuring accurate and comparable multi-point sampling and detection data.

[0010] In one optional implementation, the spatial distribution determination module includes: a data preprocessing component, a model invocation component, and an information adjustment component; The data preprocessing component is used to preprocess the mass spectrometry feature signals to obtain the target signal; The model invocation component is used to determine the target model among multiple spatial reconstruction models. The target signal and the location coordinates of the sampling port are input into the target model to obtain the spatial distribution information of the concentration of the target component. The information adjustment component is used to generate display information based on the target indicators in the spatial distribution information of concentration, and then transmit the display information to the target device.

[0011] The mass spectrometry characteristic signal is preprocessed to eliminate sampling errors; the target model is selected as needed to adapt to different cavity conditions; the spatial distribution information of concentration is output and the target index is visualized; the displayed information is uploaded to the target device to support the monitoring and control of etching process uniformity.

[0012] In one optional implementation, the data preprocessing component includes: a preprocessing unit, a first processing unit, a second processing unit, and a calibration unit; The preprocessing unit is used to obtain the length of the sampling pipeline and the average linear velocity of the component sample in the sampling pipeline, determine the correction offset based on the pipeline length and average linear velocity, and adjust the mass spectrometry characteristic signal based on the correction offset to obtain the intermediate signal. The calibration unit is used to determine the transfer coefficient of each sampling port through a step-by-step calibration method. The step-by-step calibration method includes: when the process chamber does not contain plasma, introducing a calibration gas of known concentration into the process chamber, measuring the response signal of each sampling port, and obtaining the molecular state transfer coefficient based on the response signal; when the process chamber contains plasma, determining the free ground state relative sensitivity factor based on the mass spectrometry signal of the internal standard gas and the independent reference measurement method, and obtaining the free ground state transfer coefficient based on the molecular state transfer coefficient and the free ground state relative sensitivity factor. The first processing unit is used to obtain the intensity of the first characteristic peak of the target component based on the intermediate signal when the target component sample does not contain a preset gas, and to obtain the target signal based on the ratio of the intensity of the first characteristic peak to the free ground state transfer coefficient. The second processing unit is used to obtain the first characteristic peak intensity of the target component and the second characteristic peak intensity of the preset gas based on the intermediate signal when the target component sample contains a preset gas, and to obtain the target signal based on the ratio of the first characteristic peak intensity to the second characteristic peak intensity.

[0013] The calibration unit is also used to: introduce calibration gas into the process chamber during process breaks, wherein the calibration gas is used to update the molecular state transfer coefficient.

[0014] In an optional embodiment, the calibration unit is further configured to: introduce a calibration gas into the process chamber during process intermittent periods when the process chamber does not contain plasma, wherein the calibration gas is used to update the molecular state transfer coefficient.

[0015] In one alternative implementation, the system further includes: a control module; The control module is used to determine the current process stage and, based on the current process stage, determine the sampling port set and the order in which the sampling ports in the sampling port set are called. The sampling port set is used to determine the sampling ports that need to acquire the target component sample.

[0016] The control module can automatically identify different process stages and flexibly adjust the sampling ports and acquisition sequence; it can shield sensitive ports when pressure fluctuates and reduce the sampling frequency of easily clogged ports under corrosive conditions, thus balancing the reliability of sampling data with pipeline losses.

[0017] In a second aspect, the present invention provides a method for determining the spatial distribution of the concentration of a target component, the method being applied to a system for determining the spatial distribution of the concentration of a target component in the first aspect or any corresponding embodiment described above, the method comprising: Component samples are acquired in the process chamber through sampling ports, wherein the spacing between sampling ports is greater than a first preset threshold. The spacing between sampling ports is determined according to the type of spatial reconstruction model. The first preset threshold is used to control the sampling areas where the sampling ports are located to not overlap. The target component sample is obtained by adjusting the sampling port through the current limiting element. The type and size of the current limiting element are determined according to the pressure of the target component in the process chamber and the preset rules. The target component sample is transferred to the component analysis module through the sampling pipeline. The length difference between different sampling pipelines is less than a second preset threshold. The second preset threshold is used to control the length difference between the sampling pipelines. The component analysis module generates mass spectrometry characteristic signals based on the target component sample. Based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, the spatial distribution information of the target component concentration is generated.

[0018] In one optional implementation, based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, the concentration spatial distribution information of the target component is generated, including: When the spatial reconstruction model is the first sub-model, the position coordinates are converted to polar coordinates; Based on the mass spectrometry characteristic signals, the radial concentration distribution of the target component is obtained; The radial concentration distribution is interpolated using polar coordinates to obtain the target equation, and the weights corresponding to the sampling ports are obtained based on the target equation. Based on the weights, polar coordinates, and radial concentration distribution, the spatial distribution information of the target component's concentration is obtained.

[0019] The position coordinates are converted to polar coordinates to simplify the calculation, the weights are solved to complete the radial interpolation, and the rotation generates a complete two-dimensional concentration distribution map. The calculation is simple and efficient, and the radial concentration distribution of free radicals in the chamber can be quickly restored.

[0020] In one optional implementation, based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, the concentration spatial distribution information of the target component is generated, including: When the spatial reconstruction model is the second sub-model, the mass spectrometry feature signal is used as the observation value. Based on the observation value, position coordinates and the formula corresponding to the second sub-model, the predicted mean and confidence interval of each sampling point in the process chamber are obtained. The formula corresponding to the second sub-model satisfies:

[0021] in, Let be the position coordinates of any location within the process chamber. The coordinates of another location, , Both represent the x-axis. , Both represent the ordinate. express The target active ingredient concentration at the location; Represent a Gaussian process; Represents the mean function; This represents the covariance function, used to describe any two positions. and Spatial correlation between concentration values; Based on the predicted mean and confidence interval, the spatial distribution information of the target component's concentration is obtained; When the spatial reconstruction model is the third sub-model, the process parameter vector is obtained, and the normalized signal intensity corresponding to the sampling port is obtained based on the mass spectrometry characteristic signal. The input vector is obtained based on the process parameter vector, normalized signal strength, and position coordinates; The input vector is fed into the third sub-model to obtain the spatial distribution information of the target component's concentration.

[0022] In an alternative implementation, before generating the mass spectrometry characteristic signal based on the target component sample, the method further includes: The transfer coefficient of each sampling port is determined by a step-by-step calibration method; The stepwise calibration method includes: in a process chamber without plasma, introducing a calibration gas of known concentration into the process chamber, measuring the response signal at each sampling port, and obtaining the molecular state transfer coefficient based on the response signal; in a process chamber containing plasma, using the mass spectrometry signal of the internal standard gas as an internal scale, combined with independent reference measurement methods to determine the free ground state relative sensitivity factor, and obtaining the free ground state transfer coefficient based on the molecular state transfer coefficient and the free ground state relative sensitivity factor.

[0023] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method for determining the spatial distribution of the concentration of the target component described in the second aspect or any corresponding embodiment.

[0024] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the method for determining the spatial distribution of the concentration of a target component as described in the second aspect or any corresponding embodiment.

[0025] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the method for determining the spatial distribution of the concentration of a target component as described in the second aspect or any corresponding embodiment.

[0026] In this invention, the sampling module acquires a target component sample in the process chamber through sampling ports, a current-limiting element, and sampling tubing. The spacing between sampling ports is greater than a first preset threshold. The type and size of the current-limiting element are determined based on the pressure of the target component in the process chamber and preset rules. The length difference between different sampling tubing is less than a second preset threshold. The component analysis module generates a mass spectrometry characteristic signal based on the target component sample. The spatial distribution determination module generates the spatial distribution information of the target component's concentration based on the mass spectrometry characteristic signal, the position coordinates of the sampling ports, and a spatial reconstruction model. The spatial distribution information of the target component's concentration not only reflects the concentration of the active ingredient at a local point in the chamber but also reveals the spatial distribution of the active ingredient's concentration throughout the entire chamber.

[0027] This system achieves multi-point sampling of the process chamber through multiple sampling ports. Combined with the location coordinates of the sampling ports, it simultaneously performs spatial sensing of the process chamber, enabling time-series sampling and spatial distribution reconstruction of active ingredient concentrations at multiple spatial locations within a single gas composition analyzer. This upgrades from single-point measurement to spatial sensing, providing direct data support for etching uniformity monitoring. Using a single composition analysis module with a shared switching valve assembly to connect to multiple sampling ports significantly reduces costs compared to parallel solutions with multiple analyzers. Utilizing spatial reconstruction models, mass spectrometry characteristic signals, and location coordinates, it generates spatial distribution information of the target component concentration, including the active ingredient concentration at each sampling point within the process chamber. The sampling ports employ a consumable design, resulting in a simple structure and controllable cost. They eliminate the need for expensive anti-corrosion coatings and complex heating systems, facilitating easy retrofitting of existing etching chambers and further reducing deployment and maintenance costs. Differentiated configuration of flow-limiting elements and sampling pipeline structural parameters ensures that the output flow rate of each port falls within the analyzer's linear range and reduces sample transmission delay in the sampling pipeline. The system has a wide range of applications. The spatial reconstruction model can be trained with a small amount of experimental data or pre-trained based on physical simulation and then optimized by transfer, adapting to different machines and process formulations.

[0028] In addition, the spatial distribution information of the target component concentration can be fed back to the actuators of the etching equipment, such as the air intake distribution, radio frequency power distribution, and multi-zone temperature control, to provide control input for the automatic adjustment of wafer-level etching uniformity. Attached Figure Description

[0029] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0030] Figure 1 This is a structural block diagram of a system for determining the spatial distribution of the concentration of a target component according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a multi-point sampling gas analysis system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an integrated multi-point timing injection module according to an embodiment of the present invention; Figure 4 This is a flowchart illustrating a method for determining the spatial distribution of the concentration of a target component according to an embodiment of the present invention. Figure 5 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0033] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0034] According to embodiments of the present invention, a system for determining the spatial distribution of the concentration of a target component is provided. Figure 1 This is a schematic diagram of the structure of a system for determining the spatial distribution of the concentration of a target component according to an embodiment of the present invention, as shown below. Figure 1 As shown, the system includes: a sampling module, a component analysis module, and a spatial distribution determination module; The sampling module includes a preset number of sampling ports, a preset number of current limiting elements, and a preset number of sampling pipelines; the spacing between the sampling ports is greater than a first preset threshold, and the spacing between the sampling ports is determined according to the type of the spatial reconstruction model. The first preset threshold is used to control the sampling areas where the sampling ports are located to not overlap. The sampling module is connected to the component analysis module and is used to acquire component samples in the process chamber through the sampling port; The current limiting element is connected to the sampling port; the sampling module is used to adjust the component sample obtained by the sampling port through the current limiting element to obtain the target component sample, and to transmit the target component sample to the sampling pipeline. The type and size of the current limiting element are determined according to the pressure of the target component in the process chamber and the preset rules. The sampling module is used to transfer the target component sample to the component analysis module through the sampling pipeline. The length difference between different sampling pipelines is less than a second preset threshold, which is used to control the length difference between the sampling pipelines. The component analysis module is connected to the spatial distribution determination module. It is used to generate mass spectrometry characteristic signals based on the target component sample and input the mass spectrometry characteristic signals into the spatial distribution determination module. The spatial distribution determination module is used to generate spatial distribution information of the target component's concentration based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model.

[0035] Specifically, the sampling module, for example Figure 2 The integrated multi-point timing sample introduction module 100 is included. Process chambers, for example, are chambers used for processing wafers using semiconductor etching processes (such as inductively coupled plasma etching and capacitively coupled plasma etching). Figure 2 As shown.

[0036] The sampling module includes a preset number of sampling ports, a preset number of current-limiting elements, and a preset number of sampling pipelines. The sampling ports are located within the process chamber, for example, at the top center, top edge, near the wafer on the sidewall, near the gas injection port, or near the pump port. A first preset threshold is, for example, 100 mm. The spacing between the sampling ports is greater than this threshold to ensure that the sampling areas of different ports do not overlap. The preset number indicates multiple ports, for example, 3-8, preferably 4 or 6. The spacing between the sampling ports is determined based on the type of spatial reconstruction model. For example, if the spatial reconstruction model is a Gaussian process regression model, the spacing between the sampling ports is 100 mm to 120 mm; if the spatial reconstruction model is a neural network model, the spacing between the sampling ports is 130 mm to 150 mm. The correspondence between the type of spatial reconstruction model and the spacing is adjusted according to the process chamber structure, the composition of the target component, or other requirements.

[0037] The sampling module samples active ingredients (such as free radicals like F, Cl, and CF2) within the sampling area located at the sampling port, obtaining a component sample. The sampling module then inputs the target component sample into the component analysis module. The target component is a specific active ingredient.

[0038] In this embodiment, the sampling ports are all fixed ports that are permanently installed in the process chamber and do not undergo displacement relative to the process chamber during system operation. Examples of sampling ports include: Figure 3 The sealing flange base 1011 is included. The sampling module also includes: a current-limiting element, a sampling pipeline, and a heating element, for example: Figure 3 The sampling module includes a current-limiting element 1012, a sampling conduit 1013, and a heating band 1014. Additionally, the sampling module may include a protective screen.

[0039] Types of flow-limiting elements include capillary type, microporous plate type, and microcrack type. The type of flow-limiting element is determined based on the pressure of the target component in the process chamber and preset rules; for example, the type of flow-limiting element is configured according to pressure differences at different locations. The size of the flow-limiting element is also determined based on the pressure of the target component in the process chamber and preset rules; for example, a smaller inner diameter and longer flow-limiting element is used in high-pressure areas, while a larger inner diameter and shorter flow-limiting element is used in low-pressure areas, ensuring that the output flow rate at each port falls within the analyzer's linear range.

[0040] The current limiting element is connected to the sampling port and is used to adjust the component sample acquired by the sampling port to obtain the target component sample and transmit the target component sample to the sampling pipeline.

[0041] In addition, different structural parameters can be configured for different sampling modules, such as at least one of the following: capillary inner diameter, capillary length, sampling orifice diameter, port orientation, sampling tube length, and flow restriction method. These differentiated structural parameters are set based on at least one of the following factors: the installation location of the sampling port, the expected sampling gas pressure, and the type of the target active ingredient.

[0042] The second preset threshold, for example, 20% of the sampling pipeline length, controls the length difference between sampling pipelines in different sampling modules within ±20%, reducing transmission delay. The sampling module is used to transmit the target component sample to the component analysis module through the sampling pipeline. For example, the sampling module also includes a pathway adjustment component and a pathway switching component, which sequentially transmit the target component sample to the component analysis module through the sampling pipeline, the pathway adjustment component, and the pathway switching component.

[0043] The sampling port, current limiting element, and sampling pipeline in the sampling module are simple in structure, cost-effective, and easy to retrofit and install on existing etching chambers. Differentiated configuration of structural parameters for the current limiting element and sampling pipeline ensures that the output flow rate of each port falls within the linear range of the analyzer and reduces sample transmission delay in the sampling pipeline.

[0044] Figure 2 The component analysis module 200 in the middle is, for example, a gas component analysis module. The gas component analysis module is used to receive the target component sample from the sampling module, analyze the target component sample and output the mass spectrometry characteristic signal.

[0045] The component analysis module can adopt a gas analyzer structure (including an ion source, mass analyzer, and detector), without relying on corrosion-resistant coatings or independent heating units as core corrosion protection measures. The low-cost, replaceable design of the sampling port limits corrosion impact to consumable components. The sample inlet interface of the component analysis module is, for example, a CF35 flange with a centrally located 1-2 mm diameter inlet hole (molecular leak), downstream of which is the ionization chamber of the ion source. The ion source of the component analysis module is, for example, an electron impact type, with a double yttrium iridium or tungsten filament, adjustable electron energy of 5-100 eV, and emission current of 0.1-2 mA. The mass analyzer of the component analysis module is, for example, a quadrupole mass analyzer with a rod length of 100-200 mm, a diameter of 6-12 mm, and a mass range of 1-300 amu. The detector of the component analysis module is, for example, a Faraday cup and / or a secondary electron multiplier (SEM).

[0046] The component analysis module can adopt an adaptive operating mode. The control software retrieves preset values ​​of parameters from the configuration library based on the currently connected sampling port ID and automatically adjusts them to optimize the signal-to-noise ratio and protect detector lifespan. Preset parameters include, for example, electron energy (eV), emission current (mA), and detector gain (SEM voltage). Ports with high active component concentrations use lower electron energy (15~25 eV) to reduce fragmentation interference, while ports with low concentrations use higher electron energy (30~70 eV) to improve ionization efficiency. The emission current is automatically adjusted according to the expected concentration at the port to avoid detector saturation due to high concentrations or insufficient signal-to-noise ratio due to low concentrations. The optimal detector gain is automatically selected within the linear range.

[0047] Spatial distribution determination module, for example Figure 2 The data processing and spatial reconstruction module 300 is responsible for converting discrete multi-point sampling data into continuous concentration spatial distribution information.

[0048] Mass spectrometry characteristic signals include mass spectrometry data from each sampling port, such as mass spectrometry data from each sampling port i of the gas composition analysis module, and the characteristic peak intensity of the target active ingredient. (Units are A or counts / s), internal standard gas peak intensity Internal standard gas, such as Ar, is used if the process gas contains Ar or a trace amount of Ar is actively added.

[0049] The location coordinates of the sampling port are as follows: The wafer plane is used as the xy plane, and the z-axis is perpendicular to the wafer. For ports near the wafer surface, the projected coordinates of their sampling positions on the wafer plane are taken as planar coordinates, and their height information is used independently or incorporated into the 3D reconstruction.

[0050] The spatial distribution determination module generates spatial distribution information of the target component's concentration based on the mass spectrometry feature signal, the location coordinates of the sampling port, and the spatial reconstruction model. For example, the spatial reconstruction model uses the radial basis function interpolation method, taking the location coordinates of each sampling port and the concentration value in the corresponding mass spectrometry feature signal as interpolation nodes to generate a continuous spatial distribution map; the spatial reconstruction model uses the Gaussian process regression model, taking the location coordinates of each sampling port as input and the corresponding mass spectrometry feature signal as the observation value, and predicting the concentration value of unsampled locations through the covariance function; the spatial reconstruction model uses a neural network model, taking the mass spectrometry feature signal and the location coordinates of each sampling port as input, and the reconstructed spatial distribution map of free radical concentration as output.

[0051] Concentration spatial distribution information can be categorized as follows: a two-dimensional concentration distribution matrix (size M×M, M≤512), or three-dimensional point cloud or volume data. This spatial distribution information can be transformed into a visualization method using a pseudo-color map (Jet or Hot color map) overlaid with contour lines for easy display.

[0052] This embodiment provides a system for determining the spatial distribution of target component concentration. The sampling module acquires target component samples in a process chamber through sampling ports, current-limiting elements, and sampling pipelines. The spacing between sampling ports is greater than a first preset threshold. The type and size of the current-limiting elements are determined based on the pressure of the target component in the process chamber and preset rules. The length difference between different sampling pipelines is less than a second preset threshold. The component analysis module generates a mass spectrometry characteristic signal based on the target component sample. The spatial distribution determination module generates spatial distribution information of the target component concentration based on the mass spectrometry characteristic signal, the position coordinates of the sampling ports, and a spatial reconstruction model. This system achieves multi-point sampling of the process chamber through multiple sampling ports and, combined with the position coordinates of the sampling ports, achieves spatial perception of the process chamber while sampling. Using the spatial reconstruction model, mass spectrometry characteristic signal, and position coordinates, it generates spatial distribution information of the target component concentration, ensuring that the spatial distribution information includes the concentration of the active ingredient at each sampling point within the process chamber. Furthermore, this system has a wide range of applications and can be adapted to different processes using different spatial reconstruction models. The spatial distribution information of the target component concentration can not only reflect the concentration of the active component at a local point in the chamber, but also reveal the spatial distribution of the active component concentration throughout the entire chamber.

[0053] In some optional implementations, the sampling module further includes: a path adjustment component and a path switching component; the sampling module also includes a heating belt; The pathway adjustment component is used to clean residual sample in the sampling tubing before transmitting the target component sample through the sampling tubing; The pathway switching component is connected to the pathway adjustment component through multiple first pathways. It is used to take the first pathway corresponding to the target sampling pipeline as the target pathway and connect the target pathway to the component analysis module so that the target pathway can transmit the target component sample to the component analysis module. The target sampling pipeline is the sampling pipeline that is transmitting the target component sample. Heating bands are wrapped around the outer wall of the sampling pipeline to adjust the temperature of the target component sample in the sampling pipeline.

[0054] Specifically, the pathway adjustment component, for example Figure 3 The pipeline pre-extraction and replacement unit 102. Path switching components, for example... Figure 3 The shared timing switching valve group 103 is included. The sampling module includes sampling pipelines, for example: Figure 3 Sampling line 1013 in the middle.

[0055] The pathway adjustment component is used to clean residual samples in the sampling pipeline before transmitting the target component sample through the sampling pipeline. For example, the pathway adjustment component includes a pre-pump, an independent solenoid valve, and a flow controller. Before transmitting the target component sample through the sampling pipeline, the pre-pump is used to pre-pump the sampling pipeline to remove residual samples and eliminate cross-interference. Alternatively, the independent solenoid valve and flow controller are used to purge and displace the pipeline, blowing inert gas into the sampling pipeline to displace residual samples and eliminate interference from residual gas from the previous sampling. Alternatively, the pre-pump and purge / displacement operations can be performed alternately or simultaneously.

[0056] Path switching components, for example: Figure 3 The shared timing switching valve group 103 is used in the pipeline. For example, the first passage is a pipeline connected to the three-way switching valve in the pipeline pre-extraction and replacement unit 102 via the shared timing switching valve group 103. The passage switching component is connected to the passage adjustment component via multiple first passages and to the component analysis module via a single passage.

[0057] The pathway switching assembly can employ a multi-channel VCR (Vacuum Coupling Retainer) valve group or a multi-position directional valve to achieve multi-port sequential switching, with only one channel conducting at a time. Parameters for the pathway switching assembly include, for example: port diameter of 3mm or 6mm, internal volume <0.5mL, and switching time ≤500ms.

[0058] A heating tape 1014 is wrapped around the outer wall of the sampling pipeline. The heating tape is used to adjust the temperature of the target component sample in the sampling pipeline. For example, the temperature of the heating tape is set to 80~200℃ to heat the target component sample in the sampling pipeline. This is only used to prevent the condensation and deposition of process by-products and is not used as the main means of corrosion resistance.

[0059] The pathway switching component identifies the target sampling pipeline that is currently transmitting the target component sample. For example, if the three-way switching valve 5 connects the pipeline between port 1 and the shared timing switching valve group 103, then the pipeline between port 1 and the shared timing switching valve group 103 is the target sampling pipeline. The first pathway corresponding to the target sampling pipeline is designated as the target pathway, i.e., the pipeline between the three-way switching valve 5 and the shared timing switching valve group 103 is designated as the target pathway. The pathway switching component connects the target pathway to the component analysis module, enabling the target pathway to transmit the target component sample to the component analysis module.

[0060] In this embodiment, the residual sample in the sampling pipeline is cleaned by the pathway adjustment component to eliminate sampling cross-interference; the pathway switching component controls the opening or closing of different first pathways so that the target component samples in different target sampling pipelines are input into the component analysis module separately. Multi-point time-series sampling can be achieved using only a single component analysis module, which can obtain gas samples from multiple locations at low cost and ensure that the detection data is accurate and reliable.

[0061] In some alternative implementations, the pathway adjustment components include: a three-way valve and a pre-pump; The first port of the three-way valve is connected to the sampling pipeline, the second port of the three-way valve is connected to the pre-pump, and the third port of the three-way valve is connected to the first passage. In the three-way valve, the first port and the second port are connected through the second passage, and the first port and the third port are connected through the third passage. The pathway adjustment component is used to close the third pathway corresponding to the candidate sampling pipeline and open the second pathway corresponding to the candidate sampling pipeline before transmitting the target component sample through the candidate sampling pipeline. The pathway adjustment component is also used to extract residual samples in the candidate sampling pipeline using a pre-pump, and to continue for a preset duration. The pathway adjustment component is also used to close the second pathway corresponding to the candidate sampling pipeline and open the third pathway corresponding to the candidate sampling pipeline after a preset time.

[0062] Specifically, the flow adjustment assembly includes a three-way valve and a pre-pump, the three-way valve being a three-way switching valve 1021, for example: Figure 3 Three-way switching valves 5, 6, 7, and 8 are included.

[0063] The three-way valve has three ports: a C (Common) port, an NC (Normally Closed) port, and a NO (Normally Open) port. The first port is connected to the sampling line, for example: Figure 3The C port of the three-way switching valve 5 is connected to the sampling pipeline 1013. The third port of the three-way valve is connected to the first passage, for example, the first passage is the pipeline between the three-way switching valve 5 and the shared timing switching valve group 103, and the NO port of the three-way switching valve 5 is connected to the first passage. The second port of the three-way valve is connected to the pre-pump, for example, the three-way switching valve 5 is connected to the pre-pump through the pre-pump main pipe 1022.

[0064] In the three-way valve, the first port and the second port are connected via a second passage, which is the C→NO passage of the three-way valve. The first port and the third port are connected via a third passage, which is the C→NC passage of the three-way valve.

[0065] After sampling is completed on one port and before switching to the next port, a pre-sampling operation needs to be performed. This embodiment is used to illustrate the timing of the pre-sampling operation.

[0066] The current port sampling is complete, and the analysis module finishes data recording. Before transmitting the target component sample through the candidate sampling pipeline (such as the sampling pipeline corresponding to port i), the C→NO path corresponding to the candidate sampling pipeline is closed using the pathway adjustment component, while the C→NC path corresponding to the candidate sampling pipeline is opened simultaneously, or the operation can be performed sequentially: first close C→NO, then open C→NC. This connects the section of the sampling pipeline from port i to the three-way valve to the pre-pump.

[0067] The pathway adjustment component uses a pre-pump to extract residual samples from the candidate sampling tubing and continues for a preset duration. . The value range is 2 to 10 seconds. The pre-evacuation time is determined by the length and inner diameter of the sampling pipeline. The principle is to remove at least 3 times the volume of gas in the pipeline.

[0068] After a preset time, i.e. after the pre-pump has been running for a preset time, the channel adjustment component closes the C→NC channel corresponding to the candidate sampling pipeline and opens the C→NO channel corresponding to the candidate sampling pipeline, or keeps the C→NO channel closed and waits for the next port to be connected.

[0069] Additionally, during pre-purging, purging and displacement operations can be performed alternately or simultaneously. For example, an inert purging gas (such as Ar or N2) can be introduced into the candidate sampling line to accelerate displacement. The purging gas is introduced from the NC port of the three-way valve through a separate solenoid valve and mass flow controller (MFC). The purging flow rate is set to 5-20 sccm, the duration is 1-3 seconds, and it can be performed alternately or simultaneously with pre-purging (but the pre-purging pump capacity must be sufficient).

[0070] In this embodiment, the candidate sampling pipeline is first connected to the pre-sampling pump via a three-way valve. The pre-sampling pump is then used to perform a pre-sampling operation to empty the sample remaining in the candidate sampling pipeline, thereby eliminating cross-contamination between different sampling channels and ensuring accurate and comparable multi-point sampling and detection data.

[0071] In some optional implementations, the spatial distribution determination module includes: a data preprocessing component, a model invocation component, and an information adjustment component; The data preprocessing component is used to preprocess the mass spectrometry feature signals to obtain the target signal; The model invocation component is used to determine the target model among multiple spatial reconstruction models. The target signal and the location coordinates of the sampling port are input into the target model to obtain the spatial distribution information of the concentration of the target component. The information adjustment component is used to generate display information based on the target indicators in the spatial distribution information of concentration, and then transmit the display information to the target device.

[0072] Specifically, preprocessing includes, for example, performing transmission delay correction on the mass spectrometry feature signal, followed by response coefficient normalization or internal standard normalization. The data preprocessing component preprocesses the mass spectrometry feature signal to obtain the target signal.

[0073] Input data for the spatial reconstruction model includes, for example, mass spectrometry data from each sampling port in the mass spectrometry characteristic signal, the position coordinates of each sampling port, and may also include the current process parameter vector. Mass spectrometry data may include, for example, the characteristic peak intensity of the target active ingredient. Internal standard gas peak intensity Current process parameter vectors, for example: chamber pressure. (Pa), radio frequency power (W) Bias power (W), flow rates of various process gases (sccm), chamber temperature (°C), the current process parameter vector is used as a conditional input or normalization aid for the model.

[0074] Spatial reconstruction models include: Model A, which performs 2D reconstruction based on Radial Basis Function (RBF) interpolation; Model B, which performs 2D / 3D reconstruction based on Gaussian Process Regression (GPR); and Model C, which performs end-to-end image reconstruction based on Convolutional Neural Network (CNN). Model A is applicable under conditions such as: sampling ports are distributed on the top or sidewalls of the chamber, and the active ingredient concentration is assumed to have axial symmetry in the azimuth direction (i.e., concentration varies only with radius). Model B is applicable under conditions such as: arbitrary sampling point layout, independent of the axial symmetry assumption. Model C is applicable under conditions such as: using Model C when Models A and B are not used.

[0075] Based on the applicable conditions of the aforementioned models, the model invocation component determines the spatial reconstruction model that satisfies the current process cavity scenario from multiple spatial reconstruction models and uses it as the target model. The target signal and the position coordinates of the sampling port are input into the target model to obtain the spatial distribution information of the target component's concentration.

[0076] Concentration spatial distribution information, such as: two-dimensional concentration distribution matrix (size M×M, M≤512), three-dimensional point cloud or volume data. Target indicators, such as: minimum value, maximum value, average value, and uniformity indicators (e.g., standard deviation / average, range / median).

[0077] The information adjustment component generates display information based on the target indicators in the spatial distribution information of concentration. The display information is a spatial distribution map, such as a spatial distribution map generated by overlaying a pseudo-color map (Jet or Hot color map) with contour lines. The display information is used to show the aforementioned target indicators.

[0078] Target devices include, for example, factory mainframes or etching equipment controllers. Display information is transmitted to the target device, for example, via standard industrial communication protocols. The output of the spatial distribution map is coupled to the main controller of the process equipment, used to feed back and control the process parameters of the equipment based on the uniformity evaluation indicators of the spatial distribution map. The display information is transmitted to the target device, providing control input for the automatic adjustment of wafer-level etching uniformity.

[0079] In addition, the system can compare the spatial distribution map with a set uniformity index threshold. If the threshold is exceeded, a feedback control command is sent to the main controller of the process equipment. The feedback control command is selected from at least one of adjusting the intake air flow distribution, adjusting the radio frequency power distribution, and adjusting the multi-zone temperature control setpoint.

[0080] In this embodiment, the mass spectrometry feature signal is preprocessed to eliminate sampling errors; a target model is selected as needed to adapt to different cavity operating conditions; the spatial distribution information of concentration is output and the target index is visualized; the displayed information is uploaded to the target device to support the monitoring and control of etching process uniformity.

[0081] In some optional implementations, the data preprocessing component includes: a preprocessing unit, a first processing unit, a second processing unit, and a calibration unit; the preprocessing unit is used to obtain the tube length of the sampling tube and the average linear velocity of the component sample in the sampling tube, determine the correction offset based on the tube length and the average linear velocity, and adjust the mass spectrometry characteristic signal based on the correction offset to obtain an intermediate signal. The calibration unit is used to determine the transfer coefficient of each sampling port through a step-by-step calibration method. The step-by-step calibration method includes: when the process chamber does not contain plasma, introducing a calibration gas of known concentration into the process chamber, measuring the response signal of each sampling port, and obtaining the molecular state transfer coefficient based on the response signal; when the process chamber contains plasma, determining the free ground state relative sensitivity factor based on the mass spectrometry signal of the internal standard gas and the independent reference measurement method, and obtaining the free ground state transfer coefficient based on the molecular state transfer coefficient and the free ground state relative sensitivity factor. The first processing unit is used to obtain the intensity of the first characteristic peak of the target component based on the intermediate signal when the target component sample does not contain a preset gas, and to obtain the target signal based on the ratio of the intensity of the first characteristic peak to the free ground state transfer coefficient. The second processing unit is used to obtain the first characteristic peak intensity of the target component and the second characteristic peak intensity of the preset gas based on the intermediate signal when the target component sample contains a preset gas, and to obtain the target signal based on the ratio of the first characteristic peak intensity to the second characteristic peak intensity.

[0082] Specifically, the preprocessing unit obtains the length of the sampling pipeline ( ) and the average linear velocity of the component sample in the sampling pipeline ( Average linear velocity can be measured by the pulse method or estimated based on pressure difference.

[0083] The correction offset is determined based on the pipe length and average linear velocity. ),For example: .

[0084] For sampling pipelines of different lengths, the signal arrival time at the component analysis module varies. The mass spectrometry characteristic signal is adjusted based on the correction offset to obtain an intermediate signal. Transmission delay correction is then applied to the mass spectrometry characteristic signal to align the acquisition times of the sampling data from each port in terms of timing.

[0085] The calibration unit determines the transfer coefficient of each sampling port through a step-by-step calibration method, which includes two stages: molecular state transfer coefficient calibration and determination of the relative sensitivity factor of the free ground state. During the molecular state transfer coefficient calibration stage: A known partial pressure is introduced into the process chamber without plasma present. A calibration gas (selected according to the process, such as N2, Ar, or a stable molecular form of the target gas, such as Cl2, CF4) is used to achieve a uniform concentration distribution throughout the chamber; the calibration gas is used to update the molecular state transfer coefficient. Each sampling port is sequentially connected to the component analysis module via a shared switching valve group, and the ion current intensity at each port at the target mass-to-charge ratio is recorded. And remove background signal Molecular state transfer coefficient at port i Calculate using the following formula (1).

[0086] (1) Multi-point calibration can be performed using 2 to 3 standard gases of different concentrations to establish a signal-concentration standard curve I = a•P + b, where I represents the signal, P represents the concentration, the slope a is the transfer coefficient, and the correlation index R of the standard curve is... 2 It should be greater than 0.99.

[0087] In the stage of determining the relative sensitivity factor of the free ground state: Under plasma conditions within the process chamber, an inert gas (such as Ar or Kr) that is absent in the process gas and does not react with the target free radical is selected as the internal standard gas and continuously introduced into the chamber at a known flow rate. The internal standard gas is a mixed standard gas with known composition, concentration, and leakage rate, used for calibrating its own signal. The relative sensitivity factor of the free ground state is determined based on the mass spectrometry signal of the internal standard gas and an independent reference measurement method. Under the conditions of plasma activation and free radical generation, the intensity of the characteristic peak signal of the free radical is simultaneously recorded. and internal standard gas signal strength The absolute concentration of free radicals at the sampling point is obtained using an independent reference measurement method (selected from at least one of titration, laser-induced fluorescence spectroscopy, cavity ring-down spectroscopy, or electron spin resonance). The relative sensitivity factor of the free ground state relative to the internal standard gas. Calculate according to formula (2).

[0088] (2) in, The known concentration of the internal standard gas; the transfer coefficient of the target radical to port i. It is obtained by multiplying the molecular state transfer coefficient by the relative sensitivity factor of the free ground state, as shown in formula (3).

[0089] (3) This transfer coefficient is used to represent the free radical signal intensity measured in actual detection. Converted to free radical concentration .

[0090] The first processing unit obtains the preset transfer coefficients, for example: defining the transfer function of port i as follows: ,in, The actual concentration (i.e., known partial pressure) of the standard gas at port i is given. ), The measured concentration of a known concentration standard gas at port i (i.e.) ), and All results were obtained through calibration experiments. A standard gas of known concentration was introduced into the chamber to ensure uniform concentration throughout the chamber, and the response at each port was measured. and According to the transfer function, and Calculate the preset transfer coefficient Repeated calibration during process intervals can update this coefficient, compensating for drift caused by pipeline aging and contamination. Preset transfer coefficient. Similar in essence to the molecular state transfer coefficient, both represent the "proportional relationship between the response signal and a standard gas of known concentration." When calibrated using molecular state standard gas This is the molecular state transfer coefficient.

[0091] The preset gas is an internal standard gas, such as Ar. When the target component sample does not contain the preset gas, the first processing unit obtains the intensity of the first characteristic peak of the target component based on the intermediate signal. The target signal is obtained based on the ratio of the intensity of the first characteristic peak to the preset transmission coefficient. The target signal is the first corrected signal. ),For example: .

[0092] When the target component sample contains a preset gas, the second processing unit obtains the intensity of the first characteristic peak of the target component based on the intermediate signal. ) and the intensity of the second characteristic peak of the preset gas ( The target signal is obtained based on the ratio of the intensity of the first characteristic peak to the intensity of the second characteristic peak. The target signal is the second corrected signal. ),For example: This is to eliminate signal fluctuations caused by fluctuations in total chamber pressure or unstable sampling flow rate.

[0093] In some alternative embodiments, the calibration unit is also used to: introduce calibration gas into the process chamber during process intermittent periods when the process chamber does not contain plasma, wherein the calibration gas is used to update the molecular state transfer coefficient to compensate for transfer coefficient drift caused by pipeline aging and contamination.

[0094] Specifically, during process intermittent phases, the process chamber does not contain plasma. During these intermittent phases, such as wafer transfer, chamber purging, and idle standby, the plasma is extinguished. The process intermittent phase is the optimal window for introducing calibration gas because the absence of plasma and reaction byproducts ensures uniform distribution and stable composition of the calibration gas throughout the chamber, resulting in accurate molecular state transfer coefficients. If calibration gas is introduced when the plasma is on, the plasma will dissociate or ionize the calibration gas molecules, altering the composition and generating significant background noise and fragmentation peaks, leading to complete calibration distortion.

[0095] The process of updating the molecular state transfer coefficient by calibrating the gas can be divided into two steps: Step 1: Calibration Experiment to Obtain Known Data. Introduce one or more calibration gases of precisely known concentrations into the instrument. Under steady-state conditions, record the signal intensity (e.g., ion current intensity) produced by the mass spectrometer for each calibration gas.

[0096] The second step involves calculation and updating to obtain the new molecular state transfer coefficient. The new molecular state transfer coefficient ( The transfer coefficient is calculated as follows: (I) = Instrument response signal (I) / Standard gas concentration (C), where the standard gas concentration can be a partial pressure, concentration percentage, etc. Alternatively, if only one concentration point is used, the coefficient can be directly calculated using the above formula. If multiple concentration points are used, a standard curve is fitted, and the slope of the standard curve is the transfer coefficient of the gas. A more preferred technical solution is to use multiple concentration points.

[0097] In some alternative implementations, the system further includes: a control module; The control module is used to determine the current process stage and, based on the current process stage, determine the sampling port set and the order in which the sampling ports in the sampling port set are called. The sampling port set is used to determine the sampling ports that need to acquire the target component sample.

[0098] Specifically, the control module is the module that runs the system control software, such as: Figure 2 The system control software 400 is included.

[0099] The overall control flow of the control module includes: the main controller of the control system performs periodic round-robin sampling, such as sampling → acquisition → pre-sampling → switching → next sampling port.

[0100] Process stages include, for example, the stabilization etching stage and the pressure transient stage (such as the start or end of chamber cleaning). The current process stage is determined, for example, by information from the host computer or by automatic identification of the current process stage based on pressure and power signals.

[0101] The control module dynamically adjusts the sampling port set and the order in which the sampling ports are called according to the current process stage. For example, during the stable etching stage, all sampling ports are included in the sampling port set, and the calling order is fixed, such as according to the port number from smallest to largest. During the pressure transient stage, the ports least affected by the pressure change are included in the sampling port set, and the high-pressure ports are temporarily blocked. The calling order is fixed, such as according to the port number from largest to smallest.

[0102] In addition, in processes with severe corrosion / deposition, the sampling frequency of easily clogged ports is reduced, while the sampling frequency of reference ports is increased to interpolate and fill in the missing points.

[0103] In this embodiment, the control module can automatically identify different process stages and flexibly adjust the sampling ports and acquisition sequence; it can shield sensitive ports when pressure fluctuates and reduce the sampling frequency of easily clogged ports under corrosive conditions, thus balancing the reliability of sampling data with pipeline losses.

[0104] According to an embodiment of the present invention, a method for determining the spatial distribution of the concentration of a target component is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0105] This embodiment provides a method for determining the spatial distribution of the concentration of a target component, which can be used in the aforementioned system for determining the spatial distribution of the concentration of a target component. Figure 4 This is a flowchart of a method for determining the spatial distribution of the concentration of a target component according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps: Step S401: Obtain component samples in the process chamber through sampling ports, wherein the spacing between sampling ports is greater than a first preset threshold, the spacing between sampling ports is determined according to the type of spatial reconstruction model, and the first preset threshold is used to control the sampling areas where the sampling ports are located to not overlap.

[0106] Specifically, the sampling module includes a preset number of sampling ports, a preset number of current-limiting elements, and a preset number of sampling pipelines. The sampling ports are located in the process chamber, for example: the top center, top edge, sidewall near the wafer, near the gas injection port, or near the pump port of the process chamber. The first preset threshold is, for example, 100 mm. The spacing between the sampling ports is greater than the first preset threshold to ensure that the sampling areas of different sampling ports do not overlap. The preset number indicates multiple ports, for example: 3-8, preferably 4 or 6.

[0107] By sampling through the sampling port, active ingredients such as fluorine and chlorine groups (such as free radicals such as F, Cl, and CF2) are sampled in the sampling area where the sampling port is located, and component samples are obtained.

[0108] Step S402: Adjust the component sample obtained by the sampling port through the current limiting element to obtain the target component sample. The type and size of the current limiting element are determined according to the pressure of the target component in the process chamber and the preset rules.

[0109] Specifically, the types of flow-limiting elements include capillary type, microporous plate type, and microcrack type. The type of flow-limiting element is determined based on the pressure of the target component in the process chamber and preset rules; for example, the type of flow-limiting element is configured according to the pressure difference at each location. The size of the flow-limiting element is determined based on the pressure of the target component in the process chamber and preset rules; for example, a smaller inner diameter and longer flow-limiting element is used in the high-pressure area, while a larger inner diameter / shorter flow-limiting element is used in the low-pressure area, so that the output flow rate at each port falls within the linear range of the analyzer.

[0110] The current limiting element is connected to the sampling port and is used to adjust the component sample acquired by the sampling port to obtain the target component sample and transmit the target component sample to the sampling pipeline.

[0111] Step S403: The target component sample is transferred to the component analysis module through the sampling pipeline, wherein the length difference between different sampling pipelines is less than a second preset threshold, and the second preset threshold is used to control the length difference between the sampling pipelines.

[0112] Specifically, the second preset threshold, for example, 20% of the sampling tube length, controls the length difference between sampling tubes in different sampling modules within ±20%, reducing transmission delay. The sampling module is used to transmit the target component sample to the component analysis module through the sampling tube. For example, the sampling module also includes a pathway adjustment component and a pathway switching component, which sequentially transmit the target component sample to the component analysis module through the sampling tube, the pathway adjustment component, and the pathway switching component.

[0113] The sampling port, current limiting element, and sampling pipeline in the sampling module are simple in structure, cost-effective, and easy to retrofit and install on existing etching chambers. Differentiated configuration of structural parameters for the current limiting element and sampling pipeline ensures that the output flow rate of each port falls within the linear range of the analyzer and reduces sample transmission delay in the sampling pipeline.

[0114] Step S404: The component analysis module generates a mass spectrometry characteristic signal based on the target component sample.

[0115] Specifically, the component analysis module can adopt a gas analyzer structure (including an ion source, a mass analyzer, and a detector) to receive target component samples from the sampling module, analyze the target component samples, and output mass spectrometry characteristic signals.

[0116] Step S405: Based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, generate the spatial distribution information of the target component's concentration.

[0117] Specifically, the spatial distribution determination module generates spatial distribution information of the target component's concentration based on the mass spectrometry feature signal, the location coordinates of the sampling port, and the spatial reconstruction model. For example, the spatial reconstruction model uses the radial basis function interpolation method, taking the location coordinates of each sampling port and the concentration value in its corresponding mass spectrometry feature signal as interpolation nodes to generate a continuous spatial distribution map; the spatial reconstruction model uses a Gaussian process regression model, taking the location coordinates of each sampling port as input and the corresponding mass spectrometry feature signal as the observation value, and predicting the concentration value at unsampled locations through the covariance function; the spatial reconstruction model uses a neural network model, taking the mass spectrometry feature signal and the location coordinates of each sampling port as input, and the reconstructed spatial distribution map of free radical concentration as output.

[0118] The method for determining the spatial distribution of target component concentration provided in this embodiment achieves multi-point sampling of the process chamber through multiple sampling ports, and combines the position coordinates of the sampling ports to achieve spatial sensing of the process chamber during sampling. Using a spatial reconstruction model, mass spectrometry characteristic signals, and position coordinates, spatial distribution information of the target component concentration is generated, ensuring that the spatial distribution information includes the concentration of active ingredient at each sampling point within the process chamber. Furthermore, this system has a wide range of applications and can be adapted to different processes using different spatial reconstruction models. It solves the problems that component concentration analysis results only reflect the concentration of active ingredient in a certain area of ​​the chamber, lacking spatial information of the active ingredient, and that the acquisition signals from different sampling points are not comparable.

[0119] In some optional implementations, based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, the spatial distribution information of the target component's concentration is generated, including: When the spatial reconstruction model is the first sub-model, the position coordinates are converted to polar coordinates; Based on the mass spectrometry characteristic signals, the radial concentration distribution of the target component is obtained; The radial concentration distribution is interpolated using polar coordinates to obtain the target equation, and the weights corresponding to the sampling ports are obtained based on the target equation. Based on the weights, polar coordinates, and radial concentration distribution, the spatial distribution information of the target component's concentration is obtained.

[0120] Specifically, the spatial reconstruction model is used to convert the corrected signal from N discrete points (N≥3). (or This is mapped to the concentration spatial distribution information of the entire wafer surface (or a two-dimensional plane / three-dimensional space inside the cavity), such as: active ingredient concentration distribution map.

[0121] The first sub-model is, for example, model A, which is a two-dimensional reconstruction based on radial basis function interpolation.

[0122] When the spatial reconstruction model is the first sub-model, the position coordinates are converted to polar coordinates, for example: the position coordinates of each sampling port are converted to polar coordinates. Convert to polar coordinates If it is confirmed to be an axisymmetric distribution, then Ignore, keep only .

[0123] Preset interpolation, for example, is based on radial basis function interpolation. Based on the mass spectrometry characteristic signal, the radial concentration distribution of the target component is obtained. The radial concentration distribution is then subjected to preset interpolation processing based on polar coordinates to obtain the target equation, for example, as shown in formula (4).

[0124] (4) in, The radial basis function is such as a Gaussian function or a multiple quadratic function, where σ is a shape parameter, taken as 0.3 to 0.5 times the maximum sampling interval. An example of a Gaussian function is Equation (5), and an example of a multiple quadratic function is Equation (6).

[0125] (5) (6) Where σ is a shape parameter, taken as 0.3 to 0.5 times the maximum sampling interval. Indicates the point to be predicted With the i-th known sampling point The distance between them, i.e. .

[0126] The weights corresponding to the sampling ports are determined by solving the objective equation. This ensures that the interpolation curve satisfies the following conditions at the known points. For example: Given the position coordinates of N sampling ports and their corresponding corrected signal values, convert them to polar coordinates to obtain the radial sampling points. ,in This represents the concentration observation at that point. The basic idea of ​​radial basis function interpolation is to approximate the unknown radial concentration distribution function using a weighted superposition of N radial basis functions. , as in formula (7).

[0127] (7) in, Let J be the radial coordinate of the j-th sampling point. Let j be the weight coefficient to be determined. The selected radial basis function (such as the Gaussian function). This indicates the distance from the point to be predicted (r) to the known sampling point. The distance.

[0128] To determine N unknown weight coefficients The interpolation function is required to accurately reproduce the observed value at every known sampling point, i.e., satisfy the interpolation condition: , i=1,2,...,N.

[0129] Will Substituting the expression into formula (7), we obtain a system of linear equations, which can then be written in matrix form: ,in, Given an N×N matrix, the element in the i-th row and j-th column is... ; Let be the weight vector to be determined. The weight vector to be determined is a column vector containing the weight coefficients of all basis functions, i.e. Each w corresponds to an unknown weight of a known sampling point, and what needs to be solved is this set of unknown weights. This is an observation vector, which is a column vector containing the concentration values ​​at all known sampling points. Where each y represents the corrected signal strength of one sampling port, i.e. .

[0130] Due to the radial distance of each sampling point They are all different, and their radial basis functions are... It has positive definiteness, and the coefficient matrix is... Since the matrix is ​​non-singular (i.e., it has an inverse), the system of linear equations has a unique solution, and the weight vector can be obtained through... Solve directly, or use standard numerical methods such as Gaussian elimination or LU (Lower Upper) decomposition.

[0131] Based on the aforementioned weights, polar coordinates, and radial concentration distribution, a two-dimensional distribution map is generated through axisymmetric rotation, and this two-dimensional distribution map serves as the spatial distribution information of the target component's concentration.

[0132] In this embodiment, the position coordinates are converted to polar coordinates to simplify the calculation, the weights are solved to complete the radial interpolation, and the rotation generates a complete two-dimensional concentration distribution map. The calculation is simple and efficient, and the radial concentration distribution of free radicals in the chamber can be quickly restored.

[0133] In some optional implementations, based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, the spatial distribution information of the target component's concentration is generated, including: When the spatial reconstruction model is the second sub-model, the mass spectrometry feature signal is used as the observation value. Based on the observation value, position coordinates and the formula corresponding to the second sub-model, the predicted mean and confidence interval of each sampling point in the process chamber are obtained. Based on the predicted mean and confidence interval, the spatial distribution information of the target component's concentration is obtained; When the spatial reconstruction model is the third sub-model, the process parameter vector is obtained, and the normalized signal intensity corresponding to the sampling port is obtained based on the mass spectrometry characteristic signal. The input vector is obtained based on the process parameter vector, normalized signal strength, and position coordinates; The input vector is fed into the third sub-model to obtain the spatial distribution information of the target component's concentration.

[0134] Specifically, the second sub-model is, for example, Model B, which performs two-dimensional / three-dimensional reconstruction based on Gaussian process regression. The third sub-model is, for example, Model C, which performs end-to-end image reconstruction based on convolutional neural networks (CNN).

[0135] The formula corresponding to the second sub-model, for example: Formula (8), is used to define the second sub-model.

[0136] (8) in, These are the coordinates of point A. These are the coordinates of "point B". "Point A" represents any location within the process chamber, and "point B" represents another location different from "point A". The apostrophe (...) () is used to distinguish two different points in mathematics, and does not represent differentiation or other special operations. and They jointly define the geometric distance between any two points in space, which is the core basis for Gaussian process regression models to measure the spatial correlation of concentration. For position The target concentration of the active ingredient at the location (predicted target); GP represents a Gaussian process; It means to obey, that is Obey ; The mean function can be set as a constant or a linear function of process parameters (such as chamber pressure, RF power, gas flow rate, etc.), representing the baseline trend of concentration in space. When changes in process conditions cause a shift in the overall concentration level, the mean function can absorb this shift, allowing the kernel function to focus on characterizing the spatial correlation structure of concentration. The covariance function (kernel function) represents any two positions and Spatial correlation between concentration values. The covariance function uses a squared exponential kernel.

[0137] (9) in, The signal variance controls the magnitude by which the predicted concentration value deviates from the mean (i.e., the degree of spatial variation in concentration). The higher the value, the more significant the spatial fluctuation of concentration within the chamber; , The l-value serves as the correlation length scale, controlling the smoothness of concentration changes along both directions. A larger l-value indicates smoother concentration changes and stronger correlations between distant sampling points; a smaller l-value indicates more drastic concentration changes and a rapid decay of correlation with distance. Allowed To accommodate possible anisotropic distributions within the chamber (e.g., concentration changes are more gradual in the airflow direction than in the vertical direction). The noise variance represents the random noise introduced during the measurement process (including instrument electronic noise, sampling fluctuations, etc.). This is the Kronecker function, which only takes effect when the two points being calculated are at the same sampling port. It is 1 if and only if the two positions are at the same sampling port, otherwise it is 0.

[0138] Before using the second sub-model to obtain the spatial distribution information of the target component's concentration, the second sub-model needs to be trained. For example, the hyperparameters of the second sub-model can be optimized (maximum likelihood estimation) using historical process data (including multi-point sampling values ​​under multiple different process parameters).

[0139] When the spatial reconstruction model is the second sub-model, the mass spectrometry feature signal is used as the observation value, based on the observation value and the location coordinates. Multiple sets of sampled data are obtained. The sampled data are input into the second sub-model. For a new set of sampled data, the second sub-model calculates the predicted mean value corresponding to each sampling point in the process chamber using the Gaussian process regression posterior formula (such as formula (8) and formula (9) above). And confidence intervals. Based on the predicted mean and confidence intervals, the spatial distribution information of the target component's concentration is obtained.

[0140] The second sub-model undergoes hyperparameter optimization using numerical optimization algorithms such as the conjugate gradient method, maximizing the logarithmic marginal likelihood through multiple iterations. This training process is completed during the system calibration phase, using 10-20 sets of historical multi-point sampling data under different process conditions as the training set. After training, no real-time hyperparameter optimization is required during actual operation; only subsequent prediction calculations are performed, meeting real-time requirements.

[0141] After completing the hyperparameter optimization, for any location to be predicted, such as 128×128 grid points on the wafer surface, the concentration prediction value is given by the conditional distribution of the Gaussian process.

[0142] set up The coordinate matrix formed by all the locations to be predicted. This corresponds to the concentration value to be predicted. (Observed value) Compared with the predicted value The joint distribution is shown in formula (10).

[0143] (10) in, This represents the number of observation points, i.e., the number of known sampling points used to construct the Gaussian process regression model. K = k(X, X), is the N×N observation point kernel matrix; ,for The cross kernel matrix of the observation point and the prediction point, its i-th row and i-th... Column elements are ; for Matrix, representing the prediction point kernel matrix; m, This represents the mean function value at the corresponding location. X represents a set (matrix) of coordinates for multiple points. For example, X is an N×2 matrix that stores the coordinates of all sampling ports. yes A matrix that stores the coordinates of all predicted points. Represents the coordinates of a specific point. In the formula... This describes how the kernel function value of the element in the i-th row and j-th column is calculated. Therefore, K = k(X, X) represents performing a kernel operation on the entire coordinate matrix, resulting in a large N×N matrix; while It involves performing a kernel operation on two specific points to obtain a value from the aforementioned large matrix.

[0144] Given the observed data, the predicted value is... It follows a Gaussian distribution, as shown in formula (11). The predicted mean is shown in formula (12).

[0145] (11) (12) In formulas (11) and (12), the predicted value equals the prior mean plus a weighted combination of the actual deviations of each observation point (observed value minus prior mean), with the weights determined by... The decision is made that the closer the observation point is to the point to be predicted, the greater its weight.

[0146] The confidence interval is shown in formula (13).

[0147] (13) In this context, prediction uncertainty equals prior uncertainty minus the information gain from the observed data. The closer a location is to the observation point, the smaller the prediction variance (higher confidence); conversely, the farther a location is from the observation point, the larger the prediction variance (lower confidence).

[0148] Before using the third sub-model to obtain the spatial distribution information of the target component's concentration, the third sub-model needs to be trained. For example, it can be supervised learning using a large amount of simulation or experimental data (e.g., dense distributions provided by computational fluid dynamics simulations or multi-channel optical emission spectroscopy scans as labels). The network structure of the third sub-model is as follows: First level, a fully connected layer takes the input vector V... in The first stage expands the feature map to 1024 dimensions; the second stage reshapes the one-dimensional feature map into a 32×32×1 feature map; the third stage uses several convolutional and upsampling layers (U-Net or similar structures) to gradually enlarge the map to 128×128×1; the fourth stage uses a linear activation function for the output layer.

[0149] When the spatial reconstruction model is a third sub-model, obtain the process parameter vector, such as: chamber pressure. RF power The normalized signal intensity corresponding to the sampling port is obtained based on the mass spectrometry characteristic signal. Arrange them into a 1×N vector in a fixed order. Based on the process parameter vector, normalized signal strength, and position coordinates, obtain the input vector ( ).

[0150] input vector The third sub-model is input, and it outputs a two-dimensional grid image with a resolution of, for example, 128×128. Each pixel in the grid image represents the predicted concentration of the active ingredient. This two-dimensional grid image serves as the spatial distribution information of the target ingredient's concentration.

[0151] In some alternative implementations, the method further includes, before generating the mass spectrometry characteristic signal based on the target component sample: The transfer coefficient of each sampling port is determined by a step-by-step calibration method; The stepwise calibration method includes: in a process chamber without plasma, introducing a calibration gas of known concentration into the process chamber, measuring the response signal at each sampling port, and obtaining the molecular state transfer coefficient based on the response signal; in a process chamber containing plasma, using the mass spectrometry signal of the internal standard gas as an internal scale, combined with independent reference measurement methods to determine the free ground state relative sensitivity factor, and obtaining the free ground state transfer coefficient based on the molecular state transfer coefficient and the free ground state relative sensitivity factor.

[0152] Specifically, the detailed process of the step-by-step calibration method can be found in the above embodiments, and will not be repeated here.

[0153] In some optional implementations, taking the detection of the spatial distribution of chlorine-active component (Cl) concentration in a Cl2 / BCl3 plasma etching process on a commercial 300mm silicon etching machine as an example, the workflow of the system for determining the spatial distribution of target component concentration is explained: Step S1: Parameter configuration and calibration.

[0154] Specifically, with the target component being pure Cl2, the pressure in the process chamber was stabilized at 1 Pa for calibration. The signal responses of each sampling port were collected, and the transfer coefficients were calculated. For example, the transfer coefficient for sampling port A was 1.00, the transfer coefficient for sampling port B was 0.95, the transfer coefficient for sampling port C was 0.72, and the transfer coefficient for sampling port D was 0.88.

[0155] Step S2: Multi-point time-series sampling and round-robin.

[0156] Specifically, the switching valve group selects sampling port A → the three-way valve of sampling port A cuts off the pre-evacuation circuit → the analysis module acquires data for 5 seconds and records the signal strength of the target component; the three-way valve of sampling port A switches to the pre-evacuation circuit and pre-evacuates for 3 seconds; the switching valve group selects sampling port B → the three-way valve of sampling port B cuts off the pre-evacuation circuit → the analysis module acquires data for 5 seconds and records the signal strength of the target component as 22100 counts / s; the three-way valve of sampling port B switches to the pre-evacuation circuit and pre-evacuates for 3 seconds; and so on to complete the sampling of sampling ports C and D; the total time for one complete cycle = 4 × (5s acquisition + 3s pre-evacuation + 0.5s switching) = 34 seconds.

[0157] Step S3: Data normalization preprocessing yields the corrected signal.

[0158] Specifically, the corrected signal corresponding to sampling port A is, for example, 25800 / 1.00=25800; the corrected signal corresponding to sampling port B is, for example, 22100 / 0.95=23263; the corrected signal corresponding to sampling port C is, for example, 8700 / 0.72=12083; and the corrected signal corresponding to sampling port D is, for example, 19500 / 0.88=22159.

[0159] Step S4: Spatial reconstruction model processing.

[0160] Specifically, the radial basis function interpolation method is adopted, and one-dimensional radial interpolation (Gaussian kernel, σ=50mm) is performed based on the radial coordinates of four points (radial coordinate of port A is r=0; radial coordinate of port B is r=150mm; radial coordinate of port C is r≈113mm; radial coordinate of port D is r≈141mm). Since the process chamber structure is axisymmetric (circular chamber, annular air inlet, central pump port), it can be assumed that the concentration of active ingredient does not have significant differences in the azimuth direction. Therefore, the one-dimensional radial concentration distribution can be rotated around the central axis to generate a two-dimensional full-plane distribution map.

[0161] Step S5, Output and Feedback.

[0162] Specifically, the reconstruction results show that the Cl concentration in the central region is approximately 2.6 × 10⁻⁶. -5 Pa, approximately 2.2 × 10 at the edge. -5 Pa, edge / center ratio ≈ 0.85. The system judges the uniformity as acceptable and does not provide feedback; if the ratio is lower than 0.70, an early warning is issued.

[0163] In some optional implementations, the system can also verify the correlation of the density distribution of a certain effective component by determining the spatial distribution of the target component's concentration, and can also prompt equipment maintenance personnel to replace the sampling port in a timely manner.

[0164] The process for verifying the correlation of the density distribution of a certain active ingredient is as follows: obtain the concentration distribution map of a certain active ingredient and verify its correlation with the density distribution of the active ingredient obtained by radial scanning of optical emission spectrum. The verification result is as follows: R²=0.96, where R is the Pearson correlation coefficient.

[0165] The procedure for prompting equipment maintenance personnel to replace sampling ports in a timely manner is as follows: Due to the corrosive nature of the etching gas, the sampling port adopts a consumable design. The entire sampling port module (including the flange base, flow limiting element, and sampling pipeline head section) is a replaceable consumable component and is replaced as a whole periodically. Each sampling port module should be replaced by equipment maintenance personnel after 200 hours of cumulative exposure to the plasma environment. The analysis software records the usage time of each port and automatically pops up a replacement reminder when the threshold is reached.

[0166] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0167] The following is a detailed reference. Figure 5 The diagram illustrates a structural schematic suitable for implementing an electronic device according to an embodiment of the present invention. The electronic device may include a processor 501, a memory 502, a bus 503, and a communication interface 504, through which the electronic device is connected to input devices and / or output devices.

[0168] Processor 501 may be a central processing unit, a network processor, or a combination thereof. Processor 501 may further include an integrated circuit. The integrated circuit may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GPRS), or any combination thereof. Processor 501 can perform various appropriate actions and processes according to the program stored in memory 502. Memory 502 also stores various programs and data required for the operation of the electronic device.

[0169] Memory 502 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device. Furthermore, memory 502 may include volatile memory, such as random access memory; memory may also include non-volatile memory, such as flash memory, hard disk, or solid-state drive; memory 502 may also include combinations of the above types of memory. Processor 501, memory 502, and communication interface 504 are interconnected via bus 503. Memory 502 can receive data input from input devices via bus 503 and communication interface 504, or transmit data to output devices via bus 503 and communication interface 504. Typically, input devices can be connected to communication interface 504 to input data into the electronic device; input devices include touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices can be connected to communication interface 504 to acquire data from the electronic device; output devices include liquid crystal displays, speakers, vibrators, etc. Although... Figure 5 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0170] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be installed to an electronic device via an input device or according to an installation package in memory 502. When the computer program is executed by processor 501, it performs the functions defined in the method for determining the spatial distribution of the concentration of a target component according to embodiments of the present invention.

[0171] Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0172] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the method for determining the spatial distribution of the concentration of the target component shown in the above embodiments is implemented.

[0173] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0174] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A system for determining the spatial distribution of the concentration of a target component, characterized in that, The system includes: a sampling module, a component analysis module, and a spatial distribution determination module; The sampling module includes a preset number of sampling ports, a preset number of current limiting elements, and a preset number of sampling pipelines; the spacing between the sampling ports is greater than a first preset threshold, and the spacing between the sampling ports is determined according to the type of the spatial reconstruction model. The first preset threshold is used to control the sampling areas where the sampling ports are located to not overlap. The sampling module is connected to the component analysis module and is used to obtain component samples in the process chamber through the sampling port; The current limiting element is connected to the sampling port; the sampling module is used to adjust the component sample obtained by the sampling port through the current limiting element to obtain the target component sample, and to transmit the target component sample to the sampling pipeline, wherein the type and size of the current limiting element are determined according to the pressure of the target component in the process chamber and the preset rules; The sampling module is used to transfer the target component sample to the component analysis module through the sampling pipeline, wherein the length difference between different sampling pipelines is less than a second preset threshold, and the second preset threshold is used to control the length difference between the sampling pipelines; The component analysis module is connected to the spatial distribution determination module and is used to generate a mass spectrometry feature signal based on the target component sample, and input the mass spectrometry feature signal into the spatial distribution determination module. The spatial distribution determination module is used to generate spatial distribution information of the target component's concentration based on the mass spectrometry feature signal, the location coordinates of the sampling port, and the spatial reconstruction model.

2. The system according to claim 1, characterized in that, The sampling module further includes: a path adjustment component and a path switching component; the sampling module also includes a heating belt; The pathway adjustment component is used to clean any residual sample in the sampling pipeline before transmitting the target component sample through the sampling pipeline; The pathway switching component is connected to the pathway adjustment component through multiple first pathways, and is used to take the first pathway corresponding to the target sampling pipeline as the target pathway and connect the target pathway to the component analysis module so that the target pathway transmits the target component sample to the component analysis module. The target sampling pipeline is the sampling pipeline that is transmitting the target component sample. The heating band is wrapped around the outer wall of the sampling pipeline to adjust the temperature of the target component sample in the sampling pipeline.

3. The system according to claim 2, characterized in that, The pathway adjustment assembly includes: a three-way valve and a pre-pump; The first port of the three-way valve is connected to the sampling pipeline, the second port of the three-way valve is connected to the pre-pump, and the third port of the three-way valve is connected to the first passage. In the three-way valve, the first port and the second port are connected through the second passage, and the first port and the third port are connected through the third passage. The pathway adjustment component is used to close the third pathway corresponding to the candidate sampling pipeline and open the second pathway corresponding to the candidate sampling pipeline before transmitting the target component sample through the candidate sampling pipeline. The pathway adjustment component is also used to extract the residual sample in the candidate sampling pipeline using the pre-pump, and continue for a preset duration. The pathway adjustment component is also used to close the second pathway corresponding to the candidate sampling pipeline and open the third pathway corresponding to the candidate sampling pipeline after the preset time period.

4. The system according to claim 2, characterized in that, The spatial distribution determination module includes: a data preprocessing component, a model invocation component, and an information adjustment component; The data preprocessing component is used to preprocess the mass spectrometry feature signal to obtain the target signal; The model invocation component is used to determine the target model among multiple spatial reconstruction models, input the target signal and the position coordinates of the sampling port into the target model, and obtain the spatial distribution information of the concentration of the target component; The sampling port is located in the process chamber at at least one of the following locations: the top center of the process chamber, the top edge, the near-wafer surface of the chamber sidewall, the sampling area where the gas injection port is located, and the sampling area where the pump port is located. The information adjustment component is used to generate display information based on the target indicators in the concentration spatial distribution information, and transmit the display information to the target device.

5. The system according to claim 4, characterized in that, The data preprocessing component includes: a preprocessing unit, a first processing unit, a second processing unit, and a calibration unit; The preprocessing unit is used to obtain the length of the sampling pipeline and the average linear velocity of the component sample in the sampling pipeline, determine the correction offset based on the pipeline length and the average linear velocity, and adjust the mass spectrometry feature signal based on the correction offset to obtain an intermediate signal. The calibration unit is used to determine the transfer coefficient of each sampling port through a step-by-step calibration method, wherein the step-by-step calibration method includes: when the process chamber does not contain plasma, introducing a calibration gas of known concentration into the process chamber, measuring the response signal of each sampling port, and obtaining the molecular state transfer coefficient based on the response signal; when the process chamber contains plasma, determining the free ground state relative sensitivity factor based on the mass spectrometry signal of the internal standard gas and an independent reference measurement method, and obtaining the free ground state transfer coefficient based on the molecular state transfer coefficient and the free ground state relative sensitivity factor. The first processing unit is configured to, when the target component sample does not contain a preset gas, obtain the first characteristic peak intensity of the target component based on the intermediate signal, and obtain the target signal based on the ratio of the first characteristic peak intensity to the free ground state transfer coefficient; The second processing unit is configured to, when the target component sample contains a preset gas, obtain the first characteristic peak intensity of the target component and the second characteristic peak intensity of the preset gas based on the intermediate signal, and obtain the target signal based on the ratio of the first characteristic peak intensity to the second characteristic peak intensity.

6. The system according to claim 5, characterized in that, The calibration unit is further configured to: introduce the calibration gas into the process chamber during process intermittent periods and when the process chamber does not contain plasma, wherein the calibration gas is used to update the molecular state transfer coefficient.

7. The system according to claim 1, characterized in that, The system also includes: a control module; The control module is used to determine the current process stage and determine the sampling port set and the sampling port calling order in the sampling port set according to the current process stage, wherein the sampling port set is used to determine the sampling ports that need to acquire the target component sample.

8. A method for determining the spatial distribution of the concentration of a target component, characterized in that, The method is applied to the concentration spatial distribution determination system of the target component according to any one of claims 1 to 7, and the method includes: Component samples are acquired in the process chamber through sampling ports, wherein the spacing between the sampling ports is greater than a first preset threshold, the spacing between the sampling ports is determined according to the type of spatial reconstruction model, and the first preset threshold is used to control the sampling areas where the sampling ports are located to not overlap. The target component sample is obtained by adjusting the component sample acquired by the sampling port through the current limiting element, wherein the type and size of the current limiting element are determined according to the pressure of the target component in the process chamber and the preset rules; The target component sample is transferred to the component analysis module through sampling tubing, wherein the length difference between different sampling tubing is less than a second preset threshold, and the second preset threshold is used to control the length difference between the sampling tubing; The component analysis module generates mass spectrometry feature signals based on the target component sample. Based on the mass spectrometry characteristic signal, the location coordinates of the sampling port, and the spatial reconstruction model, the spatial distribution information of the target component's concentration is generated.

9. The method according to claim 8, characterized in that, The step of generating spatial distribution information of the target component's concentration based on the mass spectrometry feature signal, the location coordinates of the sampling port, and the spatial reconstruction model includes: When the spatial reconstruction model is the first sub-model, the position coordinates are converted to polar coordinates; Based on the mass spectrometry characteristic signal, the radial concentration distribution of the target component is obtained; The radial concentration distribution is interpolated using the polar coordinates to obtain the target equation, and the weights corresponding to the sampling ports are obtained based on the target equation. Based on the weights, the polar coordinates, and the radial concentration distribution, the spatial distribution information of the target component's concentration is obtained.

10. The method according to claim 8, characterized in that, The step of generating spatial distribution information of the target component's concentration based on the mass spectrometry feature signal, the location coordinates of the sampling port, and the spatial reconstruction model includes: When the spatial reconstruction model is the second sub-model, the mass spectrometry feature signal is used as the observation value. Based on the observation value, the position coordinates and the formula corresponding to the second sub-model, the predicted mean and confidence interval of each sampling point in the process chamber are obtained. The formula corresponding to the second sub-model satisfies: in, Let be the position coordinates of any location within the process chamber. The coordinates of another location, , Both represent the x-axis. , Both represent the ordinate. express The target active ingredient concentration at the location; Represent a Gaussian process; Represents the mean function; This represents the covariance function, used to describe any two positions. and Spatial correlation between concentration values; Based on the predicted mean and the confidence interval, the spatial distribution information of the concentration of the target component is obtained; When the spatial reconstruction model is a third sub-model, the process parameter vector is obtained, and the normalized signal intensity corresponding to the sampling port is obtained according to the mass spectrometry feature signal. The input vector is obtained based on the process parameter vector, the normalized signal intensity, and the position coordinates; The input vector is input into the third sub-model to obtain the spatial distribution information of the concentration of the target component.

11. The method according to claim 8, characterized in that, Before generating the mass spectrometry feature signal based on the target component sample, the method further includes: The transfer coefficient of each sampling port is determined by a step-by-step calibration method; The stepwise calibration method includes: in the absence of plasma in the process chamber, introducing a calibration gas of known concentration into the process chamber, measuring the response signal of each of the sampling ports, and obtaining the molecular state transfer coefficient based on the response signal; in the presence of plasma in the process chamber, using the mass spectrometry signal of the internal standard gas as an internal scale, combined with an independent reference measurement method, determining the free ground state relative sensitivity factor, and obtaining the free ground state transfer coefficient based on the molecular state transfer coefficient and the free ground state relative sensitivity factor.

12. An electronic device, characterized in that, include: A memory and a processor are communicatively connected, the memory storing computer instructions, and the processor executing the computer instructions to perform the method for determining the concentration spatial distribution of the target component as described in any one of claims 8 to 11.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the method for determining the spatial distribution of the concentration of the target component as described in any one of claims 8 to 11.

14. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the method for determining the concentration spatial distribution of the target component as described in any one of claims 8 to 11.

Images