JP2008083652A - Method for inspecting line width of photomask - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for inspecting the line width of a photomask, with which the line width of every complicated pattern on a photomask is estimated, validity of measurement results is verified, and the number of measurement points is decreased to reduce man-hours. <P>SOLUTION: The method includes steps of: selecting a straight line pattern as a simple pattern that hardly induces a measurement error in a photomask and actually measuring the line width and in-plane distribution (S4); using the actual measurement value as the reference of the photomask 1 and estimating a displacement or an error by considering the distribution on the photomask, an offset component and a random error component that cause a displacement or an error in each pattern (S8); selecting only a pattern in the patterns to be measured on the basis of the estimation result, the selected pattern having the most distant estimation value from the specification (S9); actually measuring the line width or the like of the pattern (S10); and deciding whether the pattern is within the specification or not and simultaneously deciding appropriateness whether the photomask can be used for manufacturing a semiconductor device or not (S12). <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a photomask line for measuring a monitor pattern of a photomask used for forming a fine pattern in the manufacture of various solid elements such as a semiconductor element and a magnetic device element, and determining the appropriateness for use in the manufacture of the solid element. It relates to the width inspection method.

In the manufacturing process of semiconductor devices, various semiconductor elements, magnetic device elements, dielectric device elements, etc., for example, monitor pattern shapes such as elements and wiring patterns designed using a CAD apparatus on a wafer using a photomask (reticle) Manufactured by transferring.
By the way, in recent semiconductor devices, circuit / wiring patterns have been miniaturized due to the development of high integration technology. Along with this high integration and miniaturization, there is an increasing demand for high-precision line widths for monitor patterns on photomasks, and the line width, shape, and number of patterns for which measurement guarantees are required. It is increasing rapidly.
However, at present, measurement is guaranteed for the dimensions of the wiring pattern by actually measuring all the monitor patterns having a plurality of types and line widths arranged on the photomask. For this reason, it is not uncommon for a single photomask to have several hundred measurement points. Therefore, if all the wiring patterns are actually measured, the throughput is poor, and the productivity is lowered, resulting in an obstacle to product manufacture.

  Therefore, Patent Document 1 includes an automatic measuring device, a computer that controls the operation, a display that displays the output of the computer, and an external memory that stores various types of information, and prevents human error in judgment. There has been proposed an automatic measuring apparatus for semiconductor wafer inspection that can reduce the burden on the operator and can make a countermeasure immediately after the measurement is completed. However, this automatic measuring device is simply a technique for automatically determining or re-measuring the measurement result in light of the acceptance / rejection standards, and is merely a technique for preventing artificial determination errors. Further, in this apparatus, it is unclear what causes a measured value to be abnormal in the case of remeasurement. For example, even if the measurement result is out of the standard, it may have been out of the standard due to a sudden measurement error even though it is within the standard. Conversely, even if the measurement result is within the standard, there is a possibility that it may actually be out of the standard. In other words, the automatic measuring apparatus disclosed in Patent Document 1 has a problem in that it cannot make a final and clear determination as to whether or not there is an abnormality.

By the way, in the measurement technology of complex shapes, various line widths, and enormous number of patterns in recent semiconductor manufacturing processes, analysis of measurement results, that is, obtained measurement values are normal or abnormal. The determination of whether or not there is more and more difficult.
FIG. 2 is a flowchart illustrating a conventional photomask line width inspection method.
In FIG. 2, the conventional photomask line width inspection method, for example, creates measurement conditions in a recipe file and executes it according to this (step S′1). Then, the created measurement conditions are set in an SEM (scanning electron microscope) which is a measurement device that actually measures the short dimension of the line width (step S′2), and then the photomask to be inspected is SEM. The sample is conveyed and placed on the sample stage (step S′3), and the line width is measured according to the set conditions (step S′4). At this time, line width measurement was performed for all patterns. Then, after all the points are measured, the operator examines the validity of the obtained measurement values and makes an acceptance / rejection determination and a remeasurement determination (step S′5).
If the operator determines that there is no problem, the photomask is removed from the SEM (step S′8) and transported to the next mask manufacturing process (step S′9). On the other hand, if the operator's determination is not appropriate to proceed to the next process, it is determined whether or not to re-measure (step S'6), and if re-measurement, the process returns to the automatic measurement in step S'4. Thereafter, the same operation is repeated. On the other hand, when the determination is made that the re-measurement is not performed and the validity of application to the next process is maintained, the photomask to be inspected is processed or disposed of as not used in the semiconductor manufacturing process (step S′7).
However, since the conventional photomask line width inspection method measures line widths for all patterns, a sudden measurement error occurs due to the influence of complicated optical proximity correction (OPC) pattern shapes. There is a problem that it is difficult to determine whether it is an error of the pattern itself or an error of the measurement itself, and it is left to the operator to make a final decision on the appropriateness to apply to the next process. There is a problem that it is difficult to accurately determine the validity of application to the next process.

  Generally, if it is possible to estimate in advance whether a specific measurement value is abnormal or normal, it is easy to determine whether the measurement value is abnormal or normal. As a conventional technique related to a method for predicting or estimating a measurement value, for example, in Patent Document 2, a bump height is measured for a sample region on a wafer, and a bump height distribution on the wafer is estimated based on the measurement value. The part where the estimated height distribution is outside the set range is taken as the inspection area, the bump height is measured within this inspection area, and based on the measurement result, the chip area including the bump whose height is outside the set range is determined to be defective. A bump appearance inspection method for judging is disclosed. This bump appearance inspection method accumulates data of the inspection area and the inspection result, and determines the inspection density based on the accumulated data. The method of estimating the measurement value in this inspection method is the bump height of the sample area. Measure the thickness, and estimate and guarantee the distribution of the in-wafer inspection area from the measured value. However, this method is too simple and cannot be applied to the estimation of the line width of a monitor pattern having various aperture ratios, line widths, and complicated OPC shapes such as a photomask.

JP 9-27525 A JP-A-11-325859

  The present invention has been made in view of the above-described problems of the prior art, and the problem is that a line width of a complicated monitor pattern on a photomask is estimated based on an actual measurement value, and the photomask is based on an estimation result. It is an object of the present invention to provide a photomask line width inspection method capable of verifying the validity of application to the next process and reducing the number of measurement points and man-hours.

The characteristics of the photomask line width inspection method of the present invention, which is a means for solving the above problems, are listed below.
According to the photomask line width inspection method of the present invention, first, a straight line portion is selected as a simple monitor pattern that is unlikely to cause a measurement error on the photomask surface, and this monitor pattern is used as a reference monitor pattern. The width and the in-plane distribution are measured, and the measured value is used as a reference value for the photomask. Next, the line width of a monitor pattern other than the reference monitor pattern is estimated based on the line width measurement value of the reference monitor pattern. The estimation of the line width is, for example, the X of the pattern as a random error component such as a linearity of the line width, a pattern interval, a pattern shape difference, etc. as an offset component that causes a displacement or an error generated for each monitor pattern.・ Considering differences in the Y direction, measurement reproducibility, and local line width error (Local CD Error).
Next, by estimating the displacement or error based on the estimated line width, only the pattern that is out of the standard or the pattern with the largest error is selected, the line width is measured, and the line width is measured based on the measured value. It is determined whether or not the line width of the monitor pattern is within the standard, and finally it is determined whether or not the photomask to be inspected can be used for manufacturing the semiconductor device in the next process. This makes it possible to quickly and easily determine the validity of the photomask.

Select not only the monitor pattern whose estimated value is out of the standard most, but also the monitor pattern near the standard limit, the monitor pattern that has been out of the standard most in the past, and measure the line width of the monitor pattern. The validity of the photomask may be determined based on the estimated value, the reference value, and the actually measured value.
Furthermore, all monitor patterns other than the reference monitor pattern can be selected, the total number of points can be measured, and the validity of the measured value of the photomask can be judged based on the measured value and the estimated value. In this case, in determining the appropriateness of the actual measurement value of the photomask, either a sudden pattern abnormality or a sudden measurement abnormality in the monitor pattern of the photomask is mainly caused by the displacement or error of the line width of the monitor pattern. Whether it is a factor or not can be determined.

According to the photomask line width inspection method of the present invention, it is possible to determine the guarantee of the line width on the photomask without actually measuring the line widths of all monitor patterns on the photomask. The number of man-hours in photomask inspection can be reduced. That is, when the estimated line width of the monitor pattern is out of the standard, or the line width is measured only for the monitor pattern with the largest error, or the estimated value of the pattern line width is within the standard, By actually measuring the line width, the man-hour can be greatly reduced by guaranteeing the line width of the photomask based on the actually measured value and the estimated value.
Further, according to the present invention, it is possible to make a quick and accurate determination without the operator's discretion.
Furthermore, according to the present invention, when the total number of points is measured, the validity of the deviation from the actually measured value is automatically determined. Accordingly, it is not necessary for a human to determine whether the measured value is abnormal or normal.

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.
The present invention relates to a photomask line width inspection method for inspecting a monitor pattern arranged in a plane of a photomask used when manufacturing a semiconductor device, and other methods based on an actual measurement value obtained by measuring the line width of a reference monitor pattern. The monitor pattern line width is estimated, and all pattern line widths on the photomask are guaranteed based on the estimated value. To estimate the line width of other monitor patterns other than the reference monitor pattern, the line width of the reference monitor pattern and its distribution in the photomask plane, the offset component for each pattern in the monitor pattern other than the reference monitor pattern, and random Error components.

FIG. 1 is a schematic view showing the configuration of a photomask applied to the implementation of the present invention.
In FIG. 1, the photomask 1 is partitioned into a circuit region 2 in which a pattern constituting an actual circuit is formed, and a dicing region 3 provided around the circuit region 2. Then, by exposing the resist film formed on the wafer using the photomask 1, the monitor pattern formed in the circuit region 2 is transferred to the resist film. Then, after the pattern transfer, development is performed, so that the transferred pattern remains or is removed depending on the positive type / negative type of the resist film, and other portions are removed or remain, thereby leaving the resist film Formed.

FIG. 3 is a flowchart showing a photomask line width inspection method according to an embodiment of the present invention, and FIG. 4 is a diagram showing a mask pattern on the photomask applied to this embodiment.
In FIG. 3, this photomask line width inspection method is executed according to a recipe file. The recipe file includes a measurement condition file in which coordinate information, a measurement direction, and the like are written, pass / fail judgment of measured values / estimated values, and measured values. And a judgment prediction file for estimating an estimated value from (step S1).
The measurement condition parameters in the measurement condition file include pattern coordinates, measurement direction, Black / White classification (measurement pattern removal or remaining classification), line width, and measurement area.
On the other hand, measurement value estimation parameters in the judgment prediction file include measurement error of measurement device, local line width abnormality, difference in X / Y direction, etc. as random error component, and offset component as pattern shape difference , Line width linearity, pattern interval, aperture ratio, and the like. Regarding the measurement value prediction parameter, the line width and shape of the photomask are thinned or thickened for each type of photomask (for example, resist negative type or positive type, etching method is dry type or wet type). It is known that the tendency is different, and since it has unique characteristics for each manufacturing process of the photomask 1, it is preferable to acquire basic data in advance. The pass / fail judgment parameters in the judgment prediction file include line width target value / line width standard (Error, Range) and the like.
Next, the measurement condition file of the recipe file is read by a control unit of a CD-SEM (Critical Dimension Scanning Electron Microscope) that measures the line width (step S2). The reason why the CD-SEM is used for measuring the line width is that the pattern formed on the photomask 1 tends to be miniaturized. However, if an image can be actually obtained, for example, an optical microscope using ultraviolet light, laser light, or the like may be used.

Next, the photomask 1 to be inspected is transported to the sample stage in the CD-SEM (step S3), and the line width measurement of the linear pattern which is the simplest monitor pattern is performed (step S4).
Here, as shown in (1) of FIG. 4, a simple straight line pattern is most suitable as a monitor pattern because a measurement error hardly occurs. The line width of a straight line pattern is measured by selecting five points within the surface of the photomask 1, for example, points A to E as shown in FIG. (Step S5).
Here, FIG. 4 shows an example of the monitor pattern arranged on the surface of the photomask 1, (1) is a straight line pattern which is a reference monitor pattern, and (2) to (6) Exemplifies patterns of active regions, gate electrodes, element isolation regions, metal wirings, and the like.
Next, based on the data grasped in step 5, it is determined whether or not the straight line pattern of the photomask 1 is within the standard (step S6). At this time, if any of the measured value, in-plane distribution, and average value in the simplest straight line pattern is out of specification, the pattern having another complicated shape is considered out of specification as another pattern. Cancel the remaining measurements, including. Then, the photomask 1 is taken out from the CD-SEM and processed out of specification (step S7).

On the other hand, if any of the measured value, in-plane distribution, and average value in the simplest straight line pattern obtained in step 5 is within the standard, the monitor has a complicated shape other than the straight line pattern. The line width of the pattern is estimated (step S8). At this time, examples of the offset component include linearity with respect to the line width and the interval between patterns. The offset component is a value representing the position of certain data as a difference (distance) from the reference point. Among the offset components, the linearity value with respect to the line width is estimated from the following graph. FIG. 5 is a graph showing the relationship between the target line width and the amount of displacement, which is linearity with respect to the size of the line width. This is used as a function by regression processing to evaluate the estimated value. FIG. 6 is a graph showing the relationship between the pattern interval and the error amount that this pattern interval gives to the line width. Here, the pattern interval on the horizontal axis is logarithmically displayed. In FIG. 6, the error amount given to the line width increases as the pattern interval becomes narrower. This exponential effect is regressed to a function and used to evaluate the estimated value. Here, the amount of error in the pattern interval at a line width of 400 nm is shown, but in reality, the calculation formula is created by a graph that matches the line width of the target value. The amount of error varies depending on the line width even if the interval is the same.
Further, examples of the random error component that is an error amount include reproducibility of measurement by a device such as an SEM, variation in line width of the drawing apparatus, and differences in X and Y axis directions due to directionality in the drawing apparatus. Random errors are errors that do not have a fixed relationship and occur suddenly. These offset component and random error component are fixed values on the production line. This is regarded as an inherent characteristic of individual production lines.
The offset component and random error component are not limited to these. Regarding the items and the number to be used as the offset component and the random error component, the weights to the factors depending on the production line differ. Therefore, it can select suitably by the weighting of the factor which affects a manufacturing line.

  Next, the monitor pattern estimated to be out of specification in the previous estimation in step 8 or the monitor pattern with the largest error is selected (step S9). At this time, if the reference monitor pattern is within the standard, the pattern to be measured is selected based on an empirical rule that the amount of error does not increase in each pattern. Note that the selection of the pattern to be actually measured has not been performed conventionally, and until now, the total number of points of all patterns has been actually measured. This is because there is no judgment criterion to select, and all the points have to be actually measured without making a selection. In contrast, in the photomask line width inspection method according to the present embodiment, the line width of a straight line pattern having the smallest displacement amount or error amount is measured first, and based on this measurement value, the photomask 1 is measured. The line width values of other monitor patterns are estimated by using the measurement conditions such as the measurement points and the estimation parameters including the offset component and the random error component. At this time, a monitor pattern to be actually measured is selected based on the estimated value. The selection of the monitor pattern need not be limited to only one, and may be a plurality. In particular, a pattern having a large error amount whose estimated value is farthest from the target value, a monitor pattern whose estimated value is not within the standard, and the like are selected. This makes it possible to significantly reduce the number of measurement points compared to the prior art in determining whether the manufactured photomask 1 is used for manufacturing a semiconductor device, and the validity of the photomask can be reduced with fewer man-hours. Can be easily determined.

Next, the line width of the monitor pattern selected in step 9 is actually measured by the SEM (step S10), and whether the estimated value estimated in step S8 is within the standard based on the actually measured value in the selected monitor pattern. And whether or not the photomask 1 has appropriateness for application to the exposure process in the next production line is determined (step S12). In step S12, it is determined whether or not the photomask 1 has a validity to be applied to the exposure process in the next production line, regardless of whether the line width measurement value of the monitor pattern is within or outside the standard. decide. In this way, it is possible to determine the appropriateness of application of the photomask 1 to the next process without actually measuring the line width of all monitor patterns on the photomask 1, so that the number of measurement points is greatly reduced, and photo Man-hours in mask inspection can be further reduced. That is, by measuring the line width only for monitor patterns whose pattern line width is estimated to be outside the standard, and when the estimated pattern line width is within the standard, the line width is set to the standard limit pattern. By actually measuring, man-hours can be greatly reduced. This also makes it possible to make a quick and accurate determination without the operator's discretion.
In the present embodiment, even if the linear pattern is out of the standard, other patterns can be appropriately selected and measured in order to reduce the number of processes depending on the magnitude of the error. Even in this case, the number of steps in the photomask inspection can be reduced. In addition, it is possible to determine the pass / fail of the photomask quickly and accurately without the discretion of the operator.
On the other hand, regardless of whether the linear pattern is within or outside the standard, all the monitor patterns on the photomask 1 are selected and the total number of points is measured (step S11). The measured value may be compared with the estimated value to determine whether the measured value is within or outside the standard, and then the validity of the measurement result of the photomask 1 may be determined. In this case, the validity judgment of the photomask is determined by whether or not the measured value is within the range of a “predictable error region” (random component) in the vicinity of the predicted value (offset component). That is, if the actual measurement value is not within the error region, it can be determined that there is a sudden line width defect or a measurement error. If the measured value does not fall within the error area of the predicted value even after re-measurement is performed, it is determined that the pattern line width is out of specification as a sudden error in the pattern line width. Therefore, the validity of the photomask measurement result is automatically determined. Accordingly, it is not necessary for a human to determine whether the measured value is abnormal or normal.

In step S12, if the actual measurement value of the line width of the monitor pattern is within the standard and it is determined that it is appropriate to flow to the next process of the mask manufacturing line, the photomask 1 is removed from the SEM document table. (Step S16), transported to the next process (Step S17). On the other hand, if it is determined in step S12 that the measured value is out of specification and the measured value is not valid, it is determined whether or not to remeasure (step S13). At this time, in the measurement when the number of man-hours to be measured is reduced, only a specific monitor pattern may be out of the standard, and even if the actual measurement value is out of the standard, depending on the monitor pattern, the exposure process of the wafer It may not be a big problem. Therefore, the necessity of remeasurement is determined in consideration of the type of the monitor pattern measured, the difference between the measured value and the estimated value, and the like.
If it is determined in step S13 that remeasurement is not necessary, the photomask 1 is taken out from the CD-SEM and processed as out of specification (step S14).

表１において、パターン（１）は、実際の測定値を示しており、パターン（１）の測定個所Ａ〜Ｅ点における目標値に対する変位量及びその平均値が示されている。このフォトマスクの場合１０ｎｍ未満の変位量が規格内とされている。このような規格内であれば、フォトマスクとして次工程の半導体製造に用いることが可能である。
表１において、パターン（１）は規格内なので、パターン（２）〜（６）において測定されるであろう推定値を推定する。パターン（２）〜（６）の推定値は、パターン（１）の線幅に対して、パターン（２）〜（６）のそれぞれのオフセット成分を加え、さらに誤差量としてのランダムエラー成分を考慮したものである。
ここで、オフセット成分は、前述した図５及び図６に示したように、線幅予測の計算式で算出したオフセット計算値に基づいて、パターン（１）の実測した誤差量を加えて計算した。表１に示すように、単純な直線のラインパターンの線幅のオフセット計算値が５．１ｎｍで、例えば、パターン（２）のオフセット計算値が１．８ｎｍであり、従って、パターン（２）のオフセット成分は−３．３ｎｍになる。これらは、これは、図５及ぶ図６に示したオフセット値の近似式にパターン（１）の線幅、スペース幅を代入して、それぞれの値を加算して計算した。
ランダムエラー成分は、実施例に用いた製造ラインが有する固有の値で、測定再現性を±０．５ｎｍ、局所線幅エラーを±１．０ｎｍ、パターン形状による線幅のエラーを±２．５ｎｍ、Ｘ・Ｙ差を±２．０ｎｍとした。
パターン（２）〜（６）の「ランダムエラー成分」は、それぞれ
パターン（２）：測定再現性及び局所線幅エラー
パターン（３）：測定再現性及び局所線幅エラー
パターン（４）：測定再現性、局所線幅エラー及びＸ・Ｙ差
パターン（５）：測定再現性及びパターン形状による局所異常
パターン（６）：測定再現性及びパターン形状による局所異常
である。
このようにして、パターン（１）における実際に測定された測定値を基に、オフセット成分、ランダムエラー成分から各測定個所の変位量を算出することができる。算出した変位量が、測定した場合の誤差として生ずる推定値である。この推定値と目標値との差である誤差量を、表１に表した。 Specific examples of the present invention will be described below.
First, based on the flowchart shown in FIG. 3, the line width of a simple straight line pattern with the least error in the test photomask (photomask 1) was measured. Table 1 shows the measurement results, specifically the relationship between the measured monitor pattern shape and the error amount at the measurement location on the photomask 1.

In Table 1, a pattern (1) shows an actual measurement value, and a displacement amount with respect to a target value and an average value thereof at measurement points A to E of the pattern (1) are shown. In the case of this photomask, the amount of displacement of less than 10 nm is within the standard. Within such a standard, it can be used as a photomask for semiconductor manufacturing in the next process.
In Table 1, since the pattern (1) is within the standard, the estimated value that will be measured in the patterns (2) to (6) is estimated. For the estimated values of patterns (2) to (6), the offset components of patterns (2) to (6) are added to the line width of pattern (1), and a random error component as an error amount is further considered. It is what.
Here, as shown in FIG. 5 and FIG. 6 described above, the offset component was calculated by adding the actually measured error amount of the pattern (1) based on the offset calculation value calculated by the formula for calculating the line width. . As shown in Table 1, the offset calculation value of the line width of a simple straight line pattern is 5.1 nm, for example, the offset calculation value of the pattern (2) is 1.8 nm. The offset component is −3.3 nm. These were calculated by substituting the line width and space width of the pattern (1) into the approximate expression of the offset value shown in FIGS. 5 and 6 and adding the respective values.
The random error component is a unique value of the production line used in the example, measurement reproducibility is ± 0.5 nm, local line width error is ± 1.0 nm, and line width error due to pattern shape is ± 2.5 nm. The X / Y difference was set to ± 2.0 nm.
“Random error components” of patterns (2) to (6) are: pattern (2): measurement reproducibility and local linewidth error pattern (3): measurement reproducibility and local linewidth error pattern (4): measurement reproduction , Local line width error and X / Y difference Pattern (5): Local anomaly due to measurement reproducibility and pattern shape Pattern (6): Local anomaly due to measurement reproducibility and pattern shape.
In this manner, the displacement amount at each measurement location can be calculated from the offset component and the random error component based on the actually measured value in the pattern (1). The calculated displacement amount is an estimated value generated as an error when measured. Table 1 shows an error amount which is a difference between the estimated value and the target value.

  FIG. 7 is a graph showing an error amount between the estimated value and the target value in Table 1. As can be seen from FIG. 7, it is the patterns (1) and (4) that greatly deviate this error amount to the plus side, and the patterns (3) and (6) that deviate greatly to the minus side. Therefore, the patterns (2) and (5) between them are unlikely to be out of specification. A pattern having a large error amount in FIG. 7 is selected to perform actual measurement. Since this product standard is (each line width average value) <(target value ± 10 nm), it can be confirmed that there is no problem if only the measurements of patterns (3), (4), and (6) are executed. Here, all the patterns (2) to (6) were actually measured.

FIG. 8 is a graph showing an error amount which is a difference between the actually measured value and the target value of the patterns (2) to (6). As can be seen from FIG. 8, if only the patterns (1), (4), and (6) are measured as a result, the error amount of the average value of the remaining other patterns can be guaranteed.
It can be seen that the measured value can be estimated within the range of the random error component value of the estimated value. Therefore, the validity of the estimated value is proved. The accuracy can be further improved by creating a database of shapes and line widths and reflecting it in the calculation formula.
In this embodiment, the number of patterns on the photomask 1 is six for convenience, but in reality, for example, there may be a case where measurement is guaranteed within the mask surface for more than 50 types of patterns. A significant reduction in man-hours can be expected especially when the number of patterns is large.

  In the present embodiment and the above-described embodiment, the present invention has been described as a method for inspecting the applicability of the photomask to the next process. However, this inspection method can be used on a substrate such as a semiconductor substrate (wafer) or a liquid crystal display panel. The present invention can be similarly applied to the inspection of the pattern formed on the substrate.

It is the schematic which shows the structure of a photomask. It is a figure which shows the flowchart of the conventional measuring method. It is a figure which shows the flowchart of the line | wire width inspection method of the photomask in embodiment of this invention. It is a figure which shows the mask pattern on the photomask used with the line | wire width inspection method of the photomask. It is a linearity with respect to the magnitude | size of line width, Comprising: It is a graph showing the relationship between a target line width and the amount of the displacement. It is a graph which shows the relationship between the amount of error given to pattern space | interval and line | wire width. It is a graph which shows the error amount between the estimated value in Table 1, and a target value. It is a graph which shows the error amount which is a difference of the measured value and target value of pattern (2)-(6).

Explanation of symbols

1: Reticle
2: Circuit area 3: Dicing area

Claims (5)

In a photomask line width inspection method used for manufacturing a semiconductor device,
The line width inspection method of the photomask measures a line width of a reference monitor pattern formed on the photomask, estimates a line width of a monitor pattern other than the reference monitor pattern based on a measurement value,
A method for inspecting the line width of a photomask, comprising: determining whether the line width of the monitor pattern formed on the photomask is appropriate based on an estimated value. In the photomask line width inspection method according to claim 1,
The photomask line width inspection method, wherein the reference monitor pattern is a linear pattern. In the photomask line width inspection method according to claim 1,
Estimating the line width of monitor patterns other than the reference monitor pattern is as follows:
The line width of the photomask, which is performed based on the line width of the reference monitor pattern and the distribution on the photomask surface, and an offset component and a random error component for each monitor pattern other than the predetermined reference monitor pattern Inspection method. In the photomask line width inspection method according to claim 1,
The photomask line width inspection method is:
Measure only the line width of the monitor pattern that is estimated to be out of specification among the monitor patterns other than the reference monitor pattern or the monitor pattern with the largest error, and guarantee the line width of all the patterns to ensure the line of the monitor pattern. A method for inspecting a line width of a photomask, characterized by determining whether or not the width is appropriate. In the photomask line width inspection method according to claim 1,
The photomask line width inspection method is:
Measuring the line widths of all monitor patterns other than the reference monitor pattern, and re-measuring only the line widths of monitor patterns whose measured values are out of the estimated value range, and determining the validity of the measured values. Photomask line width inspection method.

Images