Systems and methods for monitoring semiconductor processes

Abstract

A system and method for monitoring a semiconductor process includes a plurality of sensors and a microcontroller. The plurality of sensors are disposed within a process chamber. The microcontroller receives data from the plurality of sensors and measures the uniformity of the semiconductor process based on the data received from the plurality of sensors. The plurality of sensors are configured to emit process signals indicative of a material process occurring proximate thereto, wherein a first sensor defines a first spatial location within the process chamber and a second sensor defines a second spatial location within the process chamber, the first spatial location having a different angular orientation than the second spatial location.

Description

Technical Field

[0001] (Related Application) This application is a provisional application filed under the relevant portions of 35 U.S.C. § 111 and 37 C.F.R. § 1.53.

[0002] This disclosure relates to the field of semiconductor manufacturing, including semiconductor manufacturing control. More particularly, in one example, a sensor system monitors semiconductor manufacturing process tools, and specifically directly monitors process uniformity within a chamber in deposition and etching processes.

Background Art

[0003] Deposition and etching processes in semiconductor manufacturing plants are widely and commonly used during device manufacturing in the semiconductor integrated circuit (IC) industry. Conventionally, the semiconductor industry's efforts to shrink dimensions previously limited by two-dimensional lithography resolution have shifted to deposition and etching process control of three-dimensional structures (e.g., 3D gates and 3D NAND). The critical dimensions of a device are more greatly affected by the control capabilities of the deposition and etching processes. Therefore, there is a need for new technologies that enable more precise process control in etching and deposition processes.

[0004] Plasma etching processes are often used to remove dielectric, semiconductor, or metal layers with an igniting gas in a plasma state (which drives the activation energy of chemical reactions). It is also possible to perform material removal by flowing a reaction gas (in a non-plasma state) or through a wet etch (liquid state) station. Deposition on chamber components and processed substrates can be performed by various methods such as plasma-enhanced (PE) chemical vapor deposition (CVD), sub-atmospheric pressure CVD, thermal CVD, atomic layer deposition (ALD), plasma-enhanced atomic layer deposition, etc. Etching and deposition processes can be made isotropic or anisotropic (such as reactive ion etching (RIE), etc.) depending on the process step.

[0005] In a substrate deposition process such as an IC manufacturing process, the deposition of many different layers on a wafer (the substrate) can be achieved through different reactions and various process material states. Example technologies include plasma (PECVD and high-density plasma (HDP)), gas sub-atmospheric pressure CVD (SACVD), and liquid (electroplating). Some examples of the main parameters for controlling the deposition layer and device manufacturing characteristics include thickness, stress, mass, resistance, particles, and refractive index. These parameters are measured and controlled not only for the average value (across the entire wafer or wafer batch), but also for the wafer-to-wafer variation and the within-wafer variation. Reducing process variation contributes to improving the manufacturing yield in the EOL (End of Line) process.

[0006] For example, in substrate etching, the following steps are used: a wafer etching step for applying a pattern to the manufactured device (in combination with a lithography step), a step for cleaning the wafer from contamination, a step for creating trenches between transistors, a step for enabling separation between contacts and isolators, and a step for reacting the wafer surface to remove photoresist prior to deposition. The main parameters for controlling the etching process on the wafer are the etching rate, thickness, stress, control of particles and defects, and critical dimensions for defined features such as other electrical and optical parameters.

[0007] Substrate etching and deposition may be a simultaneous process, either in the same process chamber, continuously within the chamber, not continuously within the chamber, or in different chambers (e.g., in some HDP processes, etching and deposition may occur either continuously or simultaneously).

[0008] During the substrate deposition sequence, by-products from the process can deposit on the chamber components. This process can be non-uniform (varying depending on the chamber location) depending on the process step or conditions. Chamber deposition is performed with no substrate in the process chamber. Coating the chamber components (sometimes referred to as "undercoat", "precoat", "seasoning", etc.) this deposition can improve the wafer uniformity and controllability. Examples of deposition between substrates that affect the wafer process include improving particle performance on the wafer by "adhering" particles in the undercoat layer, improving the uniformity of the wafer process by applying a precoat layer to the chamber components, and improving the control of the deposition rate by seasoning the chamber.

[0009] Chamber etching (during the substrate etching / washing process) is performed when there is no substrate in the chamber, removing the coating from the chamber components and cleaning the chamber from by-products. This process improves particle performance and process control. In this process, material is etched from the chamber components.

[0010] Some of the process steps can include a pretreatment or post-treatment to etch the wafer surface before or after the deposition step (onto the wafer). For example, the pretreatment (before deposition) includes removing contaminants from the wafer surface to promote better adhesion of the deposited layer, and the post-treatment includes annealing or "shrinking" the deposited layer. During these processes, by-products removed from the substrate can adhere and deposit on the chamber components and / or different chamber components can be etched. For example, during the plasma pretreatment of the substrate, by-products may deposit on the chamber walls but may also be removed from the chamber chuck (where the wafer sits).

[0011] Post-maintenance depositions and etchings, including preventive or post-failure maintenance of the process chamber, with or without substrates, are applied to enable better particle performance, process uniformity (within and / or between substrates), process control, and rate control. In this process, by-products may be deposited or removed (etched) from different chamber components.

[0012] During wafer deposition and etching, cross-impacts between different components within the chamber and along the substrate occur, for example, due to process parameters such as flow, local temperature, plasma density, pressure, etc., and may change as a result of the condition of the chamber components. Cross-impacts affect both the substrate conditions and the chamber process (etching or deposition on chamber components).

[0013] An example of cross-impact between a chamber component and a processed substrate includes a substrate processed through a clean chamber (no deposition before the substrate inlet). The deposition rate and thickness uniformity on the substrate may be different compared to, for example, a substrate processed in a chamber containing a coating material such as pre-cleaning.

[0014] Another example of cross-impact relates to hard mask gate etching, where by-products accumulated on the chamber walls affect the etching rate and critical dimensions on the substrate.

[0015] Therefore, it has an adverse effect on production yield and limits the accumulation window during wafer production. Subsequently, it is necessary to use a plasma cleaning process to clean the residues from the process chamber.

[0016] The term "local and direct process measurement using QCM" relates to a method in which the mass added to or removed from the QCM surface changes the QCM frequency (described in the Sauerbrey equation that correlates the change in the oscillation frequency of a piezoelectric crystal with the mass deposited on its surface). The effect of film stress is not explicitly described in the basic Sauerbrey equation, but it can affect the QCM frequency and the effective shear modulus. The term "direct measurement" of QCM includes the measurement of a film deposited on or removed from the crystal surface of a piezoelectric crystal. The term "local measurement" of QCM means the measurement of a film deposited locally or proximally at a specific location.

[0017] Known methods for process monitoring using integrated sensors include mass spectrometers, optical spectrometers, RF sensors, and vacuum gauges. However, such methods are not local and cannot provide detailed information about films accumulated or removed at different chamber positions. As an example of non-local process control, there are plasma cleaning methods such as emission spectroscopy, residual gas analyzers, and chamber impedance measurements. However, all of these methods measure the convolution signal from the entire chamber and do not identify the uniformity or homogeneity of the process materials at different chamber positions. Other known sensors such as temperature sensors can localize and read measurements along the surfaces of various chamber components, but they do not provide detailed information about the film state related to the coatings on these surfaces.

[0018] Patent Document 1 (Wajid) discloses a QCM that provides information about film coating or etching, but since it employs only one location, it cannot provide information about the uniformity and homogeneity of the process at different chamber locations. As the chamber size increases, the accuracy and value of the process data decrease.

[0019] Patent Document 2 (Martinson, et al.) describes a QCM probe that moves between different chamber positions. However, this solution is limited to the laboratory and is only suitable for production environments that require a vacuum for production. Furthermore, it is not easy to simultaneously monitor QCM sensors at different chamber positions.

[0020] Therefore, there is no direct and local approach for monitoring the uniformity / homogeneity of processes in chemical vapor deposition (CVD) and etch chambers.

Prior Art Documents

Patent Documents

[0021]

Patent Document 1

Patent Document 2

Summary of the Invention

[0022] Therefore, in one embodiment of the present disclosure, a system for monitoring a semiconductor process is provided. The system includes a plurality of sensors and a microcontroller. The plurality of sensors are disposed in a process chamber at various predetermined positions. The microcontroller receives data from the plurality of sensors and measures the uniformity of the semiconductor process based on the data received from the plurality of sensors. At least one of the sensors defines an angular direction different from that of another sensor. In yet another embodiment, at least two of the sensors each include a sensor surface where the angular direction of one sensor surface defines an angular direction different from that of another sensor surface.

[0023] In another embodiment of the present disclosure, a system for monitoring a semiconductor process is provided, the system comprising a plurality of sensors selectively disposed within a process chamber and configured to issue a process signal indicative of a material process occurring proximate to each of the plurality of sensors. A first sensor and a second sensor respectively define a first spatial position and a second spatial position within the process chamber, the first spatial position having an angular direction different from that of the second spatial position. A microcontroller correlates, in response to the process signal, the material processes sensed by the first and second sensors with a semiconductor process occurring within the current process chamber. The sensors acquire information useful for the implementation of a semiconductor process performed within one of the current process chamber and an associated process chamber.

[0024] In yet another embodiment, a method for monitoring a semiconductor process is provided. The method includes disposing a plurality of sensors at different positions within a process chamber and measuring the deposition or removal of a process material on a substrate disposed within the process chamber. A first sensor of the plurality of sensors has a first angular direction and a first position within the process chamber, and a second sensor of the plurality of sensors has a second angular direction and a second position within the process chamber. The method also includes monitoring the deposition or removal of the process material at different positions within the process chamber and measuring the uniformity of the semiconductor process based on data received from the plurality of sensors during the monitoring step. The first angular direction and first position of the first sensor and the second angular direction and second position of the second sensor facilitate measuring the uniformity of the semiconductor process.

[0025] In yet another embodiment, a method for monitoring a semiconductor process within a process chamber is provided, the method comprising: (i) providing a baseline substrate having a thickness profile that meets an acceptance profile; (ii) disposing a plurality of sensors within the process chamber and measuring material process data occurring proximate to each of the plurality of sensors; (iii) correlating the measured material process data with a thickness profile of a test substrate manufactured within the process chamber; (iv) comparing the thickness profile of the test substrate with the baseline substrate to determine whether the test substrate meets an acceptance criterion established by the acceptance profile of the baseline substrate; (v) recording the measured material process data of the sensors when the acceptance criterion of the test substrate is met; and (vi) processing a substrate subsequently manufactured according to the recorded material process data.

[0026] The above embodiments are merely exemplary. Other embodiments as described herein are within the scope of the disclosed subject matter.

Brief Description of the Drawings

[0027] To better understand the features of the present disclosure, a detailed description can be given by referring to specific embodiments, some of which are illustrated in the accompanying drawings. However, it should be noted that the drawings only illustrate specific embodiments, and thus, the scope of the disclosed subject matter is not to be considered limited as it equally encompasses other embodiments. The drawings are not necessarily to scale and generally focus on explaining the features of specific embodiments. In the drawings, like numerals are used to indicate like parts throughout the various figures.

[0028]

Figure 1A

Figure 1B

Figure 2

Figure 3A

Figure 3B

[0029] Corresponding reference characters indicate corresponding parts throughout the several views. The examples described herein illustrate several embodiments and should in no way be construed as limiting the scope.

DETAILED DESCRIPTION OF THE INVENTION

[0030] The present disclosure relates to the field of semiconductor manufacturing, including semiconductor manufacturing control. More specifically, in one example, a sensor system monitors semiconductor manufacturing process tools and, specifically, directly monitors the homogeneity of chamber processes in deposition and etching processes. For example, disclosed herein is a unique method for direct measurement of thin film deposition and etching using sensors that enable local monitoring of the homogeneity of a process at one or more locations within a process chamber. This provides insights into local cross-effects between the states of different chamber parts and the substrate. Advantageously, by deploying sensors at different chamber locations, it becomes possible to measure different film properties (mass density and stress) resulting from non-uniformities in processes (deposition and / or etching) within the process chamber.

[0031] In the present disclosure, many different types of sensors can be employed. For example, a quartz crystal microbalance (QCM) sensor or a microelectromechanical (MEM) sensor may be deployed. An example of the MEM sensor used in the present disclosure is a surface acoustic wave sensor. A person skilled in the art can easily understand how QCM sensors and MEM sensors are made and used. The present disclosure utilizes various such sensors disposed at different positions within the process chamber. In any embodiment discussing a QCM sensor, any other sensor such as a MEM sensor may be employed.

[0032] In one or more embodiments, any combination of the following sensor types may be used as sensors. Those types are capacitor sensors, photocathodes, photodetector sensors, microfabricated ultrasonic transducers, oscillator devices configured to measure energy or mass changes, resonant electro / optical devices, resistance measurement sensors, sensors having a dielectric waveguide in contact with a metal layer or metal pattern suitable for generating a plasmonic response, light emitting devices, electron beam sources, ultrasonic sources, optical resonators, microring resonators, photonic crystal structure resonators, temperature sensors.

[0033] In one embodiment, the system uses the local sensitivity of a quartz crystal microbalance (QCM) sensor to perform deposition and etching processes (adding or removing mass or / and film stress onto or from the QCM) that employ the resonant frequency of the quartz (at known process conditions). One embodiment of the present disclosure can also use the QCM sensor to monitor process conditions such as temperature, flow rate, pressure, etc. with a known thickness and stress accumulation to monitor local process conditions. Instead of a QCM sensor, a MEM sensor can also be used in a similar manner.

[0034] By using QCM sensors at different positions within the process chamber and sampling the QCM sensors between different process sequences (e.g., different deposition and etching steps), important information reflecting the real-time process homogeneity of the chamber can be obtained. Instead of QCM sensors, MEM sensors can also be used in the same way.

[0035] The above process can be used in multiple chambers to perform chamber-tool matching and can be used to monitor process variations (between the chamber and the tool). Another advantage is related to shortening the time to transfer a specific process chamber into production by measuring the deposition rate using multiple sensors. Wafer edge uniformity can also be obtained by using multiple sensors at the chamber radius around the wafer edge. Monitoring process homogeneity for different sequences within an HDP deposition chamber may include monitoring precoat deposition homogeneity, wafer deposition homogeneity, plasma cleaning homogeneity, etc.

[0036] Measurement of process homogeneity can be obtained by measuring QCM frequency values that start at the beginning of the deposition sequence and end at the plasma cleaning sequence (for a given manufacturing recipe). Furthermore, the difference or delta in frequency between different runs (from start to end) can provide information about the stability of the process at a given location.

[0037] Another example of measuring process homogeneity is related to the frequency difference between the start and end of wafer deposition between different wafers (of the same recipe). Then, specific correlation parameters or equations (based on the QCM position) can be calculated to predict the wafer thickness and thickness variations. This helps to avoid using test wafers for thickness measurement or can be used as feedforward or feedback information to control different process operations before or after substrate deposition. Instead of QCM sensors, MEM sensors can also be used in the same way.

[0038] Process homogeneity can also be measured by obtaining the maximum frequency during plasma cleaning from different QCM positions, enabling the user to know whether the film has accumulated due to under-etching or over-etching at a specific position. The algorithm for determining the process endpoint can use frequency information from multiple QCM sensors distributed at different locations and can be used to optimize the process endpoint (EP) of the cleaning. For example, until a threshold is reached, i.e., until the process endpoint of the cleaning is reached and the derivative of the frequency becomes much lower, the moving average of the derivative of the frequency can be monitored. For example, this over-etching or under-etching for different parts can be intentionally reached or achieved. Similar or analogous approaches can also be applied to other time-based processes that use addition or removal of materials, such as undercoats, precoats, etc.

[0039] Endpoint detection for wafer-based processes such as deposition, etching, densification, and contamination removal using plasma or heat (pretreatment or bake-out) can also be realized using signal inputs from multiple QCM sensors distributed at different positions. QCM sensors at different locations within the process chamber can measure different deposition and etching rates to provide information about the process uniformity. For example, when a nozzle is clogged (in an HDP chamber), the QCM closer associated with this specific nozzle can measure different process rates. Instead of QCM sensors, MEM sensors can also be used in a similar manner.

[0040] Furthermore, by implementing at least two QCM sensors each having a different angular direction (with respect to the wafer plane), it is possible to measure and / or calculate the process rates at different angles on the wafer and provide three-dimensional information about the process and process rates in the wafer plane. Also, by applying a variable DC bias to the QCM with respect to the chamber ground, the deposition / sputter ratio on the wafer on the QCM can be generated as a tuning factor.

[0041] The control of process uniformity is key to integrating different processes and directly affects die yield. FIG. 1A shows a chamber 100 for semiconductor processing, for example, a PECVD chamber. A plurality of QCM sensors 102 are deployed at various locations within or near the chamber 100, such as on the sidewalls, bottom, or top. A heating table 104 that supports a substrate 108 (i.e., a wafer) is deployed within the chamber, and one or more QCM sensors 102 may be deployed on the heating table 104. The chamber 100 includes a top electrode 106 and a gas inlet 110, which may also be monitored by additional QCM sensors 102. Next, the chamber includes pumps 112 for removing process gases, and those pumps 112 may also be monitored by additional QCM sensors 102. Although QCM sensors are used in this specific example, MEM sensors can be used as well.

[0042] FIG. 1B shows another chamber 101 for semiconductor processing. In the embodiment of FIG. 1B, a plurality of QCM sensors 102 are deployed at different locations within the process chamber and at different positions and / or angles relative to the substrate. Advantageously, the plurality of QCM sensors 102 enable the monitoring of process (and process rate) homogeneity at different locations on the process chamber and in different chamber parts. Based on the data collected from the plurality of QCM sensors 102, the process can be optimized, and process uniformity issues may be detected within the process chamber. For example, since this system uses a plurality of QCM sensors 102 at different angles and positions, it is possible to monitor and control the process and process rate on the substrate. Furthermore, a 3D map of process uniformity can be created using that data. Although QCM sensors are used in this specific example, MEM sensors can be used as well.

[0043] As shown in FIG. 2, the deposition (or the delta of the fundamental frequency on the wafer) measured by QCM sensors (inside the deposition chamber) at two different positions is different for wafer-to-wafer variations. The data of sensor 2 (FF2) shows wafer-to-wafer stability, while sensor 1 reflects higher process variations between wafers.

[0044] For example, the cleaning endpoint of the chamber wall may be determined by a single QCM sensor inside the chamber. Different cleaning recipes are expected to clean the chamber wall up to that endpoint. Unfortunately, even using a single QCM, the cleanliness of the entire chamber wall cannot be properly determined. It will be understood that the active area of the QCM is very small compared to the area of the chamber wall with large dispersion. Therefore, by using multiple QCM sensors at selected locations, the average cleanliness state of the chamber can be fairly represented. This not only helps to conclude that the chamber wall is sufficiently clean, but also enables understanding of the non-uniformity of the etching rate using the gradient of the etching frequency, and then using it to bring the chamber to an acceptable state.

[0045] Another advantage of using multiple sensors is related to determining the correct trade-off between under-etching conditions and over-etching conditions. For example, non-uniform etching inside the chamber may over-clean one part of the chamber and leave another part not fully cleaned, which may cause a certain over-etching of the chamber wall (the chamber material becomes particles) or / and progressive film accumulation (eventually particle detachment due to stress relaxation). Therefore, having multiple QCM sensors allows such non-uniformity-related information to be utilized in real time. Furthermore, the process can be changed to produce a satisfactory compromise until a permanent solution is found. This also helps to fine-tune a given chamber to the state of a golden chamber.

[0046] One embodiment of the present disclosure employs a system comprising a plurality of QCM sensors to monitor, learn, and control the homogeneity of a chamber process in real time. This system also provides insights regarding the cross-impact between substrate processing (etching and deposition) at different positions of different chamber parts at different or the same timestamps. By monitoring different chamber positions at the same timestamp, the process homogeneity can be measured (by comparing the QCM signals at different positions). By monitoring QCM across different process sequences (e.g., undercoat, deposition, etching, cleaning, etc.), the process homogeneity can be monitored for each process sequence or for subsequent processes in different process chambers.

[0047] One embodiment of the present disclosure enables the monitoring of the homogeneity of deposition and etching with respect to stacks of layers that can be composed of different materials, different layer stresses, and different densities. As an example, a multilayer deposition of oxide stacked on nitride used in 3D NAND technology can be mentioned. Currently, no direct measurements are made that enable in-situ process control. By installing QCM sensors at multiple locations within the stack deposition chamber, real-time monitoring becomes possible. QCM data can be fed back as a process parameter to the recipe, for example, the deposition time of a single layer within the stack can be changed in real time based on the QCM sensor data.

[0048] The system of the present disclosure can be used and compared in multiple chambers to achieve better chamber matching and reduce overall process variations. The plurality of sensors provide the advantage of verifying the reproducibility of homogeneity between wafers and the spatial homogeneity of a given wafer. Also, the data enables homogeneity analysis between steps (for the purpose of root cause analysis) by understanding which specific steps cause variations between process steps by binning the frequency shift.

[0049] In one embodiment, the system includes a plurality of single or dual QCM sensors mounted at carefully selected locations in the process chamber to monitor changes in the deposition and / or etching of thin films on the chamber walls proximate to each sensor. Such conditions may include, among other things, mass, stress, density, or material composition on the QCM sensor. Also, the QCM data can be used to correlate other process conditions (flow rate, pressure, temperature, plasma uniformity, etc.). This is because these changes affect the measured QCM fundamental frequency (at different harmonics) and, in some cases, the fundamental frequency change rate. Although QCM sensors are used in this specific example, MEM sensors can be used as well.

[0050] In such a case, the software communicates with the process tool and the QCM sensors. The SW likely receives logistic information such as the production recipe name, lot ID, slot ID, wafer ID (WID), sequence ID, etc. from the process tool, as well as other variables such as step number, RF power, pressure, etc. By combining the tool data with the QCM sensors, local process homogeneity between different process steps and within a process step can be measured.

[0051] In another example, the software may communicate with multiple chambers during the same process operation, enabling comparison and matching between different chambers. The software may also communicate with multiple chambers during different operations within the process. That is, information collected from a particular operation can be fed forward or feedback to different operations, improving process stability and reducing variability.

[0052] By measuring the wafer thickness predicted using QCM data, that data can flow into subsequent operations and be used to calculate the endpoint time in a CMP operation. A real-time 3D thickness wafer map or any other measurement for the process chamber can be developed by the methods and systems of the present disclosure. The QCM system first "learns" the correlation coefficient of the QCM sensor data to the wafer thickness (or any other measurement value) at each X / Y position on the wafer. This can be done by using a thickness measurement map of an external stand-alone measuring device (or other in-situ measurement) to "learn" the QCM sensor using the correlation of the QCM sensor data at different positions (or / and angles) within the process chamber to different X / Y positions (coordinates) on the wafer. Based on the correlation coefficient, the thickness of the wafer, stress map, or other measurement of interest can be derived from the QCM sensor system data in real time.

[0053] This method can also include providing a plurality of QCM sensors within the process chamber and performing process monitoring capable of monitoring the material thickness on the deposition chamber while recording QCM data. The thickness of the wafer can be measured by an external measurement tool at a plurality of positions on the substrate or wafer. A statistical model can be developed using QCM data or a mathematical transformation of QCM data as input and substrate thickness profile data (or a mathematical transformation of this data) at different positions on the substrate or wafer as output. Based on the calculated data, the QCM data can be used in real time to develop a three-dimensional (X / Y position, Z = thickness) surface profile. Finally, the substrate thickness profile or other measurement value can be automatically imported into the QCM sensor system and the QCM data can be automatically optimized.

[0054] Figures 3A and 3B show the correlation maps between two QCM sensors at different positions on the process chamber of FIG. 1. For example, FIGS. 3A and 3B show 49 thickness maps measured by an external measurement tool to determine which regions strongly correlate with the results of the QCM sensors. FF1 represents sensor 1 and FF2 represents sensor 2. Some regions show strong correlation while others show weak correlation.

[0055] FIG. 3A shows the correlation map in which the QCM sensor S1 is placed at approximately the 11 o'clock position with respect to the substrate or wafer. FIG. 3B shows the correlation map in which the QCM sensor S2 is placed at approximately the 6 o'clock position with respect to the substrate or wafer. Thus, by combining data for at least two sensors S1 and S2, the thickness profile of the substrate or wafer can be better understood. For example, a weighted average of the sensor data can be used, and the closer the sensor is to the region of interest, the more weight can be given to that sensor. Although QCM sensors are used in this specific example, MEM sensors can be used as well.

[0056] It may also be possible to obtain a higher correlation coefficient (between the thickness data of the X / Y position with respect to the sensor data) between adjacent sensors with respect to the X / Y position on the wafer. For example, sensor S1 shows a higher correlation with respect to the thickness data when placed near the sensor, and the same is true for sensor S2. The multiple QCM systems used can be either of the same crystallographic cut, or a combination of different cuts, selected to minimize the influence of temperature or pressure depending on the position of a given QCM sensor.

[0057] The selection of the QCM type is also based on the process step in that there can be multiple cuts at the same location for different process steps. The QCM sensors can be firmly attached to the chamber or (using a motor) dynamically stretched, tilted, and / or rotated to capture different process conditions (mass removal, mass deposition, densification, temperature, pressure, film stress) at different angles and positions within the process chamber.

[0058] By using in-situ dynamic adjustment of the position and angle of the QCM sensor during different process sequences, it becomes possible to measure and, in some cases, control different process / process speed conditions or endpoints in different parts of the chamber, and also to monitor and / or control different process sequences for different angles and positions on the wafer surface.

[0059] Also, one or more of the above-described embodiments may be modified as follows. For example, in some embodiments, the QCM (or MEM) sensor may use one or more power sources among a wired power source, wireless power charging, a battery, a capacitor, or a so-called supercapacitor or any other thermoelectric power supply. Further, in other embodiments, the signal transmission from the QCM sensor may be transmitted either wired or in any other wireless manner such as Bluetooth® or RF.

[0060] For example, the correlation method between the circular substrate and the cylindrical process chamber may be performed in the following manner. In this example, the cylindrical process chamber may be divided into eight or more sectors each having a substantially pie-shaped geometry, i.e., two radii originating from the center of the chamber, the outer arc of the cylindrical inner wall of the process chamber, and the vertical height. For each sector, a QCM or MEM sensor is placed at a predetermined position within the sector to obtain a sensed value of the mass accumulated on the sensor. The signal provided by the QCM sensor is a change in the frequency or delta frequency value indicating the accumulated mass value. When the circular substrate is positioned at the center within the process chamber, eight sectors will define eight pie-shaped regions.

[0061] Any process parameter of interest, such as the thickness value, is defined for each region of the substrate and may be denoted as thickness values T1 to T8. Next, the values of the sensor readings S1 to S8 are such that T i =F(f S1 , ···, f S8may be mathematically associated with the thickness values T1 to T8 by a mathematical equation of the form, where f Si is a function of the sensor values S1 to S8, and F is all functions of f Si . As will be understood by those skilled in the art, the function F can then be represented by an 8x8 matrix that correlates each of the sensor values with the thickness value of the substrate. This correlation takes the form T = C(f), where T is a vector of eight thickness values T1 to T8, C is a constant 8x8 matrix, and f is a vector of sensor values S1 to S8.

[0062] The applicant has discovered that the C matrix is temporally invariant for a given set of process parameters such as pressure, flow rate, plasma power, etc. Thus, one or more calibrations can be performed to determine the C matrix at a particular time. Thereafter, the values of the C matrix can be used to measure process parameters for subsequent executions of the process. For example, a calibration or baseline wafer substrate can be exposed to a given process recipe in a chamber equipped with eight sensors arranged as described above. Thereafter, the calibration wafer can be measured using any of a variety of techniques such as ellipsometry. From these measurements, the C matrix can be determined. Subsequent executions of the process on the production wafer substrate can use the above determinant to periodically check compliance with design tolerances to determine process parameters such as thickness.

[0063] Additional embodiments include any one of the above-described embodiments, wherein one or more of its components, functionality, or structure are exchanged, replaced, or enhanced with one or more of the components, functionality, or structure of the different above-described embodiments.

[0064] It should be understood that various changes and modifications to the embodiments described herein will be apparent to those skilled in the art. Such changes and modifications can be made without departing from the spirit and scope of the present disclosure and without diminishing its intended advantages. Accordingly, such changes and modifications are intended to be covered by the appended claims.

[0065] Although some embodiments of the present disclosure have been disclosed in the foregoing specification, it will be understood by those skilled in the art that many changes and other embodiments will come to mind in which the present disclosure is concerned and which have the benefit of the teachings shown in the above description and the related drawings. Accordingly, it is understood that the present disclosure is not limited to the specific embodiments disclosed above, and that many modifications and other embodiments are intended to be included within the scope of the appended claims. Further, although specific terms have been employed in this specification and the following claims, they are used only in a general and descriptive sense and not for the purpose of limiting the present disclosure and the following claims.

Claims

1. A system for monitoring a semiconductor process, comprising: a plurality of sensors selectively disposed within a process chamber, the plurality of sensors being configured to emit a process signal indicative of a material process occurring in proximity to each of them, the plurality of sensors including a first sensor defining a first spatial position within the process chamber and a second sensor defining a second spatial position within the process chamber, wherein the first spatial position has an angular direction different from that of the second spatial position; a microcontroller that receives the process signal, correlates the material process sensed by the first sensor and the second sensor with a semiconductor process occurring in the current process chamber, obtains information useful for implementing the semiconductor process performed in one of the current process chamber and an associated process chamber, and is configured to change parameters of the semiconductor process in real time based on the process signal; wherein a substrate within the process chamber defines a two-dimensional plane, and the first sensor and the second sensor have angular directions different from each other with respect to the two-dimensional plane; A system comprising.

2. The system according to claim 1, wherein the semiconductor process relates to a deposition process on a substrate or a removal process from a substrate.

3. The system according to claim 1, wherein the semiconductor process relates to a material thickness profile of a substrate. 【Claim 】 The system according to claim 3, wherein the microcontroller determines the uniformity of the material thickness profile.

5. The system according to claim 1, wherein process signals emitted by at least two sensors are correlated to obtain information regarding a semiconductor process occurring in a region between the at least two sensors.

6. The system according to claim 1, wherein the plurality of sensors are disposed near an edge of the substrate, and the microcontroller controls a semiconductor process to maintain substrate edge uniformity.

7. The system according to claim 1, wherein the microcontroller develops three-dimensional thickness profile data in relation to the substrate.

8. The system according to claim 1, wherein the plurality of sensors comprise quartz crystal microbalance (QCM) sensors.

9. The system according to claim 1, wherein the plurality of sensors includes microelectromechanical (MEM) sensors.

10. The system according to claim 1, wherein the plurality of sensors includes one or more of quartz crystal microbalance (QCM) sensors and microelectromechanical (MEM) sensors.

11. A system for monitoring a semiconductor process, comprising: a plurality of sensors selectively disposed within a process chamber, the plurality of sensors configured to monitor deposition or removal processes at different locations within the process chamber, a first sensor of the plurality of sensors having a first angular direction and a first position with respect to a substrate within the process chamber, the substrate defining a two-dimensional plane, and a second sensor of the plurality of sensors having a second angular direction and a second position with respect to the two-dimensional plane of the substrate within the process chamber; a microcontroller for receiving data from the plurality of sensors, the microcontroller measuring the uniformity of the semiconductor process based on data received from the plurality of sensors, the first angular direction and the first position of the first sensor and the second angular direction and the second position of the second sensor facilitating the measurement of the uniformity of the semiconductor process, and changing parameters of the semiconductor process in real time based on the data received from the first sensor and the second sensor; A system comprising the above.

12. The system according to claim 11, wherein the plurality of sensors includes quartz crystal microbalance (QCM) sensors.

13. The system according to claim 11, wherein the plurality of sensors includes microelectromechanical (MEM) sensors.

14. The plurality of sensors includes one or more of quartz crystal microbalance (QCM) sensors and microelectromechanical (MEM) sensors, and the microcontroller generates three-dimensional thickness profile data regarding the substrate, according to the system of claim 11.

15. The semiconductor process is a material deposition process, and the microcontroller measures the uniformity of material deposition within the chamber, according to the system of claim 11.

16. The semiconductor process consists of a material deposition process, and the microcontroller measures the material thickness in the chamber. The system according to claim 11.

17. The semiconductor process consists of a material removal process, and the microcontroller measures the uniformity of material removal in the chamber. The system according to claim 11, characterized in that.

18. The semiconductor process consists of a material removal process, and the microcontroller measures the material thickness in the chamber. The system according to claim 11.

19. A method for monitoring a semiconductor process, comprising: Providing a plurality of sensors in a process chamber, wherein the plurality of sensors include a first sensor of the plurality of sensors having a first angular direction and a first position with respect to a substrate defining a two-dimensional plane in the process chamber, and a second sensor of the plurality of sensors having a second angular direction and a second position with respect to the two-dimensional plane in the process chamber; Monitoring a deposition process or a removal process at different positions in the process chamber; Measuring the uniformity of the semiconductor process based on data received from the plurality of sensors during the monitoring step, wherein the first angular direction and the first position of the first sensor and the second angular direction and the second position of the second sensor facilitate the measurement of the uniformity of the semiconductor process; Changing parameters of the semiconductor process in real time based on the data received from the first sensor and the second sensor; A method including.

20. The plurality of sensors include a plurality of quartz crystal microbalance (QCM) sensors. The method according to claim 19.

21. The plurality of sensors include microelectromechanical (MEMS) sensors. The method according to claim 19.

22. The plurality of sensors include one or more of a quartz crystal microbalance (QCM) sensor and a microelectromechanical (MEMS) sensor. The microcontroller generates three-dimensional thickness profile data regarding the substrate. The method according to claim 19.

23. The semiconductor process is a material deposition process, and the microcontroller measures the uniformity of material deposition in the chamber. The method according to claim 19. Claim 24 The method according to claim 19, wherein the semiconductor process is a material deposition process, and the microcontroller measures the material thickness in the chamber. Claim 25 The method according to claim 19, wherein the semiconductor process is a material removal process, and the microcontroller measures the uniformity of material removal in the chamber. Claim 26 The method according to claim 19, wherein the semiconductor process is a material removal process, and the microcontroller measures the material thickness in the chamber.

Images