JP3808817B2 - Mask defect inspection method, semiconductor device manufacturing method, mask defect inspection apparatus, defect impact map creation method, and program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a mask defect inspection method for a photomask used in a lithography process of a semiconductor device, a method for manufacturing a semiconductor device including a photomask inspection process, a mask defect inspection device, and a defect used for a mask defect inspection of a photomask. The present invention relates to a program for causing a computer to execute an influence map creation method, and the mask defect inspection method and the defect influence map creation method.
[0002]
[Prior art]
2. Description of the Related Art In recent years, in the manufacture of semiconductor memory devices, higher integration of elements and wirings constituting a circuit and miniaturization of patterns of elements and wirings have been promoted. For example, in the case of a DRAM (Dynamic Random Access Memory) which is a typical semiconductor memory device, it is said that a pattern having a design rule of 0.13 μm is necessary for manufacturing a 1G DRAM.
[0003]
Conventionally, photomask inspection is performed using a die-to-die comparison inspection method that compares adjacent identical patterns or a die-to-data base comparison inspection method that compares measurement patterns and design patterns. Yes.
[0004]
FIG. 20 shows a schematic configuration diagram of a defect inspection apparatus used in a conventional die-to-die comparative inspection method. This defect inspection apparatus has two optical systems 81 and 82 for one light source 80, and displays an enlarged image of a pattern of transmitted light from two adjacent identical patterns formed on a photomask 83, such as a CCD. Are imaged on the image sensors 84 and 85 and converted into electrical signals.
[0005]
The optical systems 81 and 82 include a reflection mirror 86 and objective lenses 87 and 88, respectively. The photomask 83 is placed on the XY stage 89. The position of the XY stage 89 is controlled by the stage control mechanism 90.
[0006]
Electrical signals (sensor data) 91 and 92 obtained by the image sensors 84 and 85 are input to the comparison logic circuit 93. The comparison logic circuit 93 compares the electrical signals 91 and 92 to detect a pattern mismatch portion (defect).
[0007]
The computer 94 determines the mismatched portion on the photomask 83 based on the mismatched portion detected by the comparison logic circuit 93 and the positional information (XY coordinates) of the XY stage 89 from the stage control mechanism 90. Calculate the coordinates. This coordinate is recorded in the defect inspection apparatus as defect position information of the mask defect.
[0008]
On the other hand, in the defect inspection apparatus used for the die-to-data base comparative inspection method, as shown in FIG. 21, the comparison logic circuit 93 generates a comparison pattern using the design pattern stored in the database 95. The reference pattern 97 obtained from the circuit 96 is compared with the sensor data 91 obtained from the optical system 81, and a mismatched portion of the pattern is detected.
[0009]
The computer 94 calculates the coordinates of the mismatched portion on the photomask 83 based on the mismatched portion detected by the comparison logic circuit 93 and the XY coordinates of the XY stage 89. This coordinate is recorded in the defect inspection apparatus as defect position information of a defect (mask defect) on the photomask.
[0010]
However, the conventional die-to-die comparison inspection method and die-to-data base comparison inspection method have the following problems.
[0011]
In the photomask surface, patterns of various sizes are arranged in each part even if the design rule is the same. Therefore, the influence of the mask defect on the resist pattern (transfer pattern) on the wafer at various locations in the pattern area within the photomask surface, or the operation (characteristics) of the device formed on the wafer of the mask defect. ) Will be different.
[0012]
Here, the conventional die-to-die comparison inspection method and die-to-data base comparison inspection method are performed without considering the influence of the mask defect on the resist pattern or device operation. Specifically, With respect to the management dimension pattern determined for each design rule, in-plane inspection of a single photomask is performed with detection sensitivity at which a defect specification of a certain size should be detected.
[0013]
In this way, conventional photomask inspection is performed with uniform detection sensitivity regardless of where the mask defect affects the resist pattern or device operation and does not affect it. In some places, inspections may be performed with more severe detection sensitivity than necessary.
[0014]
When inspection is performed with a detection sensitivity that is stricter than necessary, a portion that is not originally considered as a mask defect may be detected as a mask defect (pseudo defect). Therefore, many types of defects are detected more than necessary, and it takes time to classify the detected mask defects. As a result, shortening of the manufacturing period of the photomask is hindered, and as a result, shortening of the manufacturing period of the semiconductor device is hindered.
[0015]
Also, the conventional die-to-die comparative inspection method and die-to-data base comparative inspection method do not consider the influence of mask defects on the resist pattern or device operation as described above. The distinction between mask defects that do not affect the operation (pseudo defects) and mask defects that are affected is not clear when obtaining the inspection results.
[0016]
Therefore, confirmation of the influence of each mask defect on the resist pattern based on the defect position information (defect coordinates) is performed using a lithography simulation microscope (for example, MSM100 manufactured by Carl Zeiss). It is classified as a defect. A mask defect that affects transfer is corrected and a photomask is shipped.
[0017]
Thus, after obtaining the inspection results, distinguishing between the mask pattern that does not affect the resist pattern or device operation and the mask defect that causes it increases the manufacturing process, thereby shortening the photomask manufacturing period. As a result, there is a problem that shortening the manufacturing period of the semiconductor device is hindered.
[0018]
[Patent Document 1]
JP 2000-98584 A
[0019]
[Problems to be solved by the invention]
As described above, the conventional die-to-die comparison inspection method and die-to-data base comparison inspection method do not consider the influence of mask defects on the resist pattern or the influence of mask defects on device operation. However, a portion (pseudo defect) that is not originally regarded as a mask defect is detected as a mask defect, and as a result, there is a problem that shortening of the manufacturing period of a photomask or a semiconductor device is hindered.
[0020]
The present invention has been made in view of the above circumstances, and the object of the present invention is to provide a mask defect inspection method, a semiconductor device manufacturing method, and a mask defect inspection capable of shortening the manufacturing period of a photomask and a semiconductor device. The object is to provide a device, a defect influence map creation method, and a program.
[0021]
[Means for Solving the Problems]
Of the inventions disclosed in this application, the outline of typical ones will be briefly described as follows.
[0022]
  That is, in order to achieve the above object, a mask defect inspection method according to the present invention relates to a photomask including a mask pattern, and the degree of influence of defects on the portion on a wafer for each of a plurality of portions of the mask pattern. The step of preparing the detection sensitivity of the defect at the location determined according to the step, and the step of inspecting the defect on the plurality of locations based on the detection sensitivityIn the step of preparing the detection sensitivity, a management pattern for managing the photomask and a transfer pattern formed on the wafer and corresponding to the management pattern are necessary for finishing to a predetermined dimension. 1 for each of the plurality of locations of the mask pattern, a width of the location is changed, and a transfer pattern corresponding to the location where the width is changed formed on the wafer is set to a predetermined value. Obtaining a second exposure amount necessary for finishing to a size, calculating a change rate of the second exposure amount with respect to the first exposure amount for each of the plurality of portions of the mask pattern, and For each of the plurality of locations of the pattern, shadows of defects on the locations on the wafer based on the rate of change of the second exposure dose with respect to the first exposure dose. The degree of defect detection at the location is acquired based on the degree of influence of the defect on the location on the wafer for each of the location of the mask pattern, and the plurality of the mask pattern is obtained. Based on the detection sensitivity acquired for each of the locations, an influence map of the defects on the photomask to the wafer is createdIt is characterized by that.
[0023]
A method of manufacturing a semiconductor device according to the present invention includes a step of inspecting a photomask by the mask defect inspection method according to the present invention.
[0024]
  The mask defect inspection apparatus according to the present invention has a defect detection sensitivity at each of the plurality of portions of the mask pattern in the photomask determined according to the degree of influence of the defect on the portion on the wafer. Based on a map creation unit that creates an impact map representing the degree of impact of defects on the photomask on the wafer, and an inspection unit that inspects defects on the plurality of locations based on the impact map EquippedThe map creation unit is formed on the photomask for each of the plurality of locations of the mask pattern, the width of the location being changed, and the width formed on the wafer corresponding to the changed location. Obtaining a first exposure amount necessary for finishing a transfer pattern corresponding to a management pattern for managing the photomask to a predetermined dimension, and for each of the plurality of portions of the mask pattern, A change rate of the first exposure amount with respect to a second exposure amount necessary for finishing the transfer pattern to a predetermined dimension is calculated, and the second exposure amount is determined for each of the plurality of portions of the mask pattern. On the basis of the rate of change of the first exposure amount with respect to the wafer, the degree of influence of the defect on the portion on the wafer is obtained, and the degree of the plurality of portions of the mask pattern is obtained. For each, the detection sensitivity of the defect at the location is acquired based on the degree of influence of the defect on the location on the wafer, and based on the detection sensitivity acquired for each of the multiple locations of the mask pattern Creating a map of the influence of defects on the photomask on the waferIt is characterized by that.
[0025]
  The defect influence map creating method according to the present invention relates to a photomask including a mask pattern, and for each of a plurality of portions of the mask pattern, the defect influence map is determined according to the influence of the defect on the portion on the wafer, A step of obtaining a defect detection sensitivity at a location, and an influence including a degree of influence of the defect on the photomask on the wafer based on the defect detection sensitivity obtained for each of the plurality of locations of the mask pattern. Process to create a degree mapThen, in the step of creating the influence map, for each of the plurality of locations of the mask pattern, the width of the location is changed, and the photo corresponding to the location where the width formed on the wafer is changed A plurality of locations of the mask pattern is obtained by obtaining a first exposure amount necessary for finishing a transfer pattern corresponding to a management pattern for managing the photomask formed on the mask to a predetermined dimension. The change rate of the first exposure amount with respect to the second exposure amount necessary for finishing the transfer pattern to a predetermined dimension is calculated for each of the plurality of portions of the mask pattern. 2 based on the rate of change of the first exposure amount with respect to the exposure amount of 2, the degree of influence of the defect on the location on the wafer is obtained, and the mask pattern For each of the plurality of locations, the detection sensitivity acquired for each of the plurality of locations of the mask pattern is acquired based on the degree of influence of the defects on the location on the wafer based on the degree of influence. Based on the sensitivity, a map of the degree of influence on the wafer of defects on the photomask is created.It is characterized by that.
[0026]
  The program according to the present invention relates to a photomask including a mask pattern in a computer, and for each of a plurality of locations of the mask pattern, determined in accordance with the degree of influence of a defect on the location on the wafer. A program for executing a procedure for inputting defect detection sensitivity and a procedure for inspecting defects on the plurality of locations based on the detection sensitivity in the mask dataIn the procedure of inputting the detection sensitivity, for each of the plurality of locations of the mask pattern, the width of the location is changed, and the width formed on the wafer corresponds to the location where the width is changed. A first exposure amount necessary for finishing a transfer pattern corresponding to a management pattern for managing the photomask formed on the photomask to a predetermined size is acquired, and the plurality of the mask patterns The change rate of the first exposure amount with respect to the second exposure amount necessary for finishing the transfer pattern to a predetermined dimension is calculated for each of the portions, and for each of the plurality of portions of the mask pattern, Based on the rate of change of the first exposure amount with respect to the second exposure amount, the degree of influence of the defect on the location on the wafer is acquired, and the mass For each of the plurality of locations of the pattern, the detection sensitivity of the defect at the location is acquired based on the degree of influence of the defect on the location on the wafer, and acquired for each of the plurality of locations of the mask pattern. Based on the detection sensitivity, an influence map of defects on the photomask on the wafer is created..
[0027]
  Another program according to the present invention relates to a photomask including a mask pattern in a computer, for each of a plurality of locations of the mask pattern, which is determined according to the degree of influence of a defect on the location on the wafer, A defect including the degree of influence of the defect on the photomask on the wafer based on the procedure for acquiring the defect detection sensitivity at the position and the defect detection sensitivity acquired for each of the plurality of positions of the mask pattern For executing the procedure for creating an impact mapThen, in the procedure of creating the defect influence degree map, for each of the plurality of locations of the mask pattern, the width of the location is changed to correspond to the location where the width formed on the wafer is changed. Obtaining a first exposure amount necessary for finishing a transfer pattern formed on the photomask corresponding to a management pattern for managing the photomask to a predetermined dimension, and For each of the plurality of locations, the change rate of the first exposure amount with respect to the second exposure amount necessary for finishing the transfer pattern to a predetermined dimension is calculated, and each of the plurality of locations of the mask pattern is calculated. , Based on the rate of change of the first exposure amount with respect to the second exposure amount, to obtain the degree of influence of the defect on the location on the wafer, For each of the plurality of locations of the mask pattern, the detection sensitivity of the defect at the location is acquired based on the degree of influence of the defect on the location on the wafer, and for each of the plurality of locations of the mask pattern Based on the acquired detection sensitivity, an influence map of defects on the photomask on the wafer is created.
[0028]
  Claim 1,8Inventions according to other claims can also be implemented as inventions related to programs.
[0029]
According to the present invention, it is possible to prevent detection of pseudo defects by taking into account the degree of influence of the mask defect on the wafer, for example, the influence on the transfer pattern or the influence of the mask defect on the device operation. The manufacturing period of the apparatus can be shortened.
[0030]
The above and other objects and novel features of the present invention will become apparent from the description of the present specification and the accompanying drawings.
[0031]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0032]
(First embodiment)
First, the basic concept of the photomask inspection method according to the first embodiment of the present invention will be described. FIG. 1 shows a photomask, in which A indicates a region where a line pattern having a large line width exists, and B indicates a region where a line pattern having a small line width exists.
[0033]
Here, when there are mask defects (not shown) of the same size and shape in the regions A and B, the mask defect in the region B is formed on the wafer, that is, on the wafer. The influence on the resist pattern (transfer pattern) is large. In other words, the mask defect in the region A has a small influence on the transfer pattern and can be ignored.
[0034]
In the conventional photomask inspection method / apparatus, when the design rule is the same, the detection sensitivity is the same for all patterns, and therefore a mask defect may be detected in the region A and the region B. That is, a certain location in the region A may be detected as a pseudo defect.
[0035]
Therefore, in this embodiment, for each of a plurality of portions of the mask pattern, by determining the detection sensitivity of the mask defect according to the degree of influence of the mask defect, for example, the defect detection sensitivity of the pattern in the region A is reduced, By increasing the defect detection sensitivity of the pattern in the region B, the mask inspection is performed in consideration of the influence of the mask defect on the resist pattern. As a result, the number of pseudo defects detected in the region A can be reduced, and the manufacturing period of the photomask and the semiconductor device can be shortened.
[0036]
Instead of the influence of the mask defect on the resist pattern, the influence of the mask defect on the device operation may be considered. The influence on device operation will be described in the second embodiment. Further, the influence of the mask defect on the transfer pattern and device operation may be considered simultaneously.
[0037]
Details of the photomask inspection method of this embodiment will be described below.
[0038]
First, as shown in FIG. A halftone photomask (KrF-HT mask) in which a mask pattern made of a film (KrF-HT film) that is semi-transparent to KrF exposure light is formed on the transparent substrate 1 made of
[0039]
In FIG. 2, 2 is a region where a memory cell pattern is formed (cell pattern region), 3 is a region where a peripheral circuit pattern is formed (peripheral circuit pattern region), 4 is a transparent substrate 1 that does not constitute a device pattern The KrF-HT film | membrane on the peripheral part of is shown.
[0040]
Next, as shown in FIG. 3, the division point 5 is designated in 150 nm increments for the design pattern data corresponding to the pattern in the region (inspected region) to be inspected on the photomask. In FIG. 3, 6 and 7 indicate locations corresponding to the KrF-HT film (light-shielding region) and the exposed surface (transparent region) of the transparent substrate 1 on the graphic data of the design pattern, respectively. Yes.
[0041]
As the division point 5, it is preferable to designate a portion that is an integral multiple of the drawing grid width of the mask drawing apparatus or the inspection grid width of the defect inspection apparatus with respect to the design pattern data. As a result, the division point 5 can be easily specified.
[0042]
Next, as shown in FIG. 4, each edge (edge AB, edge) defined by two adjacent upper and lower division points 5 (5A and 5B, 5B and 5C, 5D and 5E, 5E and 5F,...). (B-C, edge DE, edge EF,...)), A pattern in which the line line width is changed by Δd = ± 1, ± 2, ± 5 [nm] is prepared. FIG. 4 shows patterns PT1 to PT3 in which the line width is changed by Δd = 1, 2, 5 [nm] for the edges (edge BC) defined by the division points 5B and 5C. Although one edge side of the line is selected in FIG. 4, both edges may be selected.
[0043]
The line line width Δd may be changed for all edges defined by the dividing point 5, or the line line width Δd may be changed for one edge. That is, at least one edge is selected and the line line width Δd is changed. Basically, the inspection accuracy increases when more edges are selected. In the present embodiment, description will be made assuming that the line line width Δd is changed for all edges.
[0044]
Further, it is preferable that the selected edges do not have an optical proximity effect with each other or have a sufficiently low effect. Here, the division point 5 is designated in the vertical direction, but the direction in which the division point 5 is designated is arbitrary.
[0045]
Next, a pattern of Δd = 0 [nm] (defect-free pattern) is exposed for each edge of the management pattern (a pattern used for managing a photomask having the same design rule) by well-known exposure margin evaluation. The exposure amount Ie required at this time is acquired. Instead of the exposure amount Ie, a threshold value (Threshold) of light intensity corresponding to the exposure amount Ie may be acquired.
[0046]
Next, with respect to a pattern in which Δd = ± 1, ± 2, ± 5 [nm] is changed with respect to each edge AB, BC,..., Λ = 248 nm, NA = 0.68, σ = 0. 75, calculating the light intensity distribution formed on the resist on the wafer when using 2/3 annular illumination, exposing the respective patterns on the resist as desired from the calculated light intensity distribution, Exposure amount I required to form a resist pattern with dimensions_ijAnd the exposure amount I with respect to the exposure amount Ie_ijChange amount (exposure amount change rate) ΔI_ij(= (I_ij-Ie) / Ie) is acquired. FIG. 5 shows the light intensity distribution and the exposure amount change rate for the edge BC pattern (Δd = 0, 1, 2, 5 [nm]).
[0047]
Where ΔI_ijIn the figure, i indicates an index corresponding to each edge (edge AB, edge BC, edge DE, edge EF,...), And here, edge AB is i = 1, edge Numbers are assigned in ascending order in alphabetical order such that BC is i = 2, edge DE is i = 3, and edge EF is i = 4.
[0048]
In addition, j indicates an index corresponding to the line width variation amount Δd on the mask for each edge. Here, Δd = 1 nm is j = 1, Δd = 2 nm is j = 2, and Δd = 5 nm is j = 5. Thus, the numerical value of the line width variation [nm] is assigned.
[0049]
Exposure change rate ΔI_ijCan be positive or negative, and is negative when the exposure amount is lower than the exposure amount Ie, and positive when the exposure amount is increased. In the case where the indices i and j are not particularly necessary, hereinafter, they are simply expressed as ΔI.
[0050]
In this embodiment, regardless of the size of the pattern, the allowable dimension variation rate of the transfer pattern (resist pattern) due to the mask defect is uniformly determined to be 8%. The allowable dimensional variation rate is given by 100 × (size variation of transfer pattern) / (design size of transfer pattern).
[0051]
Although the allowable dimension variation rate may vary depending on the pattern dimension, a value of 8%, which is empirically recognized as being effective, is adopted here for the sake of simplicity. The allowable dimensional variation rate may be determined for each dimension.
[0052]
When the exposure amount change rate ΔIc of the dimension management pattern corresponding to the allowable dimension variation rate of 8% was examined, ΔIc = −0.06, and Δd corresponding to ΔIc = −0.06 was 2.7 nm. It was.
[0053]
Therefore, ΔI_ijA pattern having Δd corresponding to −0.06 larger than 2.7 nm has a smaller influence on the transfer pattern of a mask defect than a dimension management pattern, and a pattern having Δd of less than 2.7 nm is smaller than a dimension management pattern. The influence of the mask defect on the transfer pattern is large.
[0054]
From this, ΔI_ijWhen Δd corresponding to −0.06 is larger than 2.7 nm, the size of the mask defect to be inspected by the inspection apparatus is larger than the minimum detection defect size determined for the dimension management pattern. However, in the case of a pattern in which Δd is smaller than 2.7 nm, the size of the mask defect to be inspected by the inspection apparatus needs to be smaller than the minimum detected defect size. When Δd is a pattern of 2.7 nm, the size of the mask defect to be inspected by the inspection apparatus may be the same as the minimum detected defect size.
[0055]
FIG. 6 shows the results of examining the relationship between Δd and ΔI at edges AB and BC. From FIG. 6, Δd corresponding to the exposure amount change rate ΔI = −0.06 is 2.5 nm at the edge AB, and 3.2 nm at the edge BC, and the mask defect on the edge AB is transferred. It can be seen that the degree of influence on the pattern is large and that the mask defect on the edge BC has a small influence on the transfer pattern. This means that the necessary detection sensitivity can be ensured even if the detection sensitivity at the time of inspection of the edge BC is lower than the detection sensitivity at the time of inspection of the edge AB. As a result, it is possible to prevent the edge BC from being inspected with a detection sensitivity that is stricter than necessary, and to detect a pseudo defect.
[0056]
Next, Δd is calculated for each divided region of the region to be inspected (inspected region) on the photomask, and the relationship between the pattern position in the inspected region and Δd is used to calculate each region on the inspected region. In the divided area, an index (detection sensitivity K to M) indicating the degree of influence of the mask defect on the wafer is obtained, and based on the index (detection sensitivity K to M) in each divided area, as shown in FIG. A defect influence map including indices (detection sensitivities K to M) is created. As shown in FIG. 8, the defect influence degree map may have a table format in which each divided area on the inspection area and detection sensitivities K to M are associated with each other.
[0057]
For the divided region corresponding to the portion other than the edge, for example, the same index as that of the closest edge in the horizontal direction (X coordinate) is adopted. If the distance to the left edge is the same as the distance to the right edge and the left and right edges have different indexes, either one is adopted.
[0058]
The black defect detection sensitivity is changed with Δd = 2.7 as a boundary. Specifically, for a pattern of Δd ≦ 2.7 nm, 130 nm^□The above is detected as a black defect, and for patterns with Δd> 2.7 nm, 180 nm^□The above are detected as black defects. On the other hand, the white defect detection sensitivity is changed with Δd = 3.5 nm as a boundary, specifically, 180 nm for a pattern with Δd ≦ 3.5 nm.^□The above are detected as white defects, and for patterns with Δd> 3.5 nm, 220 nm^□The above are detected as white defects. The reason why the reference Δd is different between the black defect and the white defect is that the degree of influence on the transfer pattern is different between the black defect and the white defect.
[0059]
The detection sensitivities K to N are four types obtained by combining the detection sensitivities of the black defect and the white defect. Specifically, (black, white) = K (130, 180), L (130, 220), M (180, 180), N (180, 220) nm^□It becomes.
[0060]
Next, referring to the relationship between each divided region on the inspection region and the detection sensitivities K to N, each divided region is inspected based on the detection sensitivity of any of the detection sensitivities K to N, and a defect on the photomask is detected. The presence or absence of is detected.
[0061]
The inspection of the photomask using the detection sensitivity is performed by, for example, a die-to-data base comparison inspection. That is, the comparison data is used to compare the sensor data and the reference data, and those having a level difference equal to or greater than a predetermined value (threshold value) are detected as defects. The comparison circuit is composed of, for example, a differentiation circuit. The threshold value is basically determined to be higher as the detection sensitivities K to N are lower. Thereby, it is possible to reduce mask defects detected as pseudo defects.
[0062]
FIG. 9 shows an example of reference data and sensor data in a comparison circuit using the defect influence map of the present embodiment. This is for the region shown in FIG. As shown in FIG. 9A, the reference data is composed of data (256 gradations) in which the light intensity of the graphic data in each divided region is developed in a multivalued manner. The reference data that is actually used may be data that considers the corner rounding process and the resizing process corresponding to the mask process. On the other hand, as shown in FIG. 9B, the sensor data is composed of measured values (256 gradations) of the light intensity of each divided region obtained from the actual photomask inspection light.
[0063]
FIG. 10 shows a comparison result (level difference) between the reference data and the sensor data in the comparison circuit using the defect influence map of the present embodiment. This is for the region shown in FIG. FIG. 11 shows an example of the threshold value. The threshold value is an average value of the threshold value corresponding to the black defect and the threshold value corresponding to the white defect. As shown in FIG. 10, it is determined whether or not the level difference between the sensor data and the reference data in the area shown in FIG. 8 exceeds a threshold value, and the divided area is determined to exceed the threshold value. Is recognized as a defect. In the case of FIG. 10, it is recognized that there is no defect.
[0064]
FIG. 12 shows a schematic configuration diagram of a defect inspection apparatus for performing the die-to-data base comparison inspection of the present embodiment. The parts corresponding to those of the conventional defect inspection apparatus in FIG. 21 are denoted by the same reference numerals as those in FIG. 21, and detailed description thereof is omitted. A difference from the conventional defect inspection apparatus of FIG. 21 is that a map creation unit 102 to which detection sensitivity information 101 including coordinates of each region on the photomask and detection sensitivity (K to M) information is input is provided. It is in.
[0065]
As described above, according to the present embodiment, detection of pseudo defects can be prevented by considering the influence of mask defects on the transfer pattern. As a result, the manufacturing period of the photomask or the semiconductor device can be shortened, and further the cost can be reduced. Further, it is possible to detect a mask defect having a large influence on the transfer pattern with high sensitivity. This also leads to shortening of the manufacturing period and cost. Furthermore, it is not necessary to distinguish between a mask defect that does not affect the transfer pattern and an affected mask defect after obtaining the inspection result. This also leads to shortening of the manufacturing period and cost.
[0066]
In the present embodiment, the case of a halftone mask used for KrF exposure has been described. However, in the present embodiment, ArF exposure, F₂It can also be applied to halftone masks used for other light exposures such as exposure and EUV exposure, and can also be applied to masks other than halftone masks, such as Levenson masks, COG masks, etc. The present invention can also be applied to a mask used for non-optical system exposure.
[0067]
In this embodiment, the case where the die-to-data base comparison inspection method is used has been described. However, the defect influence degree map information is input in a coordinate format, and the die-to-die comparison performed using the information. Other comparative inspection methods such as an inspection method may be used. FIG. 13 shows a schematic configuration of a defect inspection apparatus for performing a die-to-die comparison inspection. Parts corresponding to those in the defect inspection apparatus of FIGS. 20 and 12 are denoted by the same reference numerals as in FIGS. 20 and 12, and detailed description thereof is omitted.
[0068]
In the present embodiment, the exposure amount change rate ΔI obtained by changing the line width variation amount Δd on the photomask._ijWe have explained the case where the wafer influence level of the mask defect is investigated using this and the detection sensitivity (threshold value) is determined based on the wafer influence degree. However, a technique that can investigate the influence degree of other mask defects on the wafer is used. The detection sensitivity may be determined.
[0069]
For example, the wafer influence degree of the mask defect is determined by using the dimensional variation rate of the pattern (resist pattern) transferred to the resist on the wafer obtained by varying the line width variation Δd with respect to the entire mask pattern on the photomask. The detection sensitivity may be determined based on this wafer influence degree.
[0070]
Another method is to apply the method disclosed in Japanese Patent Laid-Open No. 2000-098584. That is, the wafer influence degree of the mask defect may be examined using the exposure amount change rate for each part of the resist pattern by changing the line width of the entire pattern, and the detection sensitivity may be obtained based on the wafer influence degree.
[0071]
In this embodiment, the defect influence degree map is created by dividing each part of the entire pattern, and the inspection is performed based on the defect influence degree map, but the cell including the pattern having the same size as the design rule. And a peripheral circuit part including a pattern having a size larger than the pattern in the cell part, and two types of defect influence maps are defined for the cell part and the peripheral circuit part (defects in a broad sense) (Influence map creation) Inspection may be performed.
[0072]
(Second Embodiment)
This embodiment is different from the first embodiment in that a defect influence degree map is created using the influence degree of the mask defect on the device operation.
[0073]
First, as in the first embodiment, the KrF-HT mask shown in FIG. 2 is prepared, and then the dividing point 5 is designated in increments of 150 nm as shown in FIG.
[0074]
Next, as shown in FIG. 14, for each segment (segment AD, segment BE, segment CF,...) Defined by two adjacent division points 5, there is an influence on device operation. The line width variation rate (allowable line width variation rate) of the photomask pattern (wafer pattern) transferred to the resist on the wafer corresponding to the boundary is obtained.
[0075]
The allowable line width variation rate is given by 100 × (wafer pattern line width variation amount that starts to affect the device operation) / (wafer pattern design dimension). The device operation is each operation related to a device to be generally checked that affects the device operation, such as leakage current when the gate is off, and whether writing and reading can be performed accurately.
[0076]
Next, changing the dimensions of the segments A-D, B-E, etc. corresponds to changing the line width Δd at the edges A-B, B, etc., and is shown in FIG. Thus, the allowable line width variation rate in each segment is converted into the exposure amount variation rate ΔI using, for example, the results of simulation and experiment. The exposure fluctuation rate ΔI is defined in the same manner as in the first embodiment. Further, it may be converted into a change rate of light intensity corresponding to the exposure amount fluctuation rate ΔI.
[0077]
Also in the present embodiment, as in the first embodiment, the allowable dimension variation rate of the dimension management pattern is 8%, the exposure amount variation rate ΔI corresponding to the allowable dimension variation rate is 0.06, and further to the ΔI The corresponding Δd is 2.7 nm.
[0078]
Δd corresponding to ΔI of edge AB in FIG. 15 is larger than 2.7 nm from FIG. Similarly, Δd corresponding to ΔI of edge BC in FIG. 15 is also larger than 2.7 nm.
[0079]
Accordingly, the portions corresponding to the segments AD and BE have a smaller influence on the device operation of the mask defect than the dimension management pattern. Therefore, the size of the mask defect to be inspected by the inspection apparatus is smaller than that of the dimension management pattern. It may be larger than the minimum detected defect size.
[0080]
On the other hand, when Δd corresponding to ΔI is 2.7 nm or less, the mask defect has a larger influence on the device operation than the dimension management pattern. Must be smaller than the detected defect size.
[0081]
Next, an allowable line width variation rate and an exposure amount variation rate ΔI in each segment in the inspected area are obtained, and as shown in FIG. 16, the relationship between the allowable line width variation rate and the exposure amount variation rate ΔI is graphed by interpolation. (In this case, the table may be a table.) In each divided area on the inspection area, an index (in this case, detection sensitivity K to M) indicating the degree of influence of the mask defect on the wafer is acquired, and the index in each divided area is obtained. Based on (detection sensitivities K to M), as shown in FIG. 17, a defect influence degree map including indexes (detection sensitivities K to M) is formed. The defect influence map may be in a table format in which each divided area on the inspection area is associated with the detection sensitivities K to M.
[0082]
For the divided areas corresponding to the portions other than the segments, for example, the same index as that of the nearest segment in the vertical direction (Y coordinate) is adopted. When the distance to the upper segment is the same as the distance to the lower segment and the index is different between the upper and lower segments, either one is adopted. In addition, the same index is adopted for divided regions on the same segment having the same position in the vertical direction (Y coordinate) but different in the horizontal direction (X coordinate).
[0083]
The black defect detection sensitivity is changed with Δd = 2.7 as a boundary. Specifically, for a pattern of Δd ≦ 2.7 nm, 130 nm^□The above is detected as a black defect, and for patterns with Δd> 2.7 nm, 180 nm^□The above are detected as black defects. On the other hand, the detection of white defects is changed with Δd = 3.5 nm as a boundary. Specifically, for a pattern with Δd ≦ 3.5 nm, 180 nm is detected.^□The above are detected as white defects, and for patterns with Δd> 3.5 nm, 220 nm^□The above are detected as white defects.
[0084]
The detection sensitivities K to N are four types obtained by combining the detection sensitivities of the black defect and the white defect. Specifically, (black, white) = K (130, 180), L (130, 220), M (180, 180), N (180, 220) nm^□It becomes.
[0085]
Next, as in the first embodiment, each divided region on the inspection region is inspected for the presence or absence of a defect with one of the detection sensitivities K to N. This inspection is also performed using, for example, a Die-to-Data base comparison inspection method / device or a Die-to-Die comparison inspection method / device, as in the first embodiment.
[0086]
FIG. 18 shows a diagram corresponding to the multivalued development data and sensor data in FIG. 9, and FIG. 19 shows a diagram corresponding to the comparison result between the reference data and sensor data in FIG. The threshold value is the same as the threshold value shown in FIG. From FIG. 19, in the case of the present embodiment, it can be seen that a defect is detected at a location where no defect was detected in the first embodiment.
[0087]
As described above, according to the present embodiment, detection of pseudo defects can be prevented by taking into consideration the influence of mask defects on device operation. As a result, the manufacturing period of the photomask or the semiconductor device can be shortened, and further the cost can be reduced. In addition, since it is no longer possible to detect a defect having a small influence on the device operation as a defect, a defect that does not affect the device operation and does not need to be corrected is not corrected in the correction process. This also shortens the manufacturing period and reduces costs.
Also in this embodiment, various modifications described in the first embodiment are possible.
[0088]
The present invention is not limited to the above embodiment. For example, in the above-described embodiment, the case where the defect influence degree map is created using the influence degree on the transfer pattern and the influence on the device operation as the influence degree on the wafer of the mask defect has been described. It is also possible to create a defect impact map using the impact on device yield. In this case, for example, an exposure amount change rate ΔIc of a dimension management pattern corresponding to a predetermined allowable device yield rate and Δd corresponding thereto are examined to create a defect influence degree map.
[0089]
In addition, the inspection method and the defect impact map creation method of the above-described embodiment are for causing a computer to execute predetermined means (or for causing the computer to function as predetermined means, or realizing a predetermined function in the computer). Can also be implemented as a program. Furthermore, the present invention can be implemented as a computer-readable recording medium on which the program is recorded.
[0090]
For example, the program of the inspection method relates to a photomask including a mask pattern in a computer, for each of a plurality of locations of the mask pattern, which is determined according to the degree of influence of a defect on the location on the wafer, This is to execute the procedure for inputting the detection sensitivity of the defect at the location and the procedure for inspecting the defect at the plurality of locations based on the detection sensitivity in the mask data. Furthermore, various specific steps of the inspection method described above may be executed as a procedure.
[0091]
On the other hand, the program of the defect influence map is determined in accordance with the influence of the defect on the portion of the wafer on the wafer with respect to each of the plurality of portions of the mask pattern, with respect to the photomask including the mask pattern. Includes the degree of influence of the defect on the photomask on the wafer based on the procedure for obtaining the defect detection sensitivity at the location and the defect detection sensitivity obtained for each of the plurality of locations of the mask pattern. And a procedure for creating a defect influence map. Further, various specific steps of the map creation method described above may be executed as a procedure.
[0092]
Further, the above embodiments include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent requirements. For example, even if some constituent requirements are deleted from all the constituent requirements shown in the embodiment, if the problem described in the column of the problem to be solved by the invention can be solved, the configuration in which this constituent requirement is deleted Can be extracted as an invention.
[0093]
In addition, various modifications can be made without departing from the scope of the present invention.
[0094]
【The invention's effect】
As described above in detail, according to the present invention, it is possible to shorten the manufacturing period of the photomask and the semiconductor device by considering the influence of the mask defect on the wafer.
[Brief description of the drawings]
FIG. 1 is a plan view for explaining a basic concept of a photomask inspection method according to a first embodiment of the present invention;
FIG. 2 is a perspective view of a photomask.
FIG. 3 is a plan view for explaining a process of designating a division point
FIG. 4 is a diagram for explaining a process of preparing a pattern in which the line line width Δd is changed for each region defined by two adjacent upper and lower division points;
FIG. 5 is a diagram showing a light intensity distribution and an exposure amount change rate for an edge BC pattern (Δd = 0, 1, 2, 5 [nm]).
FIG. 6 is a diagram showing the result of examining the relationship between Δd and ΔI at edges AB and BC.
FIG. 7 shows a defect influence map.
FIG. 8 is a diagram showing a defect influence degree table.
FIG. 9 is a diagram illustrating an example of reference data and sensor data in a comparison circuit using a defect influence map.
FIG. 10 is a diagram showing a comparison result (level difference) between reference data and sensor data in a comparison circuit using a defect influence map;
FIG. 11 is a diagram showing an example of a threshold value
FIG. 12 is a schematic configuration diagram of a defect inspection apparatus for performing a die-to-data base comparison inspection according to the embodiment;
FIG. 13 is a schematic configuration diagram of a defect inspection apparatus for performing a die-to-die comparison inspection according to an embodiment.
FIG. 14 is a view for explaining a process of obtaining an allowable line width variation rate for each segment defined by two adjacent division points according to the second embodiment of the present invention;
FIG. 15 is a diagram for explaining a process of converting an allowable line width variation rate in each segment into an exposure amount variation rate;
FIG. 16 is a graph showing the exposure amount change rate for each segment in the inspection area;
FIG. 17 shows a defect influence map.
FIG. 18 is a diagram showing an example of reference data and sensor data in a comparison circuit using a defect influence map.
FIG. 19 is a diagram showing a comparison result (level difference) between reference data and sensor data in a comparison circuit using a defect influence map;
FIG. 20 is a schematic configuration diagram of a defect inspection apparatus used in a conventional die-to-die comparative inspection method.
FIG. 21 is a schematic configuration diagram of a defect inspection apparatus used in a conventional die-to-data base comparative inspection method.
[Explanation of symbols]
1 ... Transparent substrate
2 ... Cell pattern area
3. Peripheral circuit pattern area
4 ... KrF-HT film
5 ... Division point
6 ... Shading area on graphic data of design pattern
7 ... Transparent area on graphic data of design pattern
81, 82 ... Optical system
83 ... Photomask
84, 85 ... Image sensor
86 ... Reflection mirror
87, 88 ... Objective lens
89 ... XY stage
90 ... Stage control mechanism
91, 92 ... Electric signal (sensor data)
93. Comparison logic circuit
94 ... Calculator
95 ... Database
96 ... pattern generation circuit
97 ... Reference pattern
101 ... Detection sensitivity information
102 ... Map creation part

Claims (10)

Regarding a photomask including a mask pattern, for each of a plurality of locations of the mask pattern, a step of preparing a defect detection sensitivity at the location determined according to the degree of influence of the defect on the location on the wafer;
Based on the detection sensitivity, it possesses a step of inspecting defects on the plurality of locations,
In the step of preparing the detection sensitivity,
Preparing a management pattern for managing the photomask and a first exposure amount required to finish a transfer pattern corresponding to the management pattern formed on the wafer to a predetermined dimension;
For each of the plurality of portions of the mask pattern, the width of the portion is changed, and the second pattern necessary for finishing the transfer pattern corresponding to the portion having the changed width formed on the wafer to a predetermined dimension. And calculating a rate of change of the second exposure amount with respect to the first exposure amount for each of the plurality of locations of the mask pattern,
For each of the plurality of locations of the mask pattern, based on the rate of change of the second exposure amount with respect to the first exposure amount, obtain the degree of influence of the defects on the location on the wafer,
For each of the plurality of locations of the mask pattern, based on the degree of influence of the defects on the location on the wafer, to obtain the detection sensitivity of the defects at the location,
A mask defect inspection method , comprising: creating an influence map of defects on the photomask on the wafer based on the detection sensitivity acquired for each of the plurality of portions of the mask pattern .   The mask defect inspection method according to claim 1, wherein the influence degree of the defect on the wafer is an influence degree on a dimension of a transfer pattern formed on the wafer. Regarding a photomask including a mask pattern, for each of a plurality of locations of the mask pattern, a step of preparing a defect detection sensitivity at the location determined according to the degree of influence of the defect on the location on the wafer;
Based on the detection sensitivity, it possesses a step of inspecting defects on the plurality of locations,
The mask pattern includes a pattern constituting a device,
In the step of preparing the detection sensitivity,
Preparing a management pattern for managing the photomask and a first exposure amount required to finish a transfer pattern corresponding to the management pattern formed on the wafer to a predetermined dimension;
For each of the plurality of locations of the mask pattern, to determine the amount of change in the width of the location corresponding to the boundary of the presence or absence of the influence on the characteristics of the device,
A second exposure amount required to finish a transfer pattern corresponding to a portion whose width is changed by the amount of change formed on the wafer to a predetermined dimension is acquired, and the plurality of portions of the mask pattern For each of the above, a change rate of the second exposure amount with respect to the first exposure amount is calculated,
For each of the plurality of locations of the mask pattern, based on the rate of change of the second exposure amount with respect to the first exposure amount, obtain the degree of influence of the defects on the location on the wafer,
For each of the plurality of locations of the mask pattern, based on the degree of influence of the defects on the location on the wafer, to obtain the detection sensitivity of the defects at the location,
A mask defect inspection method , comprising: creating an influence map of defects on the photomask on the wafer based on the detection sensitivity acquired for each of the plurality of portions of the mask pattern . 4. The mask defect inspection method according to claim 3 , wherein the degree of influence of the defect on the wafer is the degree of influence on the characteristics of the device formed on the wafer. In the step of calculating the rate of change of the second exposure amount with respect to the first exposure amount, the width of the portion is changed in a plurality of ways, a plurality of the second exposure amounts are obtained, and the plurality of the second exposure amounts are obtained. for each exposure, the first mask defect inspection method according to any one of claims 1 to 4, characterized in that to calculate the rate of change with respect to the exposure amount. The method of manufacturing a semiconductor device characterized by including the claims 1 to process for inspecting a photomask by the mask defect inspection method according to any one of 5. For each of a plurality of portions of the mask pattern in the photomask, the defect on the photomask is determined based on the detection sensitivity of the defect in the portion determined according to the degree of influence of the defect on the portion on the wafer. A map creation unit that creates an impact map representing the impact on the wafer;
An inspection unit for inspecting defects on the plurality of locations based on the influence map ; and
The map creation unit
For each of the plurality of portions of the mask pattern, the width of the portion is changed, and the photomask formed on the photomask corresponding to the portion where the width is formed on the wafer is managed. Acquiring a first exposure amount necessary for finishing a transfer pattern corresponding to the management pattern to a predetermined dimension, and transferring the transfer pattern to a predetermined dimension for each of the plurality of portions of the mask pattern. Calculating the rate of change of the first exposure amount with respect to the second exposure amount necessary for finishing
For each of the plurality of locations of the mask pattern, based on the rate of change of the first exposure amount with respect to the second exposure amount, to obtain the degree of influence of the defects on the location on the wafer,
For each of the plurality of locations of the mask pattern, based on the degree of influence of the defects on the location on the wafer, to obtain the detection sensitivity of the defects at the location,
A mask defect inspection apparatus that creates an influence map of defects on the photomask on the wafer based on the detection sensitivity acquired for each of the plurality of portions of the mask pattern . Regarding a photomask including a mask pattern, for each of a plurality of locations of the mask pattern, the step of obtaining the detection sensitivity of the defect at the location determined according to the degree of influence of the defect on the location on the wafer;
Obtained for each of a plurality of positions of the mask pattern, on the basis of the detection sensitivity of the defect, possess a step of creating a degree of influence map including the degree of influence on the wafer defects on the photomask,
In the step of creating the influence map,
For each of the plurality of portions of the mask pattern, the width of the portion is changed, and the photomask formed on the photomask corresponding to the portion where the width is formed on the wafer is managed. Acquiring a first exposure amount necessary to finish a transfer pattern corresponding to the management pattern to a predetermined dimension, and setting the transfer pattern to a predetermined dimension for each of the plurality of portions of the mask pattern. Calculating a rate of change of the first exposure amount with respect to a second exposure amount necessary for finishing;
For each of the plurality of locations of the mask pattern, based on the rate of change of the first exposure amount with respect to the second exposure amount, to obtain the degree of influence of the defects on the location on the wafer,
For each of the plurality of locations of the mask pattern, based on the degree of influence of the defects on the location on the wafer, to obtain the detection sensitivity of the defects at the location,
A defect influence map , characterized in that , based on the detection sensitivity acquired for each of the plurality of portions of the mask pattern, an influence map on the wafer of defects on the photomask is created. How to make. With respect to a photomask including a mask pattern, the computer inputs a defect detection sensitivity at each of the plurality of portions of the mask pattern determined according to the degree of influence of the defect on the portion on the wafer. Procedure and
A program for inspecting defects on the plurality of locations based on the detection sensitivity in the mask data ,
In the procedure for inputting the detection sensitivity,
For each of the plurality of portions of the mask pattern, the width of the portion is changed, and the photomask formed on the photomask corresponding to the portion where the width is formed on the wafer is managed. A transfer pattern corresponding to a management pattern for obtaining a first exposure amount necessary to finish the transfer pattern to a predetermined dimension, and the transfer pattern having a predetermined dimension for each of the plurality of portions of the mask pattern. Calculating a rate of change of the first exposure amount with respect to a second exposure amount necessary for finishing;
For each of the plurality of locations of the mask pattern, based on the rate of change of the first exposure amount with respect to the second exposure amount, the degree of influence of the defect on the location on the wafer is acquired,
For each of the plurality of locations of the mask pattern, based on the degree of influence of defects on the location on the wafer, the detection sensitivity of the defects at the location is acquired,
A program for creating an influence map of defects on the photomask on the wafer based on the detection sensitivity acquired for each of the plurality of portions of the mask pattern. With respect to a photomask including a mask pattern, the computer is caused to acquire the defect detection sensitivity at the location determined in accordance with the degree of influence of the defect on the location on the wafer for each of the plurality of locations of the mask pattern. Procedure and
A step of creating a defect influence degree map including the influence degree of the defect on the photomask on the wafer based on the defect detection sensitivity acquired for each of the plurality of portions of the mask pattern. The program of
In the procedure of creating the defect impact map,
For each of the plurality of portions of the mask pattern, the width of the portion is changed, and the photomask formed on the photomask corresponding to the portion where the width is formed on the wafer is managed. A transfer pattern corresponding to a management pattern for obtaining a first exposure amount necessary to finish the transfer pattern to a predetermined dimension, and the transfer pattern having a predetermined dimension for each of the plurality of portions of the mask pattern. Calculating a rate of change of the first exposure amount with respect to a second exposure amount necessary for finishing;
For each of the plurality of locations of the mask pattern, based on the rate of change of the first exposure amount with respect to the second exposure amount, the degree of influence of the defect on the location on the wafer is acquired,
For each of the plurality of locations of the mask pattern, based on the degree of influence of defects on the location on the wafer, the detection sensitivity of the defects at the location is acquired,
A program for creating an influence map of defects on the photomask on the wafer based on the detection sensitivity acquired for each of the plurality of portions of the mask pattern.

Images