Photoetching method and photoetching machine

Abstract

The invention discloses a photoetching method and a photoetching machine. The photoetching method comprises the following steps: acquiring flatness information of a wafer surface and bending information of a photomask; according to the flatness information, determining position information of the wafer when different positions of the surface of the wafer are located on a focusing plane of a photomask pattern under the condition that the photomask is not bent; compensating the position information based on the bending information to obtain new position information; and carrying out photoetching on the wafer at the position indicated by the new position information by utilizing the photomask. According to the method and the device, the photomask bending can be accurately compensated, so that the influence of the photomask bending on the key size uniformity and the alignment deviation compensation value uniformity in the exposure unit is eliminated.

Description

Technical Field

[0001] This application relates to the field of semiconductor manufacturing technology, and in particular to a photolithography method and a photolithography machine. Background Technology

[0002] On the lithography machine, the two ends of the photomask are vacuum-adhered to guide rails. Under the influence of gravity, the middle part of the photomask sags, causing it to bend. The degree of bending increases with the duration of use. Figure 1 This is illustrated by the use of a photomask with a larger bending radius for wafers exposed later. Figure 1 The dashed line represents a horizontal line. The photomask used for the first exposure of the wafer is not bent. The bending degree of the photomask used for the second to Nth exposures (N is an integer greater than 2) increases progressively. When light passes through a bent photomask, the focal point formed on the wafer surface in different areas will change differently compared to an unbent photomask. This results in different critical dimensions (CD) at different locations within the exposure unit, and the overlay shift compensation value at different locations within the exposure unit will also be different.

[0003] In existing technologies, lithography machines can detect height data at multiple locations on the photomask surface and use this height data to fit the curvature information of the photomask surface. Then, by adjusting the height difference between the guide rails attached to both ends of the photomask using an anti-vibration device, the curvature indicated by the curvature information is compensated, thereby correcting the photomask deformation. Specifically, if Figure 1 The left edge of the photomask shown in the diagram is attached to the first guide rail, and the right edge is attached to the second guide rail. When photolithography is performed using the pattern on the left half of the photomask, the first guide rail is controlled to be lower than the second guide rail by the anti-vibration device. When photolithography is performed using the pattern on the right half of the photomask, the first guide rail is controlled to be higher than the second guide rail by the anti-vibration device, thereby compensating for the curvature shown in the flatness information.

[0004] However, the vibration damping device does not directly contact the guide rail of the adsorption photomask. The guide rail of the adsorption photomask moves under the drive of the motor by means of air flotation. There is an air flotation medium between the vibration damping device and the guide rail of the adsorption photomask, as well as a base plate used to support the guide rail when the air flotation support is not present. Therefore, the compensation performed by the vibration damping device has errors and the compensation effect is poor. Summary of the Invention

[0005] In view of the above problems, this application provides a photolithography method and a photolithography machine, which aims to accurately compensate for the bending of the photomask, thereby eliminating the influence of the bending of the photomask on the uniformity of critical dimensions and the uniformity of alignment deviation compensation value within the exposure unit.

[0006] According to a first aspect of the present invention, a photolithography method is provided, comprising: Obtain information on the flatness of the wafer surface and the curvature of the photomask; The flatness information is used to determine the position information of the wafer when different positions on the wafer surface are on the focal plane of the photomask pattern under the condition that the photomask is not bent; The position information is compensated based on the curvature information to obtain new position information; The wafer at the position indicated by the new position information is photolithographically etched using the photomask.

[0007] Optionally, compensating for the position information based on the curvature information includes: The focusing surface of the photomask under bending conditions is determined based on the bending information; Obtain key dimensional offset information of the first focusing pattern relative to the second focusing pattern. The first focusing pattern is the focusing pattern of the photomask on the focusing surface when the photomask is bent, and the second focusing pattern is the focusing pattern of the photomask on the focusing plane when the photomask is not bent. When the third focusing pattern of the photomask pattern has the key dimension offset information compared to the first focusing pattern when the photomask is bent, the position offset information of the surface where the third focusing pattern is located compared to the focusing surface is compensated to the position information. The third focusing pattern is the focusing pattern of the photomask pattern on the compensated surface when the photomask is bent.

[0008] Optionally, the photolithography method further includes: Determine whether the positional offset information of the surface where the third focusing pattern is located relative to the focusing surface is greater than half of the focusing depth of the photolithography lens; If the depth of focus is greater than half of the depth of focus, the position offset information of the surface where the third focus pattern is located relative to the focus surface is compensated to the position information.

[0009] Optionally, the photolithography method further includes: when the positional offset information of the surface where the third focusing pattern is located relative to the focusing surface is not greater than half of the focusing depth, performing photolithography on the wafer at the position indicated by the positional information using the photomask.

[0010] Optionally, obtaining the flatness information of the wafer surface includes: obtaining a detection signal for detecting the height of the wafer surface, filtering out noise in the detection signal, and determining the flatness information of the wafer surface based on the detection signal after filtering out noise.

[0011] Optionally, determining the flatness information of the wafer surface based on the detection signal after noise filtering includes: Determine whether the detection signal after noise filtering includes a first sub-signal, wherein the first sub-signal indicates that the detected wafer surface protrudes and the protrusion height is greater than a height threshold; If the detection signal after noise filtering includes the first sub-signal, the first sub-signal is removed from the detection signal after noise filtering, and the flatness information of the wafer surface is determined based on the detection signal after removing the first sub-signal.

[0012] Optionally, determining the flatness information of the wafer surface based on the detection signal after noise filtering includes: Determine whether the detection signal after noise filtering includes a second sub-signal, wherein the second sub-signal indicates that the detected wafer surface has a protruding position and the continuous width of the protruding position is greater than a width threshold. If the detection signal after noise filtering includes the second sub-signal, the second sub-signal is removed from the detection signal after noise filtering, and the flatness information of the wafer surface is determined based on the detection signal after removing the second sub-signal.

[0013] Optionally, the bending information of the photomask is obtained, including: Obtain height data at multiple locations of the photomask; Calculate the difference between the height data and the height reference value of the photomask, where the height reference value is the height of the preset plane of the photomask; The morphology curve of the photomask, including the bending information, is fitted using multiple calculated differences.

[0014] Optionally, the photomask is supported by opposing first and second edges to be positioned on the preset plane, and the photomask includes a first region from the first edge to the middle boundary line and a second region from the second edge to the middle boundary line; Acquiring height data at multiple locations of the photomask includes: acquiring height data at multiple locations within a target area, wherein the target area is either the first area or the second area; Fitting the topography curve of the photomask, which includes the bending information, using multiple calculated differences includes: fitting the topography curve of the target region using multiple calculated differences, and determining the topography curve of the photomask based on the topography curve of the target region.

[0015] According to a second aspect of the present invention, a lithography machine is provided, comprising: a processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of any of the lithography methods described in the first aspect.

[0016] The unexpected technical effect of this application is: The photolithography method provided in this application includes: acquiring flatness information of the wafer surface and bending information of the photomask; determining the wafer's position information when different positions on the wafer surface are located on the focal plane of the photomask pattern under unbent conditions, based on the flatness information; compensating the position information based on the bending information to obtain new position information; and performing photolithography on the wafer located at the position indicated by the new position information using the photomask. Therefore, the bending information of the photomask is directly compensated into the wafer's position information determined based on the wafer surface flatness information. By adjusting the wafer's position during the photolithography process for different positions on the wafer surface, the bending of the photomask is accurately compensated, thereby eliminating the influence of photomask bending on the uniformity of critical dimensions and the uniformity of alignment deviation compensation within the exposure unit. Attached Figure Description

[0017] The above and other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which: Figure 1 A schematic diagram showing the change in the bending amplitude of the photomask as the duration of photomask use increases; Figure 2 A schematic diagram of a partial structure of an exemplary photolithography machine is shown; Figure 3 This document shows a flowchart of a photolithography method provided in an embodiment of this application; Figure 4 A schematic diagram showing a plurality of grids divided in a wafer according to an embodiment of the present application; Figure 5 An exemplary curved photomask is shown according to an embodiment of this application; Figure 6a This illustration shows a protruding area on the wafer surface according to an embodiment of this application and a corresponding exemplary detection signal; Figure 6b This illustrates a noise signal present in the detection signal according to an embodiment of this application; Figure 6c This illustrates a first sub-signal present in the detection signal according to an embodiment of this application; Figure 6d This illustrates a second sub-signal present in the detection signal according to an embodiment of this application; Figure 7 This diagram illustrates an exemplary photomask pattern according to an embodiment of the present application, which compensates for photomask bending information. Figure 8 An exemplary Poisson curve is shown according to an embodiment of this application.

[0018] Figure reference numerals: 10-Lithography machine; 11-Laser source; 12-Beam expander; 13-Photomask; 14-Lens; 15-Supporting platform; 16-Wafer. Detailed Implementation

[0019] The present application will now be described in more detail with reference to the accompanying drawings. In the various drawings, the same elements are indicated by similar reference numerals. For clarity, the various parts in the drawings are not drawn to scale. Furthermore, some well-known parts may not be shown.

[0020] This application may be presented in various forms, some of which will be described below.

[0021] In semiconductor manufacturing, circuit design is performed first, and then the circuit design is used to form a layout pattern. This layout pattern is then transferred to a photomask. After the pattern is formed on the photomask, photolithography is used to transfer the photomask pattern to photoresist on the wafer surface. Next, etching is used to transfer the pattern on the photoresist to the wafer. Combined with processes such as ion implantation and metal deposition, the required semiconductor integrated circuit can be formed on the wafer.

[0022] Figure 2 The diagram shown is a partial structural schematic of an exemplary lithography machine. Figure 2 As shown, the lithography machine 10 includes a laser source 11, a beam expander 12, a photomask 13, a lens 14, and a support platform 15. Parallel light emitted from the laser source 11 is expanded by the beam expander 12 and then irradiates the photomask 13. The beam passing through the photomask 13 is focused by the lens 14 onto the wafer 16 supported by the support platform 15, thus exposing the photoresist on the surface of the wafer 16. The photomask 13 includes a light-shielding area and a light-transmitting area. The photoresist is divided into positive photoresist and negative photoresist according to its reaction mode. The exposed portion of the positive photoresist is removed during development to form an opening area, while the exposed portion of the negative photoresist is retained during development to form a blocking area. Therefore, the opening area of ​​the positive photoresist corresponds to the light-transmitting area of ​​the photomask 13, and the opening area of ​​the negative photoresist corresponds to the light-shielding area of ​​the photomask 13. Thus, the pattern constructed on the photomask 13 by the light-shielding and light-transmitting areas is projected onto the photoresist on the surface of the wafer 16.

[0023] Typically, the pattern on the photomask 13 is scaled down and projected onto the photoresist on the surface of the wafer 16. The fixed spacing between the photomask 13 and the lens 14 achieves a fixed scaling factor. When the size of the photomask 13 is 5 inches or 6 inches and the scaling factor is 4x or 5x, the exposure area obtained in one exposure is less than 50mm, usually around 30mm square. For wafers 16 ranging from 2 inches to 12 inches, one exposure cannot complete the exposure of the entire area. Multiple exposures are required to cover the entire wafer 16. Therefore, the wafer 16 is divided into multiple exposure units (shots), and one exposure unit is exposed at a time. The support platform 15 moves in the xy plane to move the wafer 16 in the xy plane, so that multiple exposure units are exposed sequentially.

[0024] In the lithography machine 10, the two ends of the photomask 13 are vacuum-adhered to the guide rails. Under the influence of gravity, the middle part of the photomask 13 sags, causing the photomask 13 to bend. Furthermore, the degree of bending of the photomask 13 increases with the duration of its use. Figure 1 The photomask 13 shown is not bent during the first exposure of the wafer, but it bends during subsequent exposures, with the bending amplitude increasing. When the photomask 13 is bent, the focal point of different areas on the surface of the wafer 16 after the light beam passes through the photomask 13 will change differently compared to when the photomask 13 is not bent. This results in different critical dimensions at different locations within the exposure unit, and the alignment deviation compensation at different locations within the exposure unit will also vary with the different critical dimensions.

[0025] In the prior art, the placement of the photomask 13 is adjusted by adjusting the height difference between the guide rails adsorbed at both ends of the photomask 13 using an anti-vibration device, thereby compensating for the bending of the photomask 13. However, the anti-vibration device does not directly contact the guide rails adsorbing the photomask, so the compensation performed by the anti-vibration device has errors and the compensation effect is poor.

[0026] In view of this, one embodiment of this application provides a photolithography method that directly compensates the bending information characterizing the bending amplitude of the photomask into the wafer position information determined based on the flatness information of the wafer surface. By adjusting the position of the wafer during the photolithography process for different positions on the wafer surface, the bending of the photomask can be accurately compensated, thereby eliminating the influence of photomask bending on the uniformity of key dimensions and the uniformity of alignment deviation compensation value within the exposure unit.

[0027] Figure 3 The diagram shows a flowchart of the photolithography method provided in an embodiment of this application. Figure 3 As shown, the photolithography method includes: Step S110: Obtain the flatness information of the wafer surface and the bending information of the photomask.

[0028] Specifically, the flatness information of the wafer surface can be composed of height data at different locations on the wafer surface, which can be used to represent the wafer surface as follows: Figure 4 The wafer surface is divided into multiple grids, each with a height data point. The more grids there are, the more accurately the flatness information, composed of the height data from multiple grids, reflects the flatness of the wafer surface. The height data in the flatness information is correlated with the position of the wafer surface represented by that height data; therefore, the flatness information can also be called a flatness map.

[0029] The bending information of the photomask can be composed of bending data at different positions of the photomask. The bending data at one position is the difference between the height data of the photomask at that position and the height reference value of the photomask. The height reference value is the height of the preset plane of the photomask. When the photomask is attached to guide rails at both ends, the height of the guide rails is the height of the preset plane. Similarly, the bending data in the bending information is correlated with the position of the photomask shape represented by the bending data. For Figure 5 The photomask shown has a curvature that is manifested by different heights along the x-axis. Therefore, the curvature data is the difference between the height data z(x) at position x and the height reference value h0. The height data z(x) of the entire photomask is a function of z changing with x.

[0030] Step S120: Determine the wafer's position information based on the flatness information when different locations on the wafer surface are within the focal plane of the photomask pattern under conditions of no photomask bending. In other words, based on the flatness information, calculate the axial compensation amount that needs to be applied to the wafer to ensure that the focal plane formed by the photomask pattern on the image side coincides with various local areas on the wafer surface under conditions of no photomask bending, and use this axial compensation amount as the wafer's position information.

[0031] It should be noted that the focal plane of the photomask pattern when the photomask is not bent is the same as the imaging plane of the photomask pattern when the photomask is not bent. Since the distance between the preset plane of the photomask and the lens on the lithography machine is fixed, the focal plane of the photomask pattern when the photomask is not bent is also fixed. Given the distance between the preset plane of the photomask and the lens and the focal length of the lens, the focal plane can be determined according to the imaging formula.

[0032] The aforementioned wafer position information can be constructed from the position data of the wafer carrier platform when different positions on the wafer surface are in the focusing plane. The wafer is placed on the surface of the carrier platform; movement of the carrier platform causes the entire wafer to move. Given a fixed position of the carrier platform, the wafer position is also uniquely determined. However, if the wafer surface is uneven, the distance between different positions on the wafer surface and the lens will be different. Therefore, when the wafer surface is uneven, the position data of the wafer carrier platform when different positions on the wafer surface are in the focusing plane will also be different.

[0033] In the embodiments of this application, the focal point of the photomask pattern on the focusing plane when the photomask is not bent is also called the automatic focus point. The above position information can be regarded as the automatic focus point position obtained by compensating for the flatness information.

[0034] Step S130: Compensate the position information based on the curvature information to obtain new position information.

[0035] It should be noted that during the process of compensating for position information based on curvature information, there is a one-to-one correspondence between the curvature data in the curvature information and the position data in the position information. Specifically, the patterns at different positions of the photomask are configured to be projected onto different preset positions on the wafer surface. The photomask position is associated with the curvature data, and the wafer surface position is associated with the position data. Therefore, the curvature data associated with a position on the photomask corresponds one-to-one with the position data associated with the preset position on which the pattern at that position on the photomask is to be projected.

[0036] Compensating for position information based on bending information involves using the new position information obtained after compensation to etch a pattern of uniform critical size onto the photoresist at different positions on the wafer surface. Since the critical size of the pattern projected onto the focal plane by the photomask pattern when the photomask is not bent is usually the critical size of the design, the aforementioned uniform critical size can be selected as the critical size of the pattern projected onto the focal plane by the photomask pattern when the photomask is not bent.

[0037] Step S140: Photolithography is performed on the wafer at the position indicated by the new position information using a photomask.

[0038] Specifically, the new position information is obtained after position information compensation. The position information is composed of the position data of the wafer carrier platform when different positions on the wafer surface are in the focusing plane. Therefore, the new position information is composed of the new position data corresponding to different positions on the wafer surface. In this way, when exposing the photoresist at different positions on the wafer surface, the wafer carrier platform is moved to the position indicated by the new position data corresponding to that position on the wafer surface.

[0039] It should be noted that since there is a one-to-one correspondence between the bending data in the bending information and the position data in the position information, the new position data in the new position information also corresponds one-to-one with the bending data in the bending information. This allows for the establishment of a correlation between the new position data and the bending data in the bending information. Therefore, when performing photolithography on a pattern at a certain location on the photomask, the new position data is determined based on this correlation between the bending data at that location and the new position data. In practice, the size of the photomask's bending amplitude can be used to determine how many areas of the pattern on the photomask share a single new position data for a single exposure. This ensures that the critical dimensions at different locations within the same exposure unit on the wafer surface are essentially the same, and that the critical dimensions within different exposure units are also essentially the same.

[0040] The photolithography method provided in this application does not compensate for the bending of the photomask by adjusting the placement of the photomask, but by adjusting the position of the wafer. The bending information of the photomask is directly compensated into the wafer position information determined based on the flatness information of the wafer surface, so that the bending of the photomask can be accurately compensated. This method can be used to eliminate the influence of photomask bending on the uniformity of key dimensions and the uniformity of alignment deviation compensation value within the exposure unit.

[0041] In an optional embodiment, the process of acquiring the bending information of the photomask in step S110 includes: acquiring bending data at multiple locations on the photomask, and using the acquired bending data to fit a topographic curve of the photomask including bending information. This eliminates the need to strictly acquire bending data at every single location on the photomask to obtain the topographic curve. Specifically, the number of acquired bending data points is determined based on the bending amplitude of the photomask; a larger bending amplitude results in more bending data points, thereby ensuring that multiple bending data points can accurately fit the topographic curve of the photomask. Figure 5 For photomasks with a small bending amplitude, bending data can be collected from each of the five dashed boxes K0 to K4 to fit the topography curve of the photomask.

[0042] Furthermore, when the photomask is supported by the opposing first and second edges to be in a preset plane, the shape of the photomask is symmetrical and the axis of symmetry is the middle boundary line of the photomask that does not intersect with either the first or second edge. The photomask includes a first region from the first edge to the middle boundary line and a second region from the second edge to the middle boundary line. The aforementioned acquisition of bending data at multiple positions of the photomask can be obtained by subtracting the height data at multiple positions within either the first or second region. The region providing the height data at multiple positions is called the target region. The aforementioned fitting of the shape curve of the photomask using the acquired bending data can include: fitting the shape curve of the target region using the multiple bending data obtained by subtraction and determining the shape curve of the photomask based on the shape curve of the target region. Specifically, the shape curve of the photomask can be determined based on the shape curve of the target region by utilizing symmetry. For Figure 5 The photomask shown can collect height data from the region to the right of the z-axis in the dashed box K0, as well as from dashed boxes K1 and K2. It is not necessary to collect height data from all five dashed boxes from K0 to K4, which helps to reduce the amount of data collected.

[0043] In another optional embodiment, the process of obtaining wafer surface flatness information in step S110 includes: obtaining a detection signal for detecting the height of the wafer surface, filtering out noise in the detection signal, and determining the wafer surface flatness information based on the noise-filtered detection signal, thereby making the determined wafer surface flatness information more accurate.

[0044] Specifically, a sufficient number of focus sensors can be set up to detect the height of the wafer surface, and the detection signal is an optical signal. In practice, the number of focus sensors is equal to the number of grids divided on the wafer, with one focus sensor detecting the height data of one grid. For example, the size of the grids divided on the wafer is 1mm × 1mm. In this case, one focus sensor detects the height of a 1mm × 1mm area on the wafer surface. The detection signals obtained by multiple focus sensors are signals distributed on a two-dimensional plane. The horizontal axis of the signal represents the position on the wafer surface, and the vertical axis represents the signal amplitude obtained by transforming the height of the corresponding detection position on the wafer surface (e.g., by taking the first derivative or a derivative higher than the first order). Based on the detection signals, a height distribution map of the wafer surface can be obtained. The accuracy of this height distribution map is at the nanometer level. In this application, the height distribution map is also referred to as the initial height distribution map of the wafer surface.

[0045] Since wafer surface protrusions exhibit a relatively obvious height change in at least one direction on the wafer surface, the detection signal has relatively obvious abnormal signal characteristics that characterize wafer surface protrusions in at least one direction. Figure 6a The illustration shows a protruding area on the wafer surface and its corresponding exemplary detection signal, such as... Figure 6a As shown, the protruding regions on the wafer surface include a height-increasing region between points Q1 and Q2, a height-maintaining region between points Q2 and Q3, and a height-decreasing region between points Q3 and Q4. The corresponding detection signals, in the direction of height change in the protruding regions, include a first signal trough near point P1, a first signal peak near point P2, a second signal peak near point P3, a second signal trough near point P4, and a signal segment between the first and second signal peaks. Here, point P1 represents the signal obtained from detecting point Q1 on the wafer surface, point P2 represents the signal obtained from detecting point Q2 on the wafer surface, point P3 represents the signal obtained from detecting point Q3 on the wafer surface, and point P4 represents the signal obtained from detecting point Q4 on the wafer surface. The protruding regions on the wafer surface are those that protrude compared to the flat regions of the wafer. Correspondingly, the signal peaks and troughs are relative to the signal amplitudes of the flat regions. The greater the protrusion height of the protruding region, the greater the peak value of the signal peak.

[0046] The aforementioned noise filtering in the detection signal can be achieved by determining whether there is a segment of signal with abnormal signal characteristics in the detection signal. If so, this segment of signal is considered noise, and then it is filtered out from the detection signal. For Figure 6a The characteristics of the abnormal signal are shown in the diagram. Figure 6bThe detection signal that does not have a signal valley but has a signal peak in the direction of the height change of the wafer surface, and the peak value of the signal peak is greater than the peak value threshold, is a noise signal.

[0047] Furthermore, the above-mentioned determination of wafer surface flatness information based on the noise-filtered detection signal may include: determining whether the noise-filtered detection signal includes a first sub-signal, wherein the first sub-signal indicates that the detected wafer surface position is protruding and the protrusion height is greater than a height threshold; if the noise-filtered detection signal includes the first sub-signal, removing the first sub-signal from the noise-filtered detection signal, and determining the wafer surface flatness information based on the detection signal after removing the first sub-signal. In this way, no new position data is subsequently determined for the wafer surface position detected by the first sub-signal. If the wafer surface position detected by the first sub-signal is not suitable for manufacturing semiconductor integrated circuits, the wafer surface position detected by the first sub-signal can be marked and notified to the manufacturer through an alarm so that the process can be improved subsequently.

[0048] for Figure 6a The characteristics of the abnormal signal are shown in the diagram. Figure 6c The signal segment located between points S3 and S4 in the direction of the height change on the schematic wafer surface is the first sub-signal. Figure 6c The signal segment shown includes a left signal valley located to the right of point S1 and adjacent to point S1, a right signal valley located to the left of point S2 and adjacent to point S2, and a signal segment between the left and right valleys. This signal segment indicates that the detected wafer surface includes two protrusions relative to the flat area of ​​the wafer and a protruding plane located between the two protrusions that further protrudes from the two protrusions. The position of the signal segment detected in the protruding plane between point S1 and point S2 has a protrusion height relative to the flat area of ​​the wafer that is greater than a height threshold.

[0049] Furthermore, the above-mentioned determination of wafer surface flatness information based on the noise-filtered detection signal may include: determining whether the noise-filtered detection signal includes a second sub-signal, wherein the second sub-signal indicates that the detected wafer surface position is protruding and the continuous width of the protruding position is greater than a width threshold; if the noise-filtered detection signal includes the second sub-signal, removing the second sub-signal from the noise-filtered detection signal, and determining the wafer surface flatness information based on the detection signal after removing the second sub-signal, so that no new position data is subsequently determined for the wafer surface position detected by the second sub-signal, and the wafer surface position detected by the second sub-signal is considered an abnormal area and is not suitable for manufacturing semiconductor integrated circuits.

[0050] for Figure 6a The characteristics of the abnormal signal are shown in the diagram. Figure 6dThe signal segment located between points S5 and S6 in the direction of height change of the wafer surface is the second sub-signal. The wafer surface area detected by the second sub-signal is uneven compared to the flat area of ​​the wafer. The width w of the second sub-signal shown in the figure corresponds to the continuous width of the detection area in the direction of height change.

[0051] In some examples, determining the flatness information of the wafer surface based on the noise-filtered detection signal can be achieved by removing both the first sub-signal and the second sub-signal from the noise-filtered detection signal, and then determining the flatness information of the wafer surface based on the detection signal after removing the first and second sub-signals.

[0052] In another optional embodiment, step S130, which compensates for the position information based on the bending information, includes: determining the focusing surface of the photomask when the photomask is bent according to the bending information; obtaining the key dimension offset information of the first focusing pattern relative to the second focusing pattern, wherein the first focusing pattern is the focusing pattern of the photomask pattern on the focusing surface when the photomask is bent, and the second focusing pattern is the focusing pattern of the photomask pattern on the focusing plane when the photomask is not bent; and, if the third focusing pattern of the photomask pattern when the first focusing pattern is bent has key dimension offset information, compensating the position offset information of the surface where the third focusing pattern is located (hereinafter also referred to as the compensation surface) relative to the focusing surface to the position information.

[0053] It should be noted that when the photomask is curved, the first and second focused patterns obtained by focusing patterns at different positions on the photomask have different critical dimension offset data. Therefore, it is necessary to have the critical dimension offset data between the first and second focused patterns obtained by focusing the pattern at a certain position on the photomask. This is achieved by compensating for the positional offset data of the curved surface where the third focused pattern is located relative to the focusing surface, and then using this positional offset data to compensate for the positional data associated with the preset position to be projected onto the wafer. The critical dimension offset information can be composed of the aforementioned critical dimension offset data corresponding to different positions on the photomask, and the positional offset information can be composed of the aforementioned positional offset data obtained at different positions on the photomask.

[0054] by Figure 7 The exemplary photomask pattern shown illustrates the above compensation process in detail, wherein, since the photomask pattern is scaled down and projected onto the photoresist on the wafer surface, therefore... Figure 7 The photomask pattern shown is located beyond twice the focal length of the lens on the object side. Figure 7 The image-side focal point F of the lens is indicated by a black square. The image-side focusing pattern and the object-side photomask pattern are both represented by black squares, but this does not imply a limitation on the specific shape of the photomask pattern.

[0055] Reference Figure 7 The photomask pattern A(r) on the lens object side is the pattern at position r on the photomask when the photomask is not bent. After the photomask is bent, the photomask pattern A(r) is transformed into the photomask pattern B(r). The photomask patterns A(r) and B(r) are, for example, light-transmitting areas on the photomask. Thus, for positive photoresist, the photomask patterns A(r) and B(r) can respectively expose the photoresist and etch the corresponding opening areas on the photoresist.

[0056] Figure 7 In the diagram, the focusing pattern A'(r) on the lens image side is the focusing pattern of the mask pattern A(r) on the focusing plane, i.e., the second focusing pattern obtained by focusing the pattern at position r on the mask when the mask is not bent; the focusing pattern B'(r) on the lens image side is the focusing pattern of the mask pattern B(r) on the focusing curved surface, i.e., the first focusing pattern obtained by focusing the pattern at position r on the mask when the mask is bent. The focusing pattern B'(r) is larger than the focusing pattern A'(r) because it is closer to twice the focal length. Correspondingly, the critical dimension of the focusing pattern B'(r) is larger than that of the focusing pattern A'(r). Subtracting the critical dimension of the mask pattern A'(r) from the critical dimension of the mask pattern B'(r) yields the critical dimension offset data CD of the first focusing pattern at position r on the mask compared to the second focusing pattern. 0 bias (r). Since the focus patterns A'(r) and B'(r) are determined based on the imaging principle, the sizes of focus patterns A'(r) and B'(r) can be determined accordingly based on the imaging principle, thereby determining the key dimension offset data CD of focus pattern A'(r) relative to focus pattern B'(r). 0 bias (r).

[0057] Furthermore, the focusing pattern B''(r) on the lens image side is the focusing pattern of the mask pattern B(r) on the compensation surface, that is, the third focusing pattern obtained by focusing the pattern at position r on the mask. The key dimension offset data of the focusing pattern B'(r) relative to the focusing pattern B''(r) can be determined from the Poisson curve of the Focus-Energy Matrix (FEM).

[0058] The focus-energy matrix is ​​a test method used to determine the optimal exposure conditions. Specifically, when exposing photoresist on the surface of a wafer, the focus value is changed in one direction with a fixed step size and the exposure energy is changed in another direction with another fixed step size. Then, the key dimensions of each exposure unit, as well as the focus value and exposure energy, are used as inputs to obtain the so-called Poisson curve. Figure 8The diagram shows an exemplary Poisson curve, in which each Poisson curve corresponds to an exposure energy. The horizontal axis represents the focus value, and the vertical axis represents the measured width of the individual lines etched on the photoresist. The vertical axis can also represent critical dimensions. Figure 8 The Poisson curve shown can be represented by the following formula (1), where CD represents the critical dimension; X represents the focal value, which can be understood as the position along the optical axis of the lens, and can be represented as the position of the focal point X. n With defocus value Y n The sum of ; A, B, and C are constants determined when obtaining the Poisson curve through focusing-energy matrix fitting.

[0059] CD=AX 2 +BX+C=A(X n +Y n ) 2 +B(X n +Y n )+C (1) It should be understood that formula (1) can be transformed into formula (2) as shown below, where A', B', and C' are constants A, B, and C in formula (1) and the position X of the focal point. n A definite constant. Furthermore, if the position X of the focal point... n If the value is 0, then formula (1) can be transformed into formula (3) as shown below.

[0060] CD=A'Y n 2 +B'Y n +C' (2) CD=AY n 2 +BY n +C (3) Therefore, based on the imaging principle, the position X of the focal point of the focusing pattern B'(r) along the optical axis of the lens is determined. n Then, the key dimension of the focused pattern B'(r) can be calculated using one of formulas (1) to (3), and then the key dimension offset data CD mentioned above can be subtracted from the key dimension of the focused pattern B'(r). 0 bias (r) obtains the key dimensions of the focused pattern B''(r), and then calculates the defocus value Y according to one of the formulas (1) to (3). n The defocus value Y here n This refers to the offset of the focused pattern B''(r) relative to the original focused pattern B'(r) along the optical axis of the lens. It should be understood that the focusing surface, as the focusing surface of the aperture when it is bent, can be understood as the image formed by the lens through the aperture when it is bent. Therefore, theoretically, the focused pattern B'(r) has... Figure 8 The critical dimension at the vertex of the Poisson curve is shown, while the critical dimension of the focusing pattern B''(r) is smaller than that of the focusing pattern B'(r), thus enabling the focusing pattern B''(r) to achieve compensation.

[0061] Furthermore, the photolithography method provided in this application embodiment may further include: determining whether the positional offset information of the curved surface where the third focusing pattern is located relative to the focusing curved surface is greater than half of the focusing depth of the photolithography lens; wherein, if it is greater than half of the focusing depth, the positional offset information of the curved surface where the third focusing pattern is located relative to the focusing curved surface is compensated to the positional information to obtain new positional information, and then step S140 is executed using the new positional information; otherwise, photolithography is directly performed on the wafer at the position indicated by the positional information using a photomask, that is, the corresponding compensation process is omitted to simplify the photolithography process when the bending of the photomask does not significantly affect the photolithography accuracy of the photoresist on the wafer surface.

[0062] In practice, the same photomask is used sequentially to perform the photolithography process on multiple wafers, and the photolithography process required for each wafer is performed according to the photolithography method described above.

[0063] Corresponding to the lithography methods provided in the above embodiments, another embodiment of this application also provides a lithography machine, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When executed by the processor, the program implements the steps of any of the lithography methods provided in the above embodiments. In some embodiments, the processor and memory can be integrated into a microprocessor for use in the lithography machine. Furthermore, the lithography machine can also be as follows... Figure 2 The image shows a laser source, a beam expander, a photomask, a lens, and a support platform. It is important to emphasize that the support platform in the lithography machine provided in this embodiment is movable in all three dimensions, thereby moving the wafer in three dimensions. This allows multiple exposure units on the wafer to be exposed sequentially while maintaining uniform critical dimensions after compensation for photomask bending information.

[0064] The embodiments disclosed herein no longer involve "straightening" the photomask, but instead allow the wafer to actively conform to the image plane of the curved photomask.

[0065] That is, the photomask bending information is directly converted into the wafer's "position compensation amount" in the Z direction, which is then superimposed on the wafer's original flatness compensation amount, so that each part of the wafer always falls on the actual focusing surface corresponding to the bent photomask, thereby eliminating the systematic unevenness of CD and overlay in one go.

[0066] It should be noted that in describing the various embodiments in this specification, the focus is on the differences from other embodiments, while the same or similar parts between the various embodiments can be understood by referring to each other. Regarding the lithography machine embodiment, since it is basically similar to the method embodiments, the relevant parts can be referred to the description of the method embodiments.

[0067] As described above, these embodiments of this application do not exhaustively cover all details, nor do they limit the application to merely the specific embodiments described. Clearly, many modifications and variations can be made based on the above description. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of this application, thereby enabling those skilled in the art to effectively utilize this application and its modifications. This application is limited only by the claims and their full scope and equivalents.

Claims

1. A photolithography method, comprising: Obtain information on the flatness of the wafer surface and the curvature of the photomask; Based on the flatness information, the axial compensation amount that needs to be applied to the wafer to make the focal plane formed by the photomask pattern in the image side coincide with each local area on the wafer surface when the photomask is not bent is calculated, and the axial compensation amount is used as the wafer position information. The position information is compensated based on the curvature information to obtain new position information; The wafer at the position indicated by the new position information is photolithographically etched using the photomask.

2. The photolithography method according to claim 1, wherein, Compensating the position information based on the curvature information includes: The focusing surface of the photomask under bending conditions is determined based on the bending information; Obtain key dimensional offset information of the first focusing pattern relative to the second focusing pattern. The first focusing pattern is the focusing pattern of the photomask on the focusing surface when the photomask is bent, and the second focusing pattern is the focusing pattern of the photomask on the focusing plane when the photomask is not bent. When the third focusing pattern of the photomask pattern has the key dimension offset information compared to the first focusing pattern when the photomask is bent, the position offset information of the surface where the third focusing pattern is located compared to the focusing surface is compensated to the position information. The third focusing pattern is the focusing pattern of the photomask pattern on the compensated surface when the photomask is bent.

3. The photolithography method according to claim 2, further comprising: Determine whether the positional offset information of the surface where the third focusing pattern is located relative to the focusing surface is greater than half of the focusing depth of the photolithography lens; If the depth of focus is greater than half of the depth of focus, the position offset information of the surface where the third focus pattern is located relative to the focus surface is compensated to the position information.

4. The photolithography method according to claim 3 further includes: When the positional offset information of the surface where the third focusing pattern is located relative to the focusing surface is not greater than half of the focusing depth, the wafer at the position indicated by the position information is photolithographically etched using the photomask.

5. The photolithography method according to claim 1, wherein, Obtaining wafer surface flatness information includes: acquiring a detection signal for detecting the height of the wafer surface, filtering out noise in the detection signal, and determining the wafer surface flatness information based on the noise-filtered detection signal.

6. The photolithography method according to claim 5, wherein, Determining the flatness information of the wafer surface based on the detection signal after noise filtering includes: Determine whether the detection signal after noise filtering includes a first sub-signal, wherein the first sub-signal indicates that the detected wafer surface protrudes and the protrusion height is greater than a height threshold; If the detection signal after noise filtering includes the first sub-signal, the first sub-signal is removed from the detection signal after noise filtering, and the flatness information of the wafer surface is determined based on the detection signal after removing the first sub-signal.

7. The photolithography method according to claim 5, wherein, Determining the flatness information of the wafer surface based on the detection signal after noise filtering includes: Determine whether the detection signal after noise filtering includes a second sub-signal, wherein the second sub-signal indicates that the detected wafer surface is protruding and the continuous width of the protruding position is greater than a width threshold. If the detection signal after noise filtering includes the second sub-signal, the second sub-signal is removed from the detection signal after noise filtering, and the flatness information of the wafer surface is determined based on the detection signal after removing the second sub-signal.

8. The photolithography method according to claim 1, wherein, Obtain the bending information of the photomask, including: Obtain height data at multiple locations of the photomask; Calculate the difference between the height data and the height reference value of the photomask, where the height reference value is the height of the preset plane of the photomask; The morphology curve of the photomask, including the bending information, is fitted using multiple calculated differences.

9. The photolithography method according to claim 8, wherein, The photomask is supported by opposing first and second edges to be positioned on the preset plane, and the photomask includes a first region from the first edge to the middle boundary line and a second region from the second edge to the middle boundary line; Acquiring height data at multiple locations of the photomask includes: acquiring height data at multiple locations within a target area, wherein the target area is either the first area or the second area; Fitting the topography curve of the photomask, which includes the bending information, using multiple calculated differences includes: fitting the topography curve of the target region using multiple calculated differences, and determining the topography curve of the photomask based on the topography curve of the target region.

10. A lithography machine, comprising: A processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the photolithography method as described in any one of claims 1-9.

Images