Composite Wavelength for Endpoint Detection in Plasma Etching

Abstract

A method for determining the endpoint of an etch process using optical emission spectroscopy (OES) data as input is described. The OES data is acquired by a spectrometer in a plasma etch process chamber. The acquired time-evolving spectral data is first filtered and de-averaged, and then converted into transformed spectral data or trends using multivariate analysis such as principal component analysis, in which previously calculated principal component weights are used to achieve the transformation. Classifying the principal component weights into two separate groups corresponding to positive and negative natural wavelengths generates separate signed trends (synthetic wavelengths).

Description

Technical Field

[0001] Cross - reference to Related Applications This application is related to and claims the benefit of priority from U.S. Patent Application Publication No. 16 / 548,333, entitled "Synthetic Wavelengths for Endpoint Detection in Plasma Etching", filed on August 22, 2019, the entire content of which is incorporated herein by reference.

[0002] This application relates to, for example, methods and systems for controlling a process of etching a structure on a substrate in semiconductor manufacturing. More specifically, the present invention relates to a method for determining an endpoint of a substrate etching process.

[0003] This application is related to U.S. Patent No. 9,330,990 ('990) entitled "Method of endpoint detection of plasma etching process using multivariate analysis" and U.S. Patent Application Publication No. 10,002,804 entitled "Method of endpoint detection of plasma etching process using multivariate analysis".

Background Art

[0004] Plasma etching processes are generally used in combination with photolithography in the manufacturing processes of semiconductor devices, liquid crystal displays (LCDs), light emitting diodes (LEDs), and some photovoltaic (PV) devices. Generally, a layer of radiation-sensitive material, such as photoresist, is first coated on a substrate and exposed to patterned light to form a latent image. Subsequently, the exposed radiation-sensitive material is developed, and the exposed (or unexposed if negative photoresist is used) radiation-sensitive material is removed, leaving a pattern of the exposed radiation-sensitive material that covers the areas that do not require etching and exposes the areas to be etched. During the etching process, for example, during plasma etching, the substrate and the pattern of the radiation-sensitive material are exposed to excited ions in a plasma processing chamber to remove the material beneath the radiation-sensitive material in order to form etching features such as via holes, trenches, etc. Following the etching of the features of the underlying material, a stripping process is used to remove the remaining radiation-sensitive material from the substrate, exposing the formed etched structure so that further processing is possible.

[0005] In many types of devices, such as semiconductor devices, when a plasma etching process is performed on a first material layer covering a second material layer and an opening or pattern is formed in the first material layer by the etching process, it is important to accurately stop the etching process without proceeding to etch the underlying second material layer.

[0006] For the purpose of controlling the etching process, various endpoint controls are utilized, some of which rely on the analysis of the chemical composition of the gas in the plasma processing chamber, for example, to estimate whether the etching process has proceeded to a lower layer having a chemical composition different from that of the layer during etching. Other processes may rely on direct in-situ measurement of the structure during etching. In the former group, optical emission spectroscopy (OES) is frequently used to monitor the chemical composition of the gas in the plasma processing chamber. The chemical species in the gas in the plasma processing chamber are excited by the plasma excitation mechanism used, and the excited chemical species generate different spectral signatures in the emission spectrum of the plasma. For example, by monitoring the change in the emission spectrum due to the cleaning of the layer during etching and the exposure of the lower layer on the substrate, the etching process can be precisely terminated, that is, by reaching the endpoint, the formation of other defects that reduce the yield, such as etching of the lower layer or undercut, can be avoided.

Summary of the Invention

Problems to be Solved by the Invention

[0007] Depending on the type of structure during etching and the etching process parameters, the change in the emission spectrum of the plasma at the endpoint of the etching process can be extremely distinct and easy to detect, or conversely, subtle and extremely difficult to detect. For example, in the etching of a structure with a very low aspect ratio, it can be difficult to detect the endpoint using the current algorithms for OES data processing. Therefore, in such difficult etching process conditions, improvements are needed to make the etching endpoint detection based on OES data more robust.

Means for Solving the Problems

[0008] The feature of the present application relates to a method for determining the endpoint of an etching process in an etching process, where, at the endpoint, when the etching process forms an opening or pattern in a first material layer, the etching process is precisely stopped without proceeding to the etching of a second material layer below.

[0009] In a non-limiting embodiment, to obtain a plurality of OES data matrices, a plurality of average OES data matrices, and one average OES data matrix, OES data for performing different etching processes is acquired. This data is used to enable the establishment of a multivariate model of the acquired OES data. Once the multivariate model of the OES data is established, it is then used for in-situ etching endpoint detection.

[0010] Using an analysis that classifies wavelengths of similar behavior, a weight vector P for converting the OES data vector into a trend region is determined. Preferably, by classifying the principal component weights into two separate groups corresponding to positive and negative natural wavelengths, separate signed trends (synthetic wavelengths) are generated.

[0011] When determining the synthetic wavelength during in-situ etching endpoint detection, the functional form of the value of the time evolution of the synthetic wavelength is plotted against time to determine the endpoint of the etching process.

[0012] For example, in one embodiment, the time evolution of the ratio of synthetic wavelengths or the time evolution of the time derivative of the ratio of synthetic wavelengths is calculated. However, in other embodiments, any other functional form, such as the square of the ratio of synthetic wavelengths or just a single signed synthetic wavelength or just a natural wavelength trend, can also be calculated.

[0013] In a further non-limiting embodiment, to compensate for OES drift between different wafers, in a principal component analysis (PCA) method, a normalized OES spectrum is used.

[0014] After the time evolution trend variable is calculated, a determination is made as to whether the endpoint has been reached. If the endpoint has actually been reached, the etching process is terminated; otherwise, the etching process is continued and the etching endpoint is continuously monitored.

[0015] The generation of the synthetic wavelength enables the acquisition of a trend similar to the natural wavelength for endpoint detection, but the signal-to-noise ratio (SNR) of the endpoint signal becomes higher.

[0016] This application will be better understood in light of the description given, which is provided non-limitingly, in conjunction with the accompanying drawings.

Brief Description of the Drawings

[0017]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6A

Figure 6B

Figure 7A

Figure 7B

Figure 8A

Figure 8B

Figure 8C

Best Mode for Carrying Out the Invention

[0018] Throughout this specification, "one embodiment" or "an embodiment" means that a particular feature, structure, material, or characteristic described in connection with that embodiment is included in at least one embodiment of the present application, but not necessarily in all embodiments. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification do not necessarily refer to the same embodiment of the present application. Further, the particular features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments.

[0019] One embodiment of the present application shown in FIG. 1 is a plasma etching processing system 10 and a controller 55, and the controller 55 is coupled to the plasma etching processing system 10. The controller 55 is configured to monitor the performance of the plasma etching processing system 10 using data obtained from various sensors disposed in the plasma etching processing system 10. For example, the controller 55 can be used to control various components of the plasma etching processing system 10, detect faults, and detect the end point of the etching process.

[0020] According to an exemplary embodiment of the present application shown in FIG. 1, the plasma etching processing system 10 includes a processing chamber 15, a substrate holder 20 to which a substrate 25 to be processed is fixed, a gas injection system 40, and a vacuum pump system 58. The substrate 25 can be, for example, a semiconductor substrate, a wafer, or an LCD. The plasma etching processing system 10 can be configured to enable the generation of plasma within a processing region 45 adjacent to the surface of the substrate 25, and the plasma is formed through the collision of heated electrons and an ionizable gas. The ionizable gas or a mixture of gases is introduced through the gas injection system 40 to adjust the processing pressure. Desirably, the plasma is utilized to create a material specific to a given material process and assist in the removal of material from the exposed surface of the substrate 25. For example, a controller 55 can be used to control the vacuum pump system 58 and the gas injection system 40.

[0021] The substrate 25 can be moved in and out of the plasma etching processing system 10 through a slot valve (not shown) and a chamber through-hole (not shown) by a robotic substrate transfer system that receives the substrate, for example, by substrate lift pins (not shown) housed within the substrate holder 20 and mechanically transports it by equipment housed within the holder. When the substrate 25 is received from the substrate transfer system, it is lowered to the upper surface of the substrate holder 20.

[0022] For example, the substrate 25 can be fixed to the substrate holder 20 via the electrostatic clamping system 28. Further, the substrate holder 20 may further include a cooling system including a recirculating coolant flow that receives heat from the substrate holder 20 and transfers the heat to a heat exchanger system (not shown) or, in the case of heating, transfers heat from the heat exchanger system. Further, gas can be delivered to the back surface of the substrate via the back surface gas delivery system 26 to improve the gas gap thermal conductivity between the substrate 25 and the substrate holder 20. Such a system can be used when temperature control of the substrate is required when the temperature rises or falls. For example, the temperature control of the substrate can be useful at a temperature above the steady state temperature achieved due to the balance between the heat flux delivered from the plasma to the substrate 25 and the heat flux removed from the substrate 25 by conduction to the substrate holder 20. In other embodiments, heating elements such as resistance heating elements or thermoelectric heaters / coolers may be included.

[0023] Continuing to refer to FIG. 1, the process gas can be introduced into the process region 45 via, for example, the gas injection system 40. The process gas can include, for example, a mixture of gases such as argon, CF4 and O2 or Ar, C4F8 and O2 for oxide etching applications or other chemicals such as O2 / CO / Ar / C4F8, O2 / CO / Ar / C5F8, O2 / CO / Ar / C4F6, O2 / Ar / C4F6, N2 / H2, etc. The gas injection system 40 includes a showerhead, and the process gas is supplied from a gas delivery system (not shown) to the process region 45 via a gas injection plenum (not shown) and a multi-opening showerhead gas injection plate (not shown).

[0024] As further shown in FIG. 1, the plasma etching processing system 10 includes a plasma source 80. For example, RF or microwave power can be coupled from a generator 82 to the plasma source 80 via an impedance matching network or tuner 84. The frequency at which RF power is applied to the plasma source is in the range of 10 MHz to 200 MHz, preferably 60 MHz, for capacitively coupled (CCP), inductively coupled (ICP), and transformer coupled (TCP) plasma sources. For microwave plasma sources 80 such as electron cyclotron resonance (ECR) and surface wave plasma (SWP) sources, the typical operating frequency of the generator 82 is 1 to 5 GHz, preferably about 2.45 GHz. An example of an SWP source 80 is a radial line slot antenna (RLSA) plasma source. Further, the controller 55 can be coupled to the generator 82 and the impedance matching network or tuner 84 to control the application of RF or microwave power to the plasma source 80.

[0025] As shown in FIG. 1, the substrate holder 20 can be electrically biased with an RF voltage by the delivery of RF power from an RF generator 30 to the substrate holder 20 via an impedance matching network 32. The RF bias can serve to attract ions from the plasma formed within the processing region 45 to facilitate the etching process. The frequency at which power is applied to the substrate holder 20 can be in the range of 0.1 MHz to 30 MHz, preferably 2 MHz. Alternatively, RF power can be applied to the substrate holder 20 at multiple frequencies. Further, the impedance matching network 32 serves to maximize the transfer of RF power to the plasma within the processing chamber 15 by minimizing the reflected power. Various match network topologies (e.g., L-type, π-type, T-type, etc.) and automatic control methods can be utilized.

[0026] Various sensors are configured to receive tool data from the plasma etching processing system 10. These sensors can include both sensors specific to the plasma etching processing system 10 and sensors outside the plasma etching processing system 10. The specific sensors can include sensors related to the functions of the plasma etching processing system 10, such as helium backside gas pressure, helium backside flow, electrostatic chuck (ESC) voltage, ESC current, substrate holder 20 temperature (or lower electrode (LEL) temperature), coolant temperature, upper electrode (UEL) temperature, forward RF power, reflected RF power, RF self-induced DC bias, RF peak-to-peak voltage, chamber wall temperature, process gas flow rate, process gas partial pressure, chamber pressure, capacitor setting (i.e., C1 and C2 positions), focus ring thickness, RF time, focus ring RF time, and measurements of any statistical values of these. Alternatively, the external sensors can include those not directly related to the functions of the plasma etching processing system 10, such as the light detection device 34 that monitors the light emitted from the plasma within the processing region 45, as shown in FIG. 1.

[0027] The light detection device 34 may include a detector such as a (silicon) photodiode or a photomultiplier tube (PMT) for measuring the total light intensity emitted from the plasma. The light detection device 34 may further include an optical filter such as a narrowband interference filter. In an alternative embodiment, the light detection device 34 may include a line CCD (charge-coupled device) or CID (charge injection device) array and an optical dispersive element such as a diffraction grating or a prism. Further, the light detection device 34 may include a monochromator (e.g., a diffraction grating / detector system) for measuring light at a given wavelength or a spectrometer (e.g., having rotating or fixed diffraction) for measuring an optical spectrum. The light detection device 34 may include a high-resolution OES sensor from a peak sensor system. Such an OES sensor has a wide spectrum spanning the ultraviolet (UV), visible (VIS), and near-infrared (NIR) optical spectra. In the peak sensor system, the resolution is about 1.4 angstroms, i.e., the sensor can collect 5550 wavelengths in the range of 240 to 1000 nm. In the peak system sensor, the sensor comprises a high-sensitivity small optical fiber UV-VIS-NIR spectrometer integrated with a 2048-pixel linear CCD array.

[0028] In one embodiment of the present application, the spectrometer receives light transmitted through a single bundled optical fiber, and the light output from the optical fiber is dispersed across a line CCD array using a fixed diffraction grating. Similar to the above-described configuration, the light transmitted through the optical vacuum window is focused on the input end of the optical fiber via a lens or a mirror. Different spectrometers individually adjusted for each given spectral range (UV, VIS, and NIR) or broadband spectrometers corresponding to UV, VIS, and NIR form sensors for the processing chamber. Each spectrometer includes an independent analog-to-digital (A / D) converter. Finally, depending on the sensor availability, the entire emission spectrum can be recorded every 0.01 to 1.0 seconds or faster.

[0029] Alternatively, in one embodiment, the light detection device 34 may use a spectrometer having all reflective optical components. Further, in one embodiment, a single spectrometer including a single diffraction grating and a single detector corresponding to the entire range of light wavelengths to be detected may be used. For example, the design and use of emission spectroscopy hardware for obtaining optical OES data using the light detection device 34 are known to those skilled in the art of optical plasma diagnostics.

[0030] The controller 55 includes a microprocessor, a memory, and digital I / O ports (potentially including D / A and / or A / D converters) capable of generating a control voltage sufficient to transmit an input to and activate the plasma etching processing system 10 and to monitor the output from the plasma etching processing system 10. As shown in FIG. 1, the controller 55 is coupled to and can exchange information with an RF generator 30, an impedance matching network 32, a gas injection system 40, a vacuum pump system 58, a backside gas delivery system 26, an electrostatic clamping system 28, and a light detection device 34. Using a program stored in the memory, it interacts with the above-described components of the plasma etching processing system 10 in accordance with the stored processing instructions. An example of the controller 55 is a DELL PRECISION WORKSTATION 530 (trademark) available from Dell of Austin, Texas. The controller 55 can be installed in the vicinity of the plasma etching processing system 10 or can be set up remotely from the plasma etching processing system 10. For example, the controller 55 can exchange data with the plasma etching processing system 10 using at least one of direct connection, intranet, and internet. The controller 55 can be coupled to the intranet, for example, at a customer site (i.e., a device manufacturer, etc.) or can be coupled to the intranet, for example, at a vendor site (i.e., a facility manufacturer). Further, for example, the controller 55 can be coupled to the internet. Further, another computer (i.e., a controller, a server, etc.) can access the controller 55 and exchange data via at least one of direct connection, intranet, and internet. The controller 55 also implements an algorithm for detecting the endpoint of the etching process being executed in the plasma etching processing system 10 based on the input data provided by the light detection device 34, as detailed herein.

[0031] In plasma etching processes, endpoint detection (EPD) using emission spectroscopy is an important technique for controlling etching consistency between wafers. By monitoring trends that change over time generated from one or two selected emission wavelengths, the endpoint at which the etching process should be stopped or terminated becomes apparent. Multivariate data analysis using synthetic wavelengths helps improve the SNR and robustness of EPD. However, synthetic wavelengths generated from multivariate data analysis generally cannot retain some of the inherent characteristics of natural wavelengths, such as having a physical meaning.

[0032] Classification of natural wavelengths using a multivariate model (a non-limiting example of which is PCA) enables trends similar to those of the natural wavelengths for EPD, but can generate synthetic wavelengths with higher SNR endpoint signals. In one example of a non-limiting embodiment of the present application, the classification includes selection of natural wavelengths that exhibit constructive or destructive contributions, and uses separate positive and negative weights for each wavelength to transform OES data into the PCA domain when classifying the wavelengths. However, other methods of classifying wavelengths may be used in generating synthetic OES data.

[0033] The endpoint determination process according to one embodiment proceeds in two phases. In the first phase, plasma etching process execution is performed within the plasma processing chamber 15 (step 110 in FIG. 2), and by using the optical detection device 34 to acquire OES data during the execution of one or more etching processes executed by the plasma etching processing system 10 (step 120), a multivariate model of the acquired OES data can be established (step 130).

[0034] Once a multivariate model for OES data is established, it can be used for in-situ etch endpoint detection in the second phase (step 140) as long as the etching process being performed during the second phase is reasonably similar to that used in one or more of the etching process executions performed in the first phase, from the perspective of the structure being etched, the etching process conditions, the etching process system used, etc. This is to ensure the validity of the multivariate model.

[0035] In a non-limiting embodiment of endpoint determination (shown in FIG. 3), a PCA analysis multivariate model is used with a specific classification of natural wavelengths (i.e., having positive and negative weights), and endpoint detection 200 is started, for example, while the etching process is being performed using a light detection device 34 and OES data is being acquired. During each plasma etching process execution, the spectrum is acquired n times (step 210 in FIG. 3), where n is an integer greater than the integer 1. The sampling time between consecutive OES data acquisitions, i.e., spectrum acquisitions, can vary in the range of 0.01 to 1.0 seconds or faster. Each acquired set of OES data, i.e., spectrum, includes m measurements of light intensity corresponding to m pixels of a CCD detector, and each pixel typically corresponds to a specific light wavelength projected onto the pixel by a diffraction grating used as a light dispersing element within the light detection device 34. The CCD detector can have 256 to 8192 pixels depending on the desired spectral resolution, but most commonly 2048 or 4096 pixels are used. For example, a 2D detector having 4k×4k pixels can be used.

[0036] Next, the OES data matrix [X] [i] is set for all plasma etching process executions i = 1, 2,... k (step 215). Each matrix [X] [i] is an n×m matrix, where the acquired spectra are arranged in the rows of the matrix, each row corresponding to the n time points at which the OES data is taken, and each column corresponding to the pixel number m. Then, all matrices [X] acquired over all i = 1, 2,... k plasma etching process executions[i] By averaging each element of, an n×m averaged OES data matrix [X] avg is optionally calculated (step 220). Optionally, OES spectrum normalization may be performed before calculating the average. If k = 1, only one wafer OES measurement exists, in which case the averaged OES matrix is not calculated.

[0037] In one embodiment, the OES data matrix [X] [i] can optionally be normalized as follows. The OES data matrix [X] [i] is an n×m matrix with components x ij , where i = 1, 2,... n, each row corresponds to an OES snapshot at time point t, and j = 1, 2,... m, each column corresponds to a trend at wavelength λ, so each column is a single wavelength trend. OES data normalization can be applied in two ways. In the first method, the method selects a reference snapshot S R = x R,j at time point R (i.e., the Rth row), and then divides all OES data by the reference snapshot, x i,j = x i,j / x R,j . There may be a single time snapshot or an averaged snapshot over a period. In the second method, the method selects a reference wavelength λ R (i.e., the Rth column), and then divides all wavelengths by the intensity of the reference wavelength, x i,j = x i,j / x i,R . Similarly, there may be a single wavelength or an average of a specific band of wavelengths. The inventors have discovered that normalization resolves intensity drifts that occur during OES execution between different wafers.

[0038] Next, as detailed in the '990 patent, noise is filtered from the averaged OES data matrix [X] avg (step 225), and the matrices [X] [i] and [X] avgRounding is performed (step 230), removing the spectrum acquired during plasma startup, and optionally following the actual etching process endpoint, the average OES data matrix [S avg is calculated (step 235), and all elements of each column are set to the average over the entire column of elements of the average OES data matrix [X] avg (i.e., over all time points), and subtracted from each of the acquired OES data matrices [X] [i] i = 1, 2,... k (step 240), performing a step of mean removal, i.e., mean subtraction, before constructing a multivariate model of the acquired OES data.

[0039] Next, in a non-limiting embodiment, a method PCA for determining the principal component weights [P] used in multivariate analysis to transform OES data (see step 242 between steps 240 and 245 in FIG. 3) is described in the following steps. Other multivariate data analysis methods, such as independent component analysis (ICA) methods, may be used. PCA is an example of an unsupervised training method. Other supervised methods may be used as long as each or some of the OES spectral target values, such as partial least squares (PLS), support vector machine (SVM) regression or classification methods, are available. The target values can be obtained from xSEM, transmission electron microscopy (TEM), optical critical dimension (OCD) spectroscopy, critical dimension scanning electron microscopy (CDSEM) or other tools.

[0040] During the execution of step 1, the average spectrum of [X] is subtracted from each row (step 240 in FIG. 3), but the data is not optionally normalized using the standard deviation of [X].

[0041] During the execution of step 2, the covariance matrix cov(λ)=[σ 2 kj is calculated. The covariance matrix is m×m. For each column (each wavelength), the average

Number

Number

[0042] The covariance of row k and column j is as follows.

Number

[0043] During the execution of step 3, the eigenvectors and eigenvalues of the covariance matrix that satisfy the equation [Covariance matrix]·[Eigenvector]=[Eigenvalue]·[Eigenvector] are calculated. This is done by performing a singular value decomposition of the covariance matrix cov(λ). P’cov(λ)P = L (3) Here, L is a diagonal matrix of the eigenvalues of cov(λ), and P is a matrix of the eigenvectors of cov(λ). The eigenvalues are arranged in descending order so that the method can find the principal component weights in order of importance. For example, in a specific software, the top three (maximum five) eigenvectors are used.

[0044] Then, the mean-removed OES data [X] [i] -[S avg is used as an input to a multivariate analysis, such as PCA using the derived weight vector P derived above (step 245), to convert the OES data vector into the PCA domain.

[0045] The inventors have found that by classifying the principal component weights Pj(λj) into two separate groups corresponding to positively and negatively weighted wavelengths, separate trends Tj are generated (T + j + T - j = Tj). T + j or T - j is a single positive trend, respectively. Therefore, all conventional trend operations, such as performing snapshot normalization and taking the ratio of any of these, are applied to T + j and T -It can be easily applied to j.

[0046] In one embodiment, the vector [P] is calculated, and then the positive vector [P + and the negative vector [P - are formed. For example, [P + is formed by setting all negative values in [P] to zero, and [P - is formed by setting all positive values in [P] to zero and then taking the absolute value (i.e., converting to a positive number).

[0047] In step 245, the mean-removed OES data [X] [i] - [S avg is used together with the determined vector [P] to derive the transformed OES data into the PCA region. [T + = ([X] - [S avg )[P + and [T] = ([X] - [S avg )[P - (4)

[0048] The method described herein generates synthetic wavelengths (corresponding to positive and negative weighted natural wavelengths) to form a single signed trend (the transformed OES vector). For example, positive and negative synthetic wavelengths are formed.

Equation

Equation

[0049] The synthetic wavelengths and the resulting trend [T+ =[Λ + and [T - =[Λ - are determined, the second phase of the endpoint detection method is to use the time-evolving values [T + and [T - in the form of a function. The trends [T + and [T - are already positive signals, so even without shifting the offset that shifts the upward trend, which is all positive, and applying that offset to a new wafer in real time, a signal enhanced by the division algorithm can be obtained. For example, in one embodiment, the ratio T + 1(t) / T + 3(t) is calculated. However, in other embodiments, any other functional form such as the square of the ratio of the composite wavelengths or simply one composite wavelength can be calculated.

[0050] The goal of the first phase is to pre-calculate multivariate model parameters useful for subsequent in-situ etch endpoint detection, and various parameters are saved for later use. In step 250, to facilitate the removal of the average of the in-situ measured OES data, the average OES data matrix [S avg is stored in a volatile or non-volatile memory medium. In this step, the vector [P] of the principal component (PC) weights is stored in a volatile or non-volatile memory medium to also facilitate the rapid conversion of the in-situ measured OES data to the transformed OES data vector [T].

[0051] In some cases, the inventors have found that the calculated values of the elements T i of the transformed OES data vector [T], i.e., the principal components, do not grow to large positive or negative values but develop over time to concentrate around the value zero, so shifting them is useful for the reliability of endpoint detection. This shift is achieved in step 255, and when a measurement is taken, at least one element T iis evaluated at each point during the etching process, and the minimum value, min(T i ) of such one or more elements is found. For this purpose, time evolution data or other data from the average OES data matrix [X avg can be used. This minimum value is then stored in a volatile or non-volatile memory medium in step 260 for later use in in-situ endpoint detection, so that the minimum value min(T i ) of the elements T i of the transformed OES data vector [T] is used to shift the time evolution values of the same elements T i of the transformed OES data vector [T] calculated from the in-situ measured optical emission spectroscopy (OES) data.

[0052] The data values stored in the volatile or non-volatile memory medium are now ready to be used in the second phase, i.e., in-situ etching endpoint detection.

[0053] FIG. 4 shows an exemplary flowchart 300 of the in-situ endpoint detection process in a plasma etching processing system 100 equipped with an optical detection device 34 where the data stored in steps 250 and 260 of flowchart 200 is available.

[0054] In steps 310 and 315, the previously determined average OES data matrix [S avg and the vector [P] of principal component (PC) weights are read from the volatile or non-volatile memory medium and loaded into the memory of the controller 55 of the plasma etching processing system 10 in FIG. 1. The controller 55 performs all the in-situ calculations necessary for determining the endpoint of the plasma process. Also, when using at least one minimum value min(T i ) of the elements T i of the transformed OES data vector [T], it can be loaded from the volatile or non-volatile medium into the memory of the controller 55 in step 320.

[0055] In step 325, the substrate 25 is loaded into the plasma etching processing system 10, and plasma is formed in the processing region 45.

[0056] In step 330, the optical detection device 34 is used to acquire in-situ, i.e., during the execution of the etching process that develops over time, the OES data.

[0057] In step 335, the read average OES data matrix [S avg elements are subtracted from each acquired set of OES data, i.e., the spectrum, and the acquired spectrum is mean-removed prior to conversion using the already developed multivariate model.

[0058] In step 340, using the already developed PCA multivariate model, the mean-removed OES data is converted into the OES data vector [T] converted using Equation 4, i.e., the read vector [P] of the principal components and the principal component (PC) weights. The reason this process is extremely fast is that it involves only simple multiplications and is thus suitable for in-situ real-time calculations. The calculated element T i , for example, the natural wavelength Λ of the converted OES data vector [T + i and Λ - i can be used for endpoint detection as it develops over time (step 345).

[0059] In step 350, each time-development element Ti of the converted OES data vector [T] can be optionally differentiated to further facilitate endpoint detection using the trend variable slope data.

[0060] After the time evolution trend variable is calculated, the controller 55 of the plasma etching processing system 10 determines whether the end point has been reached (step 355). If the end point has been reached, the etching process is terminated in step 360; otherwise, the etching process continues and each end point is continuously monitored via steps 330 to 355 of the flowchart 300.

[0061] FIG. 5 shows the time evolution of the time derivative of the trend variable of the etching process. A deep, and thus easily distinguishable, minimum value is seen where the distinguished trend variable passes through the etching end point. The lowest group of the trajectories corresponds to the trend obtained using the combined wavelength applied trend Λ + 1(t) / Λ + 3(t). Various trajectories corresponding to different wafers used in different etching runs are shown. FIG. 5 includes the time evolution of the time derivative of the single wavelength trend at λ = 656 nm and the time evolution of the time derivative of the ratio of two single wavelength trends λ = 656 nm and λ = 777 nm (for different wafers). The time evolutions of other types of trends are also shown. As seen in FIG. 5, the end point occurs when about 32 seconds have elapsed in the etching process.

[0062] FIG. 6A shows the time evolution of the trend Λ + 1(t) / Λ + 3(t) of different wafers, and FIG. 6B shows the time evolution of the time derivative of the trend Λ + 1(t) / Λ + 3(t) of different wafers in another etching process run.

[0063] FIG. 7A shows the time evolution of the single wavelength trend at λ = 656 nm of different wafers, and FIG. 7B shows the time evolution of the time derivative of the single wavelength at λ = 656 nm of different wafers in another etching process run.

[0064] FIG. 8A shows the time evolution of the time derivative of the single wavelength trend at λ = 260 nm of different wafers, and FIG. 8B shows the trend Λ +1(t) / Λ + Shows the time evolution of the time derivative of the trend obtained using 3(t).

[0065] FIG. 8C shows the time evolution of the time derivative of the synthetic wavelength trend using the above normalization. The plot shows only the synthetic wavelength, not the ratio, because the normalization is applied beforehand.

[0066] After the time evolution trend variable is calculated, the controller 55 of the plasma etching processing system 10 needs to determine whether the end point has been reached in step 355. If the end point has actually been reached, the etching process is terminated in step 360; otherwise, the etching process continues and the end point of etching is continuously monitored via steps 330 to 355 of the flowchart 300.

[0067] In light of the above teachings, many improvements and modifications can be made to this application. Therefore, it should be understood that this application can be implemented within the scope of the appended claims, unless otherwise described herein.

Claims

Claim 1 A method for determining endpoint data of an etching process in a plasma processing system, comprising: performing a plasma etching process in a plasma processing chamber of the etching processing system; acquiring emission spectroscopy (OES) data from the plasma processing chamber during one or more etching processes; performing multivariate data analysis on the OES data by classifying wavelengths emitted during the plasma etching process to generate composite OES data from the OES data; using the composite OES data for in-situ determination of an etching process endpoint and a method comprising the same. Claim 2 The step of generating the composite OES data includes obtaining a transformed OES data vector [T]. [T] = ([X] - [S ａｖｇ )[P] where [X] is an OES data matrix, [P] is a weight vector, and [S ａｖｇ is an n×m average OES data matrix, each element of the n×m average OES data matrix being the average value of the n elements of the corresponding column of the [X ａｖｇ which is an n×m average OES data matrix, each element of the [X ａｖｇ being calculated as the average of the corresponding elements of the OES data matrix [X] over k executions of the etching process, n corresponding to the time point at which the OES data is taken, and m corresponding to the number of light intensities measured by the detector within the plasma processing chamber, the method according to claim 1. Claim 3 The method according to claim 2, wherein the step of generating the composite OES data includes classifying wavelengths corresponding to positive and negative weights. Claim 4 The weight vector [P] is calculating eigenvectors and eigenvalues of a covariance matrix associated with a matrix [X]; sorting the eigenvalues representing the weight vector [P] in descending order; By setting all negative components of [P] to zero, a positive weight vector [P ＋ is set, and by setting all positive components of [P] to zero and taking its absolute value, a negative weight vector [P － is set, and and is determined by the method according to claim 3. Claim 5 The step of obtaining the transformed OES database vector [T+] or [T － is further included, [T ＋ = ([X] - [S ａｖｇ )[P ＋ , [T － = ([X] - [S ａｖｇ )[P － ​ The method according to claim 4. Claim 6 The method according to claim 5, further comprising the step of selecting a functional form involving elements of the transformed OES database vector [T ＋ or [T － and calculating the time evolution of the selected functional form. Claim 7 The method according to claim 6, further comprising calculating a time derivative of the selected functional form and calculating a time evolution of the time derivative of the selected functional form. Claim 8 The function form is [T ＋ , [T － , ratio [T ＋ / [T － , ratio [T ＋ / [T － to the power of, or any mathematical form using the transformed OES database vector [T ＋ or [T － or a single element of [T ＋ and / or [T － , the method according to claim 7. Claim 9 The method according to claim 1, further comprising performing k plasma etching processes in the plasma processing chamber, where k is an integer greater than 0, and each of the k plasma etching processes includes: loading a substrate to be processed into the plasma processing chamber, wherein the plasma processing chamber includes a spectrometer having a detector including m pixels, and each pixel corresponds to a different optical wavelength; forming a plasma in the plasma etching process chamber; acquiring OES data from the plasma processing chamber during one or more etching processes and forming an OES data matrix [X] for each of the k plasma etching processes. and a method according to claim 1. Claim 10 n×m average OES data matrix [X] ａｖｇ A step of calculating, wherein each element is calculated as the average of the corresponding elements of the OES matrix [X] over the k etching process executions. The average OES data matrix [X] ａｖｇ filtering noise from Each OES data matrix [X] and [X] ａｖｇ A step of performing truncation on the data obtained during plasma startup and over the number of times exceeding the etching process end point, where the data is discarded Step of calculating an n×m average OES data matrix [S ａｖｇ , where each element is calculated as the average value of n elements of the corresponding column of [X ａｖｇ , To average-remove the OES data, for each k, subtract [S ａｖｇ from the matrix [X] The method according to claim 9, further comprising the same. Claim 11 For each execution of the plasma etching process, after forming the OES data matrix [X] and before calculating the n×m average OES data matrix [X] ａｖｇ The method according to claim 2, wherein the OES data matrix [X] is normalized. Claim 12 The OES data matrix normalization selects the reference snapshot x at time point R and then divides all OES data by the reference snapshot, x Ｒ，ｊ = x ｉ，ｊ = x ｉ，ｊ / x Ｒ，ｊ , and includes the step, the method according to claim 11 Claim 13 The method according to claim 12, wherein the reference snapshot is a single time snapshot or a snapshot averaged over a period of time.

14. The OES data matrix normalization selects a reference wavelength λ Ｒ and then divides all OES data by the intensity at the reference wavelength, x ｉ，ｊ = x ｉ，ｊ / x ｉ，Ｒ , the method according to claim 11, including the step.

15. The method according to claim 14, wherein the reference wavelength is a single wavelength or an average of band wavelengths.

16. A method for determining endpoint data of an etching process in a plasma processing system, comprising: performing a plasma etching process in a plasma processing chamber of an etching process system; acquiring optical emission spectroscopy (OES) data from the plasma processing chamber during one or more etching processes; performing multivariate data analysis on the OES data by classifying wavelengths corresponding to positive and negative weights associated with natural wavelengths to generate composite OES data from the OES data; using the composite OES data for in-situ determination of an etching process endpoint. The method includes the above steps.

17. The method according to claim 16, wherein the multivariate data analysis is performed using independent component analysis.

18. The method according to claim 16, wherein the multivariate data analysis is performed using a supervised multivariate data analysis method, and the supervised multivariate data analysis method includes support vector machine regression.

19. The step of obtaining the converted OES database vector [T+] or [T － is further included, [T ＋ = ([X] - [S ａｖｇ )[P ＋ , [T － = ([X] - [S ａｖｇ )[P － ​ where [X] is the OES data matrix, [P ＋ is the positive weight vector, [P － is the negative weight vector, and [S ａｖｇ is the n×m average OES data matrix, and each element of the n×m average OES data matrix is calculated as the average value of n elements of the corresponding column of the [X ａｖｇ which is the n×m average OES data matrix, and each element of the [X ａｖｇ is calculated as the average of the corresponding elements of the OES data matrix [X] over k executions of the etching process, n corresponds to the time point at which the OES data is taken, and m corresponds to the number of light intensities measured in the plasma processing chamber by the detector, the method according to claim 16.

20. the converted OES database vector [T ＋ or [T － , and further includes the step of calculating the time evolution of the selected functional form, the method according to claim 19.

21. The method according to claim 20, further comprising calculating a time derivative of the selected functional form and calculating a time evolution of the time derivative of the selected functional form.

22. The function form is [T ＋ , [T － , ratio [T ＋ / [T － , ratio [T ＋ / [T － to the power of, or any mathematical form using the transformed OES database vector [T ＋ or [T － , a single element of [T ＋ and / or [T － , the method according to claim 21.

Images