Filtering scanning based etching endpoint detection method and system

Abstract

The application relates to a filter scanning-based etching endpoint detection method and system. The etching endpoint detection method comprises the following steps: irradiating infrared light emitted by a light source to an absorption cell to output detection light after a target gas is absorbed; controlling a filter to perform wavelength scanning on the detection light, and identifying at least one characteristic wavelength point with an absorption peak; setting a fixed working frequency of the filter, so that the filter allows the detection light of the characteristic wavelength point to pass; scanning the detection light of the characteristic wavelength point to obtain a first light intensity vector containing interference; calibrating a sidelobe response function matrix S of the filter; calculating the first light intensity vector according to the sidelobe response function matrix S to obtain a second light intensity vector in which the interference is eliminated; calculating a characteristic concentration c of the target gas according to the second light intensity vector, and performing etching endpoint judgment. The application can improve the accuracy of etching endpoint judgment.

Description

Technical Field

[0001] This invention relates to the field of semiconductor inspection technology, and in particular to an etching endpoint detection method and system based on filtered scanning. Background Technology

[0002] In advanced semiconductor manufacturing processes, the accuracy and real-time performance of endpoint detection for critical steps such as plasma etching and chamber cleaning directly impact device performance and production yield. Traditional endpoint detection methods typically employ optical emission spectroscopy (OES) or laser interferometry (IEP), but these methods have limitations when monitoring complex chemical reactions and trace gas components. In recent years, gas detection technology based on mid-infrared spectroscopy has been increasingly introduced into online analysis of semiconductor process exhaust gases due to its ability to identify characteristic absorption peaks of various gas molecules, used to reflect the etching process and byproduct formation in real time. Acousto-optically tunable filters (AOTFs), as electrically tunable filter devices without moving parts, offer advantages such as fast scanning speed, flexible wavelength switching, and high stability, and have been explored for use in mid-infrared spectral detection systems. However, in practical applications for semiconductor exhaust gas monitoring, AOTFs exhibit the following inherent drawbacks in the mid-infrared band. First, there is a trade-off between resolution and light transmittance. Achieving high spectral resolution requires increasing the RF drive power and optimizing the crystal design, but this leads to decreased diffraction efficiency and reduced effective utilization of the aperture. When detecting low-concentration gases, the signal strength is weak, and the sharp decrease in light transmittance deteriorates the signal-to-noise ratio, making it difficult to accurately invert the gas concentration. This problem is particularly prominent in scenarios requiring high sensitivity for etching endpoint determination. Secondly, spectral interference is introduced by the sidelobe effect: the tuning characteristics of AOTF are accompanied by strong sidelobes, and the non-target wavelength radiation generated by them overlaps with the absorption spectrum of the target gas. When detecting multi-component mixed gases (such as CF4, SF6, NF3, and hydrocarbons, which are common etching products), the sidelobes cause severe spectral crosstalk. This not only reduces the recognition of characteristic peaks but may also misjudge gas composition and concentration changes, thus affecting the accuracy of endpoint detection. In addition, in the semiconductor mass production environment, the complex process gas background, temperature and pressure variations, and optical window contamination further amplify the above problems, making it difficult for existing AOTF-based detection systems to meet the requirements of high-precision and high-reliability endpoint detection.

[0003] Therefore, in view of the problems of resolution-transmittance contradiction and sidelobe interference in the application of existing acousto-optic tunable filters in the mid-infrared band for semiconductor exhaust gas detection, there is an urgent need to provide a new etching endpoint detection method and system based on filter scanning, so as to improve the sensitivity and accuracy of spectral detection in complex gas environments, thereby achieving rapid and reliable judgment of the etching process endpoint. Summary of the Invention

[0004] The technical problem to be solved by this application is to provide a method and system for detecting etching endpoints based on filtered scanning with high detection accuracy.

[0005] To address the aforementioned technical problems, according to embodiments of this application, a method for detecting etching endpoints based on filtered scanning is provided, comprising the following steps: irradiating an absorption cell with infrared light emitted from a light source to output detection light after absorbing the target gas; controlling a filter to perform wavelength scanning on the detection light to identify at least one characteristic wavelength point with an absorption peak. Set a fixed operating frequency for the filter so that it allows the characteristic wavelength point to pass through. The detection light; for the characteristic wavelength point The detection light is scanned to obtain a first light intensity vector containing interference. ; Calibrate the sidelobe response function matrix S of the filter; Apply the sidelobe response function matrix S to the first light intensity vector Calculations are performed to obtain the second light intensity vector to eliminate interference. According to the second light intensity vector Calculate the characteristic concentration c of the target gas and determine the etching endpoint.

[0006] According to an embodiment of this application, the control filter performs wavelength scanning on the detection light to identify at least one characteristic wavelength point with an absorption peak. This includes controlling the filter to apply a first integral time and greater than 10 ... The filter's radio frequency is changed sequentially with varying step sizes to collect light intensity at each wavelength, obtaining a low-resolution spectrum. The low-resolution spectrum is then analyzed to identify absorption peaks with a peak depth greater than three times the noise standard deviation, and their corresponding wavelengths are determined as the characteristic wavelengths. .

[0007] According to an embodiment of this application, setting a fixed operating frequency for the filter allows the filter to pass through the characteristic wavelength point. The detection light includes selecting the filter to pass through the characteristic wavelength point. The operating frequency for detecting light is used as the fixed operating frequency.

[0008] According to an embodiment of this application, calibrating the sidelobe response function matrix S of the filter includes controlling the center of the main channel of the filter to align with the corresponding characteristic wavelength point. The corresponding characteristic wavelength points are output sequentially using a monochromatic light source. and adjacent wavelengths, to obtain the characteristic wavelength point. and the diffraction intensity corresponding to the adjacent wavelengths ; for the diffracted light intensity Perform normalization processing to obtain the characteristic wavelength points. Sidelobe response vector The set of sidelobe response vectors Form the sidelobe response function matrix S.

[0009] According to an embodiment of this application, the intensity of the diffracted light is... Perform normalization processing to obtain the characteristic wavelength points. Sidelobe response vector ,include,

[0010]

[0011] in, The wavelength variable of the monochromatic light source; Align the main channel of the filter with the characteristic wavelength point. At that time, the intensity of the diffracted light after the monochromatic light source is diffracted by the filter; Align the main channel of the filter with the characteristic wavelength point. And the monochromatic light source and the characteristic wavelength point The maximum diffraction intensity corresponding to the same value.

[0012] According to an embodiment of this application, the first light intensity vector is adjusted based on the sidelobe response function matrix S. Calculations are performed to obtain the second light intensity vector to eliminate interference. This includes obtaining the corresponding submatrix based on the sidelobe response function matrix S. According to the submatrix Calculate the second light intensity vector to eliminate interference. ,

[0013]

[0014] in, submatrix The inverse matrix.

[0015] According to an embodiment of this application, the step of using the second light intensity vector... Calculating the characteristic concentration c of the target gas includes, based on the second light intensity vector Calculate the characteristic wavelength point First absorbance and the second absorbance of the adjacent wavelength ,

[0016]

[0017]

[0018] in, To the characteristic wavelength point Reference light intensity for the corresponding wavelength; The reference light intensity is the wavelength corresponding to the adjacent wavelength. To the characteristic wavelength point The corresponding second light intensity vector ; The second light intensity vector corresponding to adjacent wavelengths ; for the first absorbance and the second absorbance Perform the difference calculation to obtain the differential absorbance. According to the differential absorbance Perform concentration inversion to obtain the characteristic concentration c.

[0019]

[0020] Where L is the absorption optical path length; At temperature T, the characteristic wavelength point Absorption cross section at the location; Let T be the absorption cross section at adjacent wavelengths.

[0021] According to an embodiment of this application, the etching endpoint determination includes setting a concentration threshold; when the characteristic concentration c is less than or equal to the concentration threshold and continues for a set time, it is determined that the etching endpoint has been reached.

[0022] An etching endpoint detection system based on filtered scanning is used in the above-mentioned etching endpoint detection method. The etching endpoint detection system includes: a light source for emitting infrared light; an absorption cell connected to the exhaust gas pipe of a semiconductor device and having a light inlet and a light outlet; the infrared light emitted by the light source illuminates the absorption cell through the light inlet, the exhaust gas in the absorption cell absorbs part of the infrared light, and the remaining infrared light is emitted through the light outlet to form a detection light beam; a filter for receiving and modulating the detection light beam; and a detector for receiving the detection light beam modulated by the filter and calculating the characteristic concentration c of the target gas to determine the etching endpoint.

[0023] According to an embodiment of this application, it further includes a collimating lens, a polarizing mirror, and a focusing lens; the collimating lens and the polarizing mirror are both disposed between the filter and the light output port, so that the detection light passes through the light output port, passes through the collimating lens and the polarizing mirror in sequence, and is received by the filter; the focusing lens is disposed between the filter and the detector, so as to focus the detection light modulated by the filter onto the detector.

[0024] By employing the above technical solution, an infrared broadband light source illuminates the absorption cell connected to the tail gas of the etching cavity. A filter is used to scan and dynamically identify the characteristic absorption peaks of the target gas, ensuring that the detection is always focused on the most effective spectral region. Subsequently, the radio frequency is locked to the corresponding peak position, and a long integral is performed on the diffracted light to obtain the first intensity vector containing sidelobe crosstalk. The second intensity vector is calculated based on the sidelobe response matrix S. This allows for the calculation of the characteristic concentration c of the target gas and the determination of the endpoint, thereby improving the accuracy of endpoint determination. Attached Figure Description

[0025] Figure 1 This is a step diagram of an etching endpoint detection method according to an embodiment of the present invention.

[0026] Figure 2 This is an optical path diagram of an etching endpoint detection system according to an embodiment of the present invention.

[0027] Figure label:

[0028] 110. Light source; 120. Absorption cell; 130. Filter; 140. Detector; 150. Collimating lens; 160. Polarizing mirror; 170. Focusing lens. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art. The terms "comprising" and similar expressions used herein mean that the element or object preceding the word covers the element or object listed after the word and its equivalents, but does not exclude other elements or objects.

[0030] The following is in conjunction with the appendix Figure 1-2 The specific embodiments of the present invention will be further described in detail below.

[0031] Embodiments of the present invention provide an etching endpoint detection method and system based on filtered scanning. The etching endpoint detection method is used to detect the etching endpoint during wafer processing. Specifically, the etching endpoint detection system includes: a light source 110 for emitting infrared light; an absorption cell 120 connected to the exhaust pipe of the semiconductor equipment and having an inlet and an outlet; the infrared light emitted by the light source 110 illuminates the absorption cell 120 through the inlet, the exhaust gas in the absorption cell 120 absorbs part of the infrared light, and the remaining infrared light is emitted through the outlet to form a detection beam; a filter 130 for receiving and modulating the detection beam for filtered scanning; and a detector 140 for receiving the detection beam modulated by the filter 130 and calculating the characteristic concentration c of the target gas for etching endpoint determination. An off-axis parabolic mirror is also provided between the light source 110 and the absorption cell 120 to reflect the light emitted by the light source 110 back to the inlet and into the absorption cell 120. This is well known to those skilled in the art and is not limited herein.

[0032] In some embodiments, the light source 110 emits a broadband infrared light source 110 (such as a silicon carbide rod or a quantum cascade laser), with the emitted light covering the mid-to-far infrared band (8-15μm), adapted to the absorption characteristics of semiconductor gases. The absorption cell 120 is connected to the exhaust pipe of the semiconductor device and has a light inlet and a light outlet; the infrared light emitted by the light source 110 illuminates the absorption cell 120 through the light inlet, the exhaust gas in the absorption cell 120 absorbs part of the infrared light, and the remaining infrared light is emitted through the light outlet to form the detection light. The absorption cell 120 is well known to those skilled in the art and will not be described in detail here.

[0033] In some embodiments, the filter 130 is disposed on the light outlet side for receiving and modulating the detection light; the detector 140 is disposed on the light outlet side, so that the filter 130 is placed between the detector 140 and the light outlet. The detector 140 is used to receive the detection light after passing through the filter and to analyze and calculate the detection light to calculate the characteristic concentration c of the target gas for etching endpoint determination.

[0034] In some embodiments, the etching endpoint detection system further includes a collimating lens 150, a polarizing mirror 160, and a focusing lens 170; the collimating lens 150 and the polarizing mirror 160 are both disposed between the filter 130 and the light outlet, so that the detection light passes through the light outlet, passes through the collimating lens 150 and the polarizing mirror 160 in sequence, and is then received by the filter 130; the polarizing mirror 160 only allows the polarization component parallel to the incident surface of the filter 130 to pass through, and the focusing lens 170 is disposed between the filter 130 and the detector 140 to focus the detection light modulated by the filter 130 onto the detector 140.

[0035] The etching endpoint detection method disclosed in this application includes the following steps:

[0036] S1. Infrared light emitted by light source 110 is irradiated into absorption cell 120 to output detection light after absorbing target gas;

[0037] S2, the control filter 130 performs wavelength scanning on the detection light to identify at least one characteristic wavelength point with an absorption peak. ;

[0038] S3. Set the fixed operating frequency of filter 130 so that filter 130 allows the characteristic wavelength point to pass through. The detection light;

[0039] S4. For characteristic wavelength points The detection light beam is scanned to obtain the first light intensity vector containing interference. ;

[0040] S5, the sidelobe response function matrix S of calibration filter 130;

[0041] S6. Based on the sidelobe response function matrix S, the first intensity vector... Calculations are performed to obtain the second light intensity vector to eliminate interference. ;

[0042] S7, According to the second light intensity vector Calculate the characteristic concentration c of the target gas and determine the etching endpoint.

[0043] In some embodiments, during the determination of the etching endpoint, it is necessary to detect the change in the concentration of the target gas in the exhaust gas discharged from the semiconductor device; that is, the infrared light emitted by the light source 110 is irradiated into the absorption cell 120 to absorb the detection light after the target gas is absorbed; wherein, the light source 110 is a broadband infrared light source 110 (such as a silicon carbide rod or a quantum cascade laser), so that the light emitted covers the mid- and far-infrared band (e.g., 8-15μm), thereby adapting to the absorption characteristics of semiconductor gases, such as the target gas in the exhaust gas. The absorption wavelength is Target gas The absorption wavelength is Meanwhile, after the exhaust gas interacts with infrared light in the absorption cell 120, only specific wavelengths are absorbed by the corresponding gas. Therefore, the detection light emitted after passing through the absorption cell 120 carries information about the target gas, which facilitates detection.

[0044] In some embodiments, the detection light, after absorbing the target gas in the absorption cell 120, illuminates the filter 130. The filter 130 is then controlled to perform a wavelength scan on the detection light to determine which target gases are present in the exhaust gas and the wavelengths of their absorption peaks. This allows the acquisition of one or more wavelength positions in the infrared band of the target gas that exhibit the strongest absorption and minimal overlap with other gases, i.e., characteristic wavelength points. .

[0045] In some embodiments, after obtaining the characteristic wavelength points Then, a fixed operating frequency is set for filter 130, so that the operating frequency of filter 130 is close to the characteristic wavelength. The operating frequency of the point-time filter 130 is consistent with that of the filter 130, thus allowing the filter 130 to pass through the characteristic wavelength point. The detection light is used to detect specific target gases. Specifically, at characteristic wavelength points... After the detection light beam is passed into the filter 130, the filter 130 modulates the corresponding detection light beam and selects the characteristic wavelength point. The detection light beam is scanned and the scanned data is transmitted to the detector 140. The detector 140 receives the data scanned by the filter 130 and calculates the first light intensity vector. The first light intensity vector at this time This is the light intensity vector that includes interference.

[0046] In some specific embodiments, a fixed operating frequency is set for the filter 130, allowing the filter 130 to pass through characteristic wavelength points. The detection light includes selecting filter 130 to pass through the characteristic wavelength point. The operating frequency for detecting light is used as a fixed operating frequency.

[0047] In some embodiments, after obtaining the first light intensity vector The sidelobe response function matrix S of filter 130 needs to be calibrated; after calibrating the sidelobe response function matrix S of filter 130, the first light intensity vector is adjusted according to the sidelobe response function matrix S. Calculations are performed to obtain the second light intensity vector to eliminate interference. ; and according to the second light intensity vector Calculate the characteristic concentration c of the target gas and determine the etching endpoint.

[0048] In some specific embodiments, the control filter 130 performs wavelength scanning on the detection light to identify at least one characteristic wavelength point with an absorption peak. This includes controlling the filter 130 to guide the detection light within a predetermined wavelength band with a first integral time and greater than... The step size of the filter 130 kHz frequency is changed sequentially to collect the light intensity at each wavelength point and obtain a low-resolution spectrum. The low-resolution spectrum is analyzed to identify absorption peaks with a peak depth greater than 3 times the noise standard deviation, and the corresponding wavelength points are determined as characteristic wavelength points. .

[0049] In some specific embodiments, because the energy of filter 130 is limited, scanning with a small step size would result in a long scanning time, and the etching process would be completed after the scanning, making detection impossible. Therefore, a larger step size and a shorter integration time are used for scanning to obtain a low-resolution spectrum. The low-resolution spectrum is then analyzed to identify abrupt changes, and these abrupt changes are detected more precisely to determine the characteristic wavelength points. To achieve wavelet detection.

[0050] In some more specific embodiments, the predetermined band is selected as Within this band, the target gas The maximum absorption band is Target gas The maximum absorption band is Both fall within this interval and do not overlap, allowing for simultaneous monitoring; simultaneously, filter 130 exhibits optimal sidelobe suppression ratio within this band; the first integration time is set to 1 ms, and this data is used to acquire the low-resolution spectrum. The wavelengths corresponding to absorption peaks with peak depths greater than 3 times the noise standard deviation in the low-resolution spectrum are determined as characteristic wavelength points. Among them, the noise standard deviation in 3 times the noise standard deviation is the standard deviation of the flat area.

[0051] In some embodiments, calibrating the sidelobe response function matrix S of the filter 130 includes aligning the center of the main channel of the filter 130 with the corresponding characteristic wavelength point. Using a monochromatic light source 110, the corresponding characteristic wavelength points are output sequentially. and adjacent wavelengths, to obtain characteristic wavelength points. and the diffraction intensity corresponding to adjacent wavelengths ; for diffracted light intensity Perform normalization processing to obtain characteristic wavelength points Sidelobe response vector Set of sidelobe response vectors Form the sidelobe response function matrix S.

[0052] In some specific embodiments, filter 130 can only allow one wavelength to undergo high-frequency diffraction at a time, therefore it is necessary to target the characteristic wavelength point. Only by aligning with the center of the main channel of the control filter 130 can the characteristic wavelength point be targeted. Detection is performed on itself and adjacent wavelengths. The center of the main channel of the control filter 130 is aligned with the corresponding characteristic wavelength point. Then, the corresponding characteristic wavelength points are output sequentially using a monochromatic light source 110. and adjacent wavelengths, wherein broadband light contains multiple characteristic wavelength points. Therefore, the corresponding characteristic wavelength point cannot be determined. The light intensity, while using a monochromatic light source 110 can input only a single characteristic wavelength point each time. This facilitates subsequent calculations and ensures the accuracy of the calculations.

[0053] In some specific embodiments, characteristic wavelength points and the diffraction intensity corresponding to adjacent wavelengths It can be directly detected by detector 140, which will not be elaborated here; after obtaining the diffracted light intensity After that, it is necessary to adjust the intensity of the diffracted light. Perform normalization processing to obtain characteristic wavelength points Sidelobe response vector Specifically, without sidelobe response vectors... Therefore, it is impossible to obtain the characteristic wavelength point. During detection, how many pairs of light from adjacent wavelengths are at the characteristic wavelength point? The detection process is disrupted, leading to significant errors. More specifically, this occurs when acquiring all characteristic wavelength points. Sidelobe response vector Then, the obtained sidelobe response vectors The sidelobe response function matrix S is formed, thus completing the calibration of the sidelobe response function matrix S.

[0054] In some embodiments, the intensity of the diffracted light is... Perform normalization processing to obtain characteristic wavelength points Sidelobe response vector ,include,

[0055]

[0056] in, For the wavelength variable of the monochromatic light source 110; Align the main channel of filter 130 with the characteristic wavelength point. At that time, the intensity of the diffracted light after the monochromatic light source 110 is diffracted by the filter 130; Align the main channel of filter 130 with the characteristic wavelength point. Furthermore, the monochromatic light source 110 and the characteristic wavelength point The maximum diffraction intensity corresponding to the same value.

[0057] In some specific embodiments, the diffracted light intensity is normalized. This is transformed into a dimensionless ratio with a value ranging from 0 to 1. This ratio will not change when the equipment or light source 110 is replaced in the future, thus ensuring that the sidelobe response function matrix S is a universal matrix, which facilitates the acquisition of the subsequent second light intensity vector. .

[0058] In some embodiments, the first intensity vector is adjusted according to the sidelobe response function matrix S. Calculations are performed to obtain the second light intensity vector to eliminate interference. ,include,

[0059] Obtain the corresponding submatrix from the sidelobe response function matrix S. ;

[0060] According to the submatrix Calculate the second light intensity vector to eliminate interference. ,

[0061]

[0062] in, submatrix The inverse matrix.

[0063] In some specific embodiments, during the calibration phase, a total of m wavelengths were scanned, and the resulting sidelobe response function matrix S was m. The matrix is ​​a partial matrix, but in actual detection, only n absorption peaks are detected, where n < m. Therefore, it is necessary to extract the portion corresponding to the n detected absorption peaks from the sidelobe response function matrix S to obtain the corresponding submatrix. At this time, the submatrix For n The matrix.

[0064] In some specific embodiments, due to the sidelobe effect of filter 130 introducing energy from adjacent wavelengths into the current channel, the first intensity vector... The second light intensity vector The linear superposition between the diffracted light intensity and external interference factors, while the aforementioned steps will reduce the diffracted light intensity. This is transformed into a dimensionless ratio with a value ranging from 0 to 1. This ratio was measured and fixed during calibration. Therefore, the superposition quantity is decomposed according to the inverse matrix, thereby restoring the interference-free second light intensity vector. .

[0065] In some embodiments, based on the second light intensity vector Calculating the characteristic concentration c of the target gas includes, based on the second light intensity vector Calculate characteristic wavelength points First absorbance and the second absorbance at adjacent wavelengths Through characteristic wavelength points First absorbance The target gas can be identified, and the second absorbance at adjacent wavelengths... As a background factor, by calculating the difference between the two, factors such as background drift, light source jitter, and window contamination can be completely offset, thereby improving the accuracy of the calculation of the characteristic concentration c of the target gas. Specifically, the calculation formula is as follows.

[0066]

[0067]

[0068] in, To at the characteristic wavelength point Reference light intensity for the corresponding wavelength; The reference light intensity is the wavelength corresponding to the adjacent wavelength. To at the characteristic wavelength point The corresponding second light intensity vector ; The second intensity vector corresponding to adjacent wavelengths .

[0069] For the first absorbance Second absorbance Perform the difference calculation to obtain the differential absorbance. ;

[0070] Based on differential absorbance Perform concentration inversion to obtain the characteristic concentration c.

[0071]

[0072] Where L is the absorption optical path length; At temperature T, the characteristic wavelength point Absorption cross section at the location; Let T be the absorption cross-section at adjacent wavelengths at temperature T. Through concentration inversion, the second intensity vector can be used to determine the absorption cross-section. Obtain the characteristic concentration c, which facilitates comparison with the threshold.

[0073] In some embodiments, the etching endpoint determination includes setting a concentration threshold; when the characteristic concentration c is less than or equal to the concentration threshold and continues for a set time, it is determined that the etching endpoint has been reached.

[0074] In some specific embodiments, different concentration thresholds are set for different gases, with a range of 5-40 ppm. For example, the target gas in the same chamber and the same etching step. With target gas The endpoint of both processes signifies the end of etching; therefore, the concentration thresholds for both can be set to 10 ppm for easier subsequent logic checks. If the process requires stopping the flow first... Then stop the flow. It can target gas The concentration threshold is set to 10 ppm for the target gas. The concentration threshold is set to 5 ppm. The time is set to 0.5-5 seconds. When the characteristic concentration c of the target gas is less than or equal to the concentration threshold and remains so for the set time, the etching endpoint is reached, and the etching process ends.

[0075] The implementation principle of the etching endpoint detection method and system based on filtered scanning in this application embodiment is as follows: an infrared broadband light source 110 illuminates an absorption cell 120 connected to the tail gas of the etching cavity; a filter 130 scans and identifies the characteristic absorption peaks of the target gas; then, the radio frequency is locked to the corresponding peak position; and a long integral is performed on the diffracted light to obtain the first light intensity vector containing sidelobe crosstalk. The second intensity vector is calculated based on the sidelobe response matrix S. This allows for the calculation of the characteristic concentration c of the target gas, facilitating endpoint determination and thus improving the accuracy of endpoint determination.

[0076] While embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations can be made to these embodiments. However, it should be understood that such modifications and variations fall within the scope and spirit of the invention as set forth in the claims. Furthermore, the invention described herein may have other embodiments and can be implemented or carried out in various ways.

Claims

1. A method for detecting etching endpoints based on filtered scanning, characterized in that, Includes the following steps: Infrared light emitted from the light source is directed onto the absorption cell to output the detection light after absorbing the target gas; The control filter performs wavelength scanning on the detection light to identify at least one characteristic wavelength point with an absorption peak. ; The filter is set to a fixed operating frequency so that it allows the characteristic wavelength point to pass through. The detection light; For the characteristic wavelength point The detection light is scanned to obtain a first light intensity vector containing interference. ; The center of the main channel of the filter is aligned with the corresponding characteristic wavelength point. ; The corresponding characteristic wavelength points are output sequentially using a monochromatic light source. and adjacent wavelengths, to obtain the characteristic wavelength point. and the diffraction intensity corresponding to the adjacent wavelengths ; Regarding the diffracted light intensity Perform normalization processing to obtain the characteristic wavelength points. Sidelobe response vector ; Set the sidelobe response vectors Form the sidelobe response function matrix S; Based on the sidelobe response function matrix S, the first light intensity vector Calculations are performed to obtain the second light intensity vector to eliminate interference. ; According to the second light intensity vector Calculate the characteristic concentration c of the target gas and determine the etching endpoint.

2. The etching endpoint detection method according to claim 1, characterized in that, The control filter performs wavelength scanning on the detection light to identify at least one characteristic wavelength point with an absorption peak. ,include, The filter is controlled to apply the detection light within a predetermined wavelength band with a first integral time and greater than... The step size is changed sequentially to change the radio frequency of the filter, and the light intensity at each wavelength point is collected to obtain a low-resolution spectrum; The low-resolution spectrum is analyzed to obtain absorption peaks with a peak depth greater than 3 times the noise standard deviation, and the corresponding wavelength points are determined as the characteristic wavelength points. .

3. The etching endpoint detection method according to claim 1, characterized in that, The filter is set to a fixed operating frequency so that it allows the characteristic wavelength point to pass through. The detection light includes, Select the filter through the characteristic wavelength point The operating frequency for detecting light is used as the fixed operating frequency.

4. The etching endpoint detection method according to claim 1, characterized in that, The intensity of the diffraction light Perform normalization processing to obtain the characteristic wavelength points. Sidelobe response vector ,include, in, The wavelength variable of the monochromatic light source; Align the main channel of the filter with the characteristic wavelength point. At that time, the intensity of the diffracted light after the monochromatic light source is diffracted by the filter; Align the main channel of the filter with the characteristic wavelength point. And the monochromatic light source and the characteristic wavelength point The maximum diffraction intensity corresponding to the same value.

5. The etching endpoint detection method according to claim 1, characterized in that, The first light intensity vector is adjusted according to the sidelobe response function matrix S. Calculations are performed to obtain the second light intensity vector to eliminate interference. ,include, Obtain the corresponding submatrix based on the sidelobe response function matrix S. ; According to the submatrix Calculate the second light intensity vector to eliminate interference. , in, submatrix The inverse matrix.

6. The etching endpoint detection method according to claim 1, characterized in that, According to the second light intensity vector Calculating the characteristic concentration c of the target gas includes, According to the second light intensity vector Calculate the characteristic wavelength point First absorbance and the second absorbance of the adjacent wavelength , in, To the characteristic wavelength point Reference light intensity for the corresponding wavelength; The reference light intensity is the wavelength corresponding to the adjacent wavelength. To the characteristic wavelength point The corresponding second light intensity vector ; The second light intensity vector corresponding to adjacent wavelengths ; For the first absorbance and the second absorbance Perform the difference calculation to obtain the differential absorbance. ; According to the differential absorbance Perform concentration inversion to obtain the characteristic concentration c. Where L is the absorption optical path length; At temperature T, the characteristic wavelength point Absorption cross section at the location; Let T be the absorption cross section at adjacent wavelengths.

7. The etching endpoint detection method according to claim 1, characterized in that, The process of determining the etching endpoint includes, Set a concentration threshold; When the characteristic concentration c is less than or equal to the concentration threshold and continues for a set time, it is determined that the etching endpoint has been reached.

8. An etching endpoint detection system based on filtered scanning, characterized in that, For implementing the etching endpoint detection method according to any one of claims 1-7, the etching endpoint detection system comprises: A light source used to emit infrared light; An absorption cell is connected to the exhaust pipe of a semiconductor device and has a light inlet and a light outlet. Infrared light emitted by the light source is irradiated into the absorption cell through the light inlet. The exhaust gas in the absorption cell absorbs part of the infrared light, and the remaining infrared light is emitted through the light outlet to form a detection light beam. A filter for receiving and modulating the detection light; The detector receives the detection light modulated by the filter and calculates the characteristic concentration c of the target gas to determine the etching endpoint.

9. The etching endpoint detection system according to claim 8, characterized in that, It also includes a collimating lens, a polarizing mirror, and a focusing lens; the collimating lens and the polarizing mirror are both disposed between the filter and the light outlet, so that the detection light passes through the light outlet and then sequentially through the collimating lens and the polarizing mirror before being received by the filter; the focusing lens is disposed between the filter and the detector to focus the detection light modulated by the filter onto the detector.

Images