JP2017015632A - Determination method for correction amount of pattern measurement, pattern measurement method and mold for imprint - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a correction amount determination method and a pattern measurement method for measuring fine patterns, such as a pattern of a mold used by an imprint method and a pattern formed by the imprint method, with high precision using CD-SEM, and a mold for imprint enabling high-precision pattern measurement.SOLUTION: A mold 11 for imprint has a base material 12, a main pattern positioned in a main pattern region 14 set on the base material 12, and a reference pattern positioned in a reference pattern region 15 set on the base material 12. The reference pattern has a plurality of kinds of unevenness periodic structure having projection parts and recessed parts arrayed alternately, and various unevenness periodic structures have projection parts equal in width, but different in pitch.SELECTED DRAWING: Figure 17

Description

  The present invention relates to a pattern measurement method, in particular, a correction amount determination method and a pattern measurement method for measuring a fine pattern such as a mold pattern used in an imprint method and a pattern formed by the imprint method with high accuracy. The present invention relates to an imprint mold capable of measuring a pattern with high accuracy.

In recent years, a pattern forming technique based on imprint lithography using an imprint method has attracted attention as a fine pattern forming technique that replaces the photolithography technique. The imprint method is a pattern formation technique in which a pattern member is transferred at an equal magnification by using a mold member (mold) having a fine uneven structure pattern and transferring the uneven structure to a resin material to be molded.
A pattern provided in such an imprint mold or a pattern formed by transfer using this imprint mold requires higher precision as the pattern becomes finer. Must be strictly managed as a CD (Critical Dimension). For such a pattern measurement, a scanning electron microscope for length measurement (hereinafter referred to as a CD-SEM) that has been conventionally used for measuring a photomask pattern (Patent Document 1) is used ( Patent Document 2).

JP 2013-72642 A JP 2011-124389 A

  However, when the result of measuring the concave portion of the pattern included in the imprint mold using the CD-SEM was examined, the difference between the measured value and the designed value varied depending on the size of the concave portion. Moreover, when the measurement result of the concave portion of the pattern included in the imprint mold by the CD-SEM is compared with the measurement result of the corresponding convex portion of the pattern transferred using this mold, there is a dimensional difference between the two. When such a dimensional difference is present, there has been a problem in confirming the accuracy of the mold and the pattern produced using the mold. Furthermore, when the result of measuring the concave portion of the pattern provided in the mold is measured by the X-ray small angle scattering method and the measurement result by CD-SEM, there is a difference between the two, and further investigation on the measurement result by CD-SEM is possible. It was necessary.

Therefore, the cross section of the concave portion of the pattern provided in the mold is actually measured by destructive inspection, and when this result is compared with the measurement result by the CD-SEM, in the measurement by the CD-SEM, the measured value is close to the actual measurement size of the bottom portion of the concave portion. It was confirmed that there was a tendency. On the other hand, the measurement result of the X-ray small angle scattering method substantially coincides with the result of actually measuring the cross section of the concave portion of the pattern by the destructive inspection, and it has been found that the measurement result is higher than the measurement by the CD-SEM. However, the X-ray small angle scattering method requires a relatively wide measurement area, and is not suitable for measurement of a mold having a narrow region having the same concavo-convex structure, and a pattern having a size exceeding 100 nm has a large size. As the distance between the fringes of the diffracted X-rays becomes narrower, the X-ray small angle scattering method becomes difficult to analyze, and the concavo-convex structure is provided with patterns of several dimensions of several tens to several hundreds of nanometers. There was a problem that it was not suitable for mold measurement. For this reason, it is practical to use a CD-SEM to measure a pattern included in an imprint mold or to measure a pattern formed by imprint using a mold. There is a demand for improved measurement accuracy.
The present invention has been made in view of the above circumstances, and a fine pattern such as a pattern of a mold used in an imprint method or a pattern formed by the imprint method is high using a CD-SEM. An object of the present invention is to provide a correction amount determination method, a pattern measurement method, and an imprint mold capable of high-precision pattern measurement for measuring with high accuracy.

  The present inventor has examined pattern measurement of a fine dimension using a CD-SEM. For example, when measuring a fine concavo-convex structure of a line / space shape with a CD-SEM, the half pitch, particularly a space (concave part) is measured. The width, the depth of the space, and the side wall angle of the space affect the measurement value, and in the tapered concave portion having the side wall angle such that the opening portion is wider than the bottom portion, the width of the concave portion decreases. The recess dimensions measured by the CD-SEM were separated from the dimensions of the bottom portion of the recesses (actually measured dimensions by destructive inspection), and found to tend to be dimensions in a portion near the opening portion of the recess, and the present invention was conceived.

  That is, the pattern measurement correction amount determination method of the present invention is a reference pattern having a plurality of concave and convex periodic structures in which convex portions and concave portions are alternately arranged, and the width of the convex portions among various concave and convex periodic structures is A reference pattern that is the same but has a different pitch of the convex portion is formed in advance under the same conditions as the pattern to be measured, and the convex portion width of each concavo-convex periodic structure of the reference pattern is measured with a scanning electron microscope. One kind of irregularity periodic structure is set as a reference irregularity periodic structure from the kind of irregularity periodic structure, and the measurement values of the projections of the other irregularity periodic structure excluding the reference irregularity periodic structure are measured by the scanning electron microscope. Calculate the error from the measurement value of the convex part of the periodic structure with the scanning electron microscope, and in each concave and convex periodic structure of the reference pattern, the value obtained by subtracting the measurement value of the convex part from the pitch of the convex part Calculated as the width of the concave portion of the period structure, and created a correction curve indicating the error corresponding to the width of the concave portion from the relationship between the width of the concave portion and the amount of error in each concave-convex periodic structure of the reference pattern, and using a scanning electron microscope The measured value of the measured concave portion of the pattern to be measured is applied to the correction curve, and the corresponding error is identified and used as the correction amount.

As another aspect of the present invention, each concavo-convex periodic structure constituting the reference pattern is configured to have concave portions having the same width with at least one convex portion interposed therebetween.
As another aspect of the present invention, the concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the base portion of the convex portion. The width of the bottom portion of the concave portion and the width of the top portion of the convex portion are the same.
As another aspect of the present invention, the concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the top portion of the convex portion. It was set as such a structure.

  The pattern measurement method of the present invention is a reference pattern formed under the same conditions as the pattern to be measured, and has a plurality of uneven periodic structures in which convex portions and concave portions are alternately arranged. The reference pattern measuring step of measuring the width of the convex part of each concave-convex periodic structure with a scanning electron microscope for a reference pattern having the same width but different pitch of the convex part, and a plurality of types of the concave-convex periodic structure One kind of concave / convex periodic structure is set as the reference concave / convex periodic structure, and the projections of the convex / concave periodic structure other than the reference concave / convex periodic structure are scanned with the scanning electron microscope. In addition to calculating the error from the measurement value obtained by a scanning electron microscope, in each uneven periodic structure of the reference pattern, the value obtained by subtracting the measured value of the convex part from the pitch of the convex part is the concave width of the concave / convex periodic structure. A correction curve creating step for creating a correction curve indicating an error corresponding to the recess width from the relationship between the error and the recess width in each concave-convex periodic structure of the reference pattern; A pattern measurement step measured by a scanning electron microscope and a measurement value of a concave portion of the pattern to be measured are applied to the correction curve, a corresponding error is specified as a correction amount, and the correction amount is calculated from the measurement value of the concave portion. And a correction step of subtracting the recess width.

As another aspect of the present invention, in the correction step, the correction amount is added to a measured value of a convex portion adjacent to the concave portion to obtain a convex portion width.
As another aspect of the present invention, each concavo-convex periodic structure constituting the reference pattern is configured to have concave portions having the same width with at least one convex portion interposed therebetween.
As another aspect of the present invention, the concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the base portion of the convex portion. The width of the bottom portion of the concave portion and the width of the top portion of the convex portion are the same.

  The mold for imprinting of the present invention has a base material, a main pattern located in a main pattern region set on the base material, and a reference pattern located in a reference pattern region set on the base material. The reference pattern has a plurality of uneven periodic structures in which convex portions and concave portions are alternately arranged, and the width of the convex portions is the same among various concave and convex periodic structures, but the pitch of the convex portions is different. The configuration is as follows.

As another aspect of the present invention, the base material includes a convex structure portion having a flat surface formed through a step and a periphery on one surface, and the main pattern region and the reference are formed on the surface of the convex structure portion. The configuration is such that a pattern is set.
As another aspect of the present invention, the main pattern area and the reference pattern area are separated from each other.
As another aspect of the present invention, the reference pattern region is positioned within the main pattern region.
As another aspect of the present invention, the base material includes a convex structure portion having a flat surface formed through a periphery and a step on one surface, and the main pattern region is set on the surface of the convex structure portion. The reference pattern is configured to be set on the base material around the convex structure portion.

As another aspect of the present invention, each concavo-convex periodic structure constituting the reference pattern is configured to have concave portions having the same width with at least one convex portion interposed therebetween.
As another aspect of the present invention, the concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the base portion of the convex portion. The width of the bottom portion of the concave portion and the width of the top portion of the convex portion are the same.
As another aspect of the present invention, the concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the top portion of the convex portion. It was set as such a structure.

As another aspect of the present invention, the width of the concave portion in the concave / convex periodic structure having the smallest convex portion pitch among the plurality of types of concave / convex periodic structures constituting the reference pattern is the largest among the concave portions constituting the main pattern. It was set as the structure which has below the width | variety of a recessed part with a narrow width | variety.
As another aspect of the present invention, the width of the concave portion in the concave-convex periodic structure having the widest convex pitch among the plural types of concave-convex periodic structures constituting the reference pattern is 200 nm or more. .
As another aspect of the present invention, the reference pattern has an inverted uneven periodic structure in which the unevenness of the plurality of types of uneven periodic structures is inverted. The pitch is different.

The correction amount determination method and the pattern measurement method of the present invention make it possible to measure a fine pattern with high accuracy using a CD-SEM.
Further, the imprint mold of the present invention can measure the pattern provided with high accuracy using a CD-SEM.

FIG. 1 is a partial cross-sectional view showing a cross section of a concave portion of a line / space shape pattern. FIG. 2 is a diagram showing the difference between the measured value by the CD-SEM and the actually measured width of the bottom portion in the concave portion. FIG. 3 is a diagram showing the relationship between the measured value of the concave portion by CD-SEM and the actually measured width of the bottom portion of the concave portion due to the difference in the width of the opening portion of the concave portion. FIG. 4 is a diagram for explaining the shadowing effect. FIG. 5 is a diagram for explaining the shadowing effect. FIG. 6 is a diagram for explaining the incident angle dependency of secondary electron generation. FIG. 7 is a diagram for explaining the edge effect. FIG. 8 is a diagram showing a contrast distribution of an electron beam by simulation. FIG. 9 is a diagram for explaining various line / space shapes in which the width of the convex portion is constant and the width of the concave portion is different. FIG. 10 is a diagram showing the contrast distribution of the electron beam obtained by the simulation in the line / space shape as shown in FIG. FIG. 11 is a diagram showing the contrast distribution of the electron beam obtained by simulation in the line / space shape as shown in FIG. FIG. 12 is a diagram for explaining the measurement position of the width of the concave portion due to the difference in the width of the opening portion of the concave portion. FIG. 13 shows the contrast distribution of the electron beam obtained by changing the inclination angle of the side wall of the recess and changing the width of the opening of the recess by simulation, and calculating the calculated recess width and the bottom of the recess. It is the figure which computed the dimensional difference with the width | variety, and plotted so that the dimensional difference of 190 nm with the widest space width might become a reference | standard. FIG. 14 is a diagram for explaining the determination of the correction amount of the measurement value of the width of the concave portion from the measurement value of the width of the convex portion. FIG. 15 is a diagram illustrating an example of a correction curve indicating the relationship between the error and the width of the recess. FIG. 16 is a diagram illustrating an example of a correction curve indicating an error corresponding to the recess width. FIG. 17 is a plan view showing an embodiment of an imprint mold of the present invention. FIG. 18 is a plan view showing another embodiment of the imprint mold of the present invention. FIG. 19 is a diagram for explaining an uneven periodic structure of a reference pattern in the imprint mold of the present invention. FIG. 20 is a plan view showing another embodiment of the imprint mold of the present invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
Note that the drawings are schematic or conceptual, and the dimensions of each member, the ratio of sizes between the members, etc. are not necessarily the same as the actual ones, and represent the same members. However, in some cases, the dimensions and ratios may be different depending on the drawing.

  FIG. 1 is a partial cross-sectional view showing a cross section of a concave portion of a line / space shape pattern. In the line / space pattern 1 shown in FIG. 1, lines (convex portions) 2 and spaces (concave portions) 3 are alternately arranged, and the depth D of the concave portions 3 corresponds to the height of the convex portions 2. The inclination angle θ of the side wall 3a of the recess 3 is a predetermined angle of 90 ° or less. The width dimension of the concave portion 3 of such a pattern 1 is measured with a scanning electron microscope for length measurement (hereinafter referred to as CD-SEM), and this is taken as a measurement value Lm. On the other hand, the cross section of the concave portion 3 of the pattern 1 is actually measured by destructive inspection, and the width Lb of the bottom portion 3b of the concave portion 3 and the width Lt of the opening portion 3c of the concave portion 3 are obtained. Then, in the line / space pattern 1 having various dimensions and varying the width Lt of the opening portion 3c with the inclination angle θ constant, the measured value Lm in the CD-SEM and the actually measured width of the bottom portion 3b in the recess 3 Compare with Lb. FIG. 2 is a diagram showing the difference between the measured value Lm by the CD-SEM and the actually measured width Lb of the bottom portion 3b in the concave portion 3 where the inclination angle θ of the side wall portion 3a is 83 °. As shown by a triangular plot in FIG. 2, as the width Lt of the opening 3c of the recess 3 becomes narrower, the measured value Lm measured by the CD-SEM becomes larger than the measured width Lb of the bottom 3b. This is the size of the recess 3 near the opening 3c. That is, as shown in FIG. 3, when the width Lt of the opening 3c of the recess 3 is wide (FIG. 3A) and the case where the width Lt is narrow (FIG. 3B), the opening 3c of the recess 3 is compared. When the width Lt is large, the measured value Lm measured by the CD-SEM is close to the actually measured width Lb of the bottom portion 3b, but when the width Lt of the opening portion 3c of the recess 3 is narrow, the CD-SEM The measured value Lm to be measured is a dimension at a portion near the opening portion 3c of the recess 3 and is larger than the actually measured width Lb of the bottom portion 3b. Therefore, as the pattern dimension decreases, the amount of deviation of the measurement result using the CD-SEM increases and the reliability of the measurement value decreases.

  The deviation between the measured value Lm measured by the CD-SEM as described above and the measured width Lb of the bottom portion 3b increases as the width of the concave portion 3 becomes narrower, and the secondary effect generated at the side wall portion 3a. This is thought to occur because the amount of detected electrons changes. The shadowing effect and the amount of secondary electrons generated below are described in CGFrace, E. Buhr, K. Dirscherl, Meas. Sci. Technol., Vol. 18, pp 510-519 (2007), “ Scanning electron microscope for nanotechnology ", PP36-46, edited by Japan Surface Science Society, Maruzen Co., Ltd., February 25, 2004

  This shadowing effect will be described with reference to FIGS. As shown in FIG. 4A, the irradiated electrons collide with the point P on the side wall 3a of the recess 3 (shown on the left side of FIG. 4A), and the point P on the bottom 3b of the recess 3 2 (shown on the right side of FIG. 4A), secondary electrons (partially indicated by chain arrows) are generated from the surface of the recess 3. The amount of secondary electrons generated increases in the direction in which the emission angle α is small. However, when the width of the opening 3c of the recess 3 is narrow, the range of secondary electrons generated from the surface of the recess 3 and passing through the opening 3c (the region indicated by sand in FIG. 4A) is narrow and detected. The amount of secondary electrons detected by the device is small. Secondary electrons that cannot pass through the opening 3c collide with the side wall 3a of the recess 3 and are reabsorbed. As the position of the point P on the side wall 3a of the recess 3 approaches the opening 3c, secondary electrons passing through the opening 3c include those having a smaller emission angle α and are detected by the detector. The amount of secondary electrons increases. FIG. 4B conceptually shows the amount (indicated by a chain line) of the secondary electrons generated in the recess 3 passing through the opening 3c in this manner corresponding to the side wall 3a and the bottom 3b of the recess 3. FIG. As shown in the figure, when the width of the opening 3c of the recess 3 is narrow, the detection amount of secondary electrons is low when the position of the point P is close to the bottom 3b of the side wall 3a, and the detection becomes closer to the opening 3c. The amount of secondary electrons to be increased. In FIG. 4B, the inclination angle of the side wall 3a is made small in order to make it easy to display the change in the detected amount of secondary electrons.

On the other hand, as shown in FIG. 5A, when the width of the recess 3 is increased, the emission angle among the secondary electrons generated at the point P located at the same depth as the above in the side wall portion 3a of the recess 3. Secondary electrons having a smaller α also pass through the opening 3c. Further, the amount of secondary electrons generated at the point P of the bottom portion 3b of the recess 3 also increases through the opening portion 3c. FIG. 5B conceptually shows the amount of secondary electrons generated in the concave portion 3 passing through the opening portion 3c (indicated by a chain line) corresponding to the side wall portion 3a and the bottom portion 3b of the concave portion 3. FIG. As shown in the figure, when the width of the opening 3c of the recess 3 is wide, even if the position of the point P is closer to the bottom 3b of the recess 3 compared to FIG. The amount of secondary electrons increases. In FIG. 5B, the inclination angle of the side wall 3a is made small in order to make it easy to display the change in the detected amount of secondary electrons.
Apart from such shadowing effect, as shown in FIG. 6, when the inclination angle θ (see FIG. 1) of the side wall 3a of the recess 3 is large, and therefore the incident angle θ of electrons is large (FIG. 6A). ) Secondary electron generation amount (shown for convenience with a chain line) compared to secondary electron generation amount (shown with a chain line for convenience) when the electron incident angle θ is small (FIG. 6B). It is incident angle dependent. Further, as shown in FIG. 7, the secondary electrons generated by the absorption of the electrons in the convex portion 2 are diffused, and the secondary electrons are emitted from the side wall portion 3 a of the concave portion 3 (a part is indicated by a chain line) and detected. There is a so-called edge effect. In FIG. 7, the amount of secondary electrons generated from the side wall 3a due to the edge effect is indicated by a chain line for convenience. The incident angle dependence and etching effect of secondary electron generation are not affected by the width of the recess 3 unlike the shadowing effect.

Therefore, a simulation of the contrast distribution of the electron beam detected in the measurement of the concave portion 3 using the CD-SEM is performed in consideration of the above shadowing effect, incident angle dependency, and etch effect.
FIG. 8 shows a contrast distribution obtained by such a simulation with a two-dot chain line. In this case, the inclination angle of the side wall 3a of the recess 3 is 83 °. The distance (the distance indicated by the arrow in FIG. 8) at the location where the contrast distribution has the maximum slope (the location where the absolute value of the differential curve of the contrast distribution (shown by the chain line in FIG. 8) becomes the maximum) is set as the width of the recess. To do. With the simulation in such a setting, the width of the concave portion 3 is calculated for a line / space shape pattern with various dimensions in which the width of the opening 3c is changed, and is shown by the two-dot chain line in FIG. -It substantially coincides with the measured value of the recess 3 using SEM. Therefore, it can be confirmed that the correction amount in the pattern measurement using the CD-SEM can be examined by the simulation of the electron beam contrast distribution in consideration of the shadowing effect, the incident angle dependency, and the etching effect. In the following, the amount of correction in pattern measurement using a CD-SEM will be examined by simulation.

  In the above simulation, the distance at the position where the contrast distribution of the detection electron beam has the maximum inclination is set as the width of the concave portion, but the distance between the peaks of the contrast distribution as shown by the two-dot chain line in FIG. Also in the simulation for setting the width of the recess and the simulation for setting the distance at the location that is about 60% of the peak value of the contrast distribution as the width of the recess, the measured value of the recess 3 using the CD-SEM. Is approximately the same.

  In the examination of the correction amount in the pattern measurement using the CD-SEM, first, as shown in FIG. 9, various line / space shapes in which the width of the convex portion 2 is constant and the width of the concave portion 3 is different are described above. The contrast distribution of the electron beam is obtained by the simulation. In FIG. 9, a line / space in which the width of the concave portion 3 is narrow is indicated by a solid line, and a line / space in which the width of the concave portion 3 extends in the direction of arrow a is partially indicated by a two-dot chain line. 10 and 11 are diagrams showing the contrast distribution of the electron beam in such a line / space shape. 10, the contrast distribution (solid line) of the electron beam when the inclination angle θ of the side wall 3a of the recess 3 is 85 °, the depth D of the recess 3 is 64 nm, and the width Lt of the opening 3c of the recess 3 is 28 nm. And the contrast distribution (two-dot chain line) of the electron beam when the width Lt of the opening 3c of the recess 3 is 100 nm. Below these contrast distributions, the secondary electron generation amount (solid line) when the width Lt of the opening 3c of the recess 3 is 28 nm, and the secondary when the width Lt of the opening 3c of the recess 3 is 100 nm. The amount of electron generation (two-dot chain line) is shown. When the width Lt is 100 nm, only the electron beam contrast distribution and the secondary electron generation amount (two-dot chain line) in one side wall 3a of the recess 3 are shown. In the example shown in FIG. 10, the location K1 where the contrast distribution of the electron beam has the maximum inclination when the width Lt of the opening 3c of the recess 3 is 28 nm is the same as that when the width Lt of the opening 3c of the recess 3 is 100 nm. The electron beam contrast distribution is located on the outer side (left side in the figure) of the portion K2 where the maximum inclination is obtained. This is because, as the width Lt of the opening 3c of the recess 3 becomes narrower, the position on the pattern corresponding to the recess width calculated by the CD-SEM (hereinafter referred to as the measurement position of the recess width) is the opening of the recess 3. It shows that it becomes close to 3c.

  In FIG. 11, the contrast distribution of the electron beam when the inclination angle θ of the side wall 3a of the recess 3 is 88 °, the depth D of the recess 3 is 64 nm, and the width Lt of the opening 3c of the recess 3 is 28 nm ( The solid line) and the electron beam contrast distribution (two-dot chain line) when the width Lt of the opening 3c of the recess 3 is 100 nm. Below these contrast distributions, the secondary electron generation amount (solid line) when the width Lt of the opening 3c of the recess 3 is 28 nm, and the secondary when the width Lt of the opening 3c of the recess 3 is 100 nm. The amount of electron generation (two-dot chain line) is shown. When the width Lt is 100 nm, only the electron beam contrast distribution and the secondary electron generation amount (two-dot chain line) in one side wall 3a of the recess 3 are shown. Also in the example shown in FIG. 11, the location K1 where the contrast distribution of the electron beam has the maximum inclination when the width Lt of the opening 3c of the recess 3 is 28 nm is when the width Lt of the opening 3c of the recess 3 is 100 nm. Is located on the outer side (left side in the figure) of the portion K2 where the contrast distribution of the electron beam becomes the maximum inclination. This indicates that as the width Lt of the opening portion 3 c of the recess 3 becomes narrower, the measurement position of the recess width becomes closer to the opening portion 3 c of the recess 3.

  From the contents shown in FIGS. 10 and 11, it can be said as shown in FIG. That is, when the width Lt of the opening 3c of the recess 3 is narrow (FIG. 12A), the point K1, which is the measurement position of the recess, is closer to the opening 3c of the recess 3, and the width Lm of the recess 3 at the point K1 A dimensional difference with the width Lb of the bottom part 3b of the recessed part 3 becomes large. On the other hand, when the width Lt of the opening 3c of the recess 3 is wide (FIG. 12B), the point K2 that is the measurement position of the recess width is closer to the bottom 3b of the recess 3, and the measured value Lm at the point K2 and the recess 3 The dimensional difference with the width Lb of the bottom portion 3b of the metal plate becomes small.

  In FIG. 13, the depth D of the recess 3 is 75 nm, the inclination angle θ of the side wall 3a of the recess 3 is three types of 88 °, 83 °, and 78 °, and the width Lt of the opening 3c of the recess 3 is set. The contrast distribution of the electron beam when it is changed stepwise from 28 nm to 190 nm is obtained by the above-mentioned simulation, and the distance of the portion where these contrast distributions have the maximum inclination is set as the width of the recess 3. The width Lm is calculated, the dimensional difference between the calculated recess width Lm and the width Lb of the bottom portion of the recess 3 is calculated, and the dimensional difference at 190 nm, which is the widest width Lt of the opening 3c of the recess 3, is zero. That is, it is plotted so that a dimensional difference of 190 nm having the widest space width becomes a reference. The width Lb of the bottom portion of the concave portion 3 is obtained by [Lt−2D / tan θ] from the depth D of the concave portion, the inclination angle θ of the side wall portion 3a of the concave portion 3, and the width Lt of the opening portion 3c of the concave portion 3. . And the approximated curve was calculated | required using the least squares method from these 3 types of plots shown by this FIG. 13, and these were shown with the chain line. Of these three types of approximate curves, paying attention to the approximate curve in which the inclination angle θ of the side wall 3a of the recess 3 is 83 °, the dimensional difference when the width Lt of the opening 3c of the recess 3 is 24 nm is about 8 nm. This corresponds to the amount of difference in the measured value Lm in the CD-SEM with respect to the measured width Lb of the bottom portion 3b in the recess 3 when the width Lt of the opening 3c of the recess 3 shown in FIG. 2 is 24 nm. Also from this point, it can be confirmed that the correction amount in the pattern measurement using the CD-SEM can be examined by the simulation as described above.

  Therefore, the approximate curve indicated by the chain line in FIG. 13 can be used as a correction curve when the inclination angle θ of the side wall 3a of the recess 3 is 88 °, 83 °, and 78 °. In the example shown in FIG. 13, the correction accuracy is highest when the width Lt = 190 nm of the opening 3c of the recess 3 is the reference, for example, when the width Lt = 150 nm of the opening 3c of the recess 3 is the reference. The accuracy of correction is slightly reduced. In addition, when the inclination angle θ of the side wall 3a of the recess 3 is 88 °, high-accuracy correction is possible even with the width Lt = 130 nm of the opening 3c of the recess 3 as a reference.

  Here, when the width of the concave portion is measured at the point where the contrast distribution of the electron beam has the maximum inclination, the width of the convex portion is also measured at the point where the contrast distribution of the electron beam has the maximum inclination. As shown in FIG. 12 described above, when the width Lt of the opening 3c of the recess 3 is narrow (FIG. 12A), the measured value Lm at the point K1 at which the contrast distribution of the electron beam has the maximum inclination is the width of the recess 3. Since the uneven pitch is P1, the width of the convex portion 2 at the point K1 is (P1-Lm). And since the point K1 which is a measurement position of a recessed part width is near the opening part 3c of the recessed part 3, the width | variety (P1-Lm) of the convex part 2 becomes a thing narrower than the width | variety in the base 2b of the convex part 2. FIG. On the other hand, when the width Lt of the opening 3c of the recess 3 is wide (FIG. 12B), the measured value Lm at the point K2 where the contrast distribution of the electron beam has the maximum inclination is measured as the width of the recess 3, and the pitch of the unevenness Is P2, the width of the convex portion 2 at the point K2 is (P2-Lm). And since the point K2 which is a measurement position of a recessed part width is near the bottom part 3b of the recessed part 3, the width | variety (P2-Lm) of the convex part 2 becomes a thing close | similar to the width | variety in the base 2b of the convex part 2. FIG. This is because the measurement of the width of the convex part 2 by the CD-SEM is affected by the shadowing effect in the adjacent concave part 3, and the width of the convex part 2 becomes smaller as the width of the adjacent concave part 3 becomes narrower. It shows that it becomes the width | variety near top part 2c. Therefore, the convex part 2 can be measured and a correction amount can be determined based on this. This will be described below.

  As shown in FIG. 14, a line / space-shaped uneven periodic structure is assumed in which the width of the convex portion 2 is constant and the width of the concave portion 3 is different. 14A has a pitch (projection cycle) of the convex portions 2 of P1, and the projection / recess cycle structure shown in FIG. 14B has a pitch pitch of the projections 2 of P2 (projection cycle) of P2. In the concavo-convex periodic structure shown in FIG. 14C, the pitch (concave / convex period) of the convex portions 2 is P3, and P1 <P2 <P3. In such three types of concavo-convex periodic structures, the width of the convex portion 2 is constant, and each of the concave and convex portions 3 has the same width with at least one convex portion 2 interposed therebetween. This is because it is necessary that the concave portions 3 exist on both sides of the convex portion 2 in order for the measurement of the width of the convex portion 2 by the CD-SEM to be affected by the shadowing effect in the adjacent concave portion 3.

  When the width of the convex portion 2 is measured with respect to the above three types of concave / convex periodic structures, the measurement position of the convex portion 2 is the opening portion of the concave portion 3 as the width of the concave portion 3 becomes narrower as described in FIG. It is close to 3c. In FIG. 14, the measurement position of the convex part 2 with the pitch P1 is shown as a point K1, the measurement position of the convex part 2 with the pitch P2 is shown as a point K2, and the measurement position of the convex part 2 with the pitch P3 is shown as a point K3. Then, assuming that the measured values of the width of the convex portion 2 at the points K1, K2, and K3 are Lm1, Lm2, and Lm3, respectively, Lm1 <Lm2 <Lm3 as described in FIG. Since the point K3 is closer to the base 2b of the convex part 2, the measured value Lm3 of the width of the convex part 2 at the point K3 is close to the width of the base 2b of the convex part 2. 14C is set as a reference irregularity periodic structure, and the measured value Lm3 of the projection 2 of the reference irregularity periodic structure and the projection 2 An error with respect to the measured value of the convex part 2 of the concave-convex periodic structure whose pitch is smaller than P3 is obtained. That is, the error from the measured value Lm1 of the convex part 2 of the concave-convex periodic structure (FIG. 14A) in which the pitch of the convex part 2 is P1 is (Lm3-Lm1). Further, an error from the measured value Lm2 of the convex part 2 of the concave-convex periodic structure (FIG. 14B) in which the pitch of the convex part 2 is P2 is (Lm3−Lm2). On the other hand, the width of the concave portion 3 in each concave-convex periodic structure in which the pitch of the convex portion 2 is P1, P2, and P3 is (P1-Lm1), (P2) from the measured values Lm1, Lm2, and Lm3 of the convex portion 2, respectively. -Lm2), (P3-Lm3). That is, the width of the concave portion 3 in the concave-convex periodic structure (FIG. 14A) in which the pitch of the convex portions 2 is P1 is the width at the point K1. Similarly, the width of the concave portion 3 in the concave-convex periodic structure (FIG. 14B) in which the pitch of the convex portions 2 is P2 is the width at the point K2, and the concave-convex periodic structure in which the pitch of the convex portions 2 is P3 (see FIG. 14 (C)), the width of the recess 3 is the width at the point K3.

  Then, the horizontal axis represents the width of the concave portion 3, and the vertical axis represents the error between the measured value of the convex portion 2 of the reference concave / convex periodic structure and the measured value of the convex portion 2 of the concave / convex periodic structure having a small pitch, Three points: width (P1-Lm1) and error (Lm3-Lm1), width (P2-Lm2) and error (Lm3-Lm2), width (P3-Lm3) and error (0) Is plotted as shown in FIG. By obtaining an approximate curve from the plot shown in FIG. 15 using the least square method, this approximate curve becomes a correction curve indicating the relationship of the error to the width of the recess 3. In FIG. 15, the correction curve is indicated by a two-dot chain line for convenience. Therefore, for example, when the measurement value Lm is obtained as the width of the recess 3 to be measured, a correction amount ΔLm corresponding to the measurement value Lm is obtained from the correction curve as shown in FIG. Then, by subtracting the correction amount ΔLm from the measurement value Lm of the recess 3 to be measured, the width of the recess 3 can be obtained as (Lm−ΔLm). By performing such correction, a width approximate to the width of the bottom portion 3b can be obtained for the recess 3 to be measured.

14 and 15, for the sake of convenience, it has been described that the correction curve is obtained from three types of concave / convex periodic structures in which the pitch of the convex portions 2 (concave / convex cycle) is P1 <P2 <P3. By increasing the number of concave and convex periodic structures having different pitches, the accuracy of the correction curve can be made higher.
From the examination of the correction amount in the pattern measurement using the CD-SEM by such a simulation, the determination method of the correction amount of the pattern measurement, the pattern measurement method, and the imprint mold of the present invention were created.

[Method of determining the correction amount for pattern measurement]
In the pattern measurement correction amount determination method of the present invention, the reference pattern is formed in advance under the same conditions as the pattern to be measured.
The reference pattern to be formed is a reference pattern having a plurality of irregular periodic structures in which convex portions and concave portions are alternately arranged. As shown in FIG. 14 described above, the widths of the convex portions are the same, but the pitches of the convex portions are different among the plurality of types of irregular periodic structures constituting this reference pattern. The reason why such a reference pattern is formed under the same conditions as the pattern to be measured is that the inclination angle of the side wall portion of the concave portion of the reference pattern is the same as the inclination angle of the side wall portion of the concave portion of the measurement pattern. Accordingly, if the reference pattern and the pattern to be measured are formed under the same conditions, the formation of the reference pattern may be performed simultaneously with the formation of the pattern to be measured, or as a separate process from the formation of the pattern to be measured. Good.
The type of the concave-convex periodic structure constituting the reference pattern, that is, the type of the pitch of the convex portion affects the accuracy of the correction curve created as described later, and the more it is preferable, the more the number is five or more, preferably ten As mentioned above, More preferably, it is 20 or more types. In the present embodiment, for convenience, it is assumed that N types (N is an integer of 5 or more) of pitches P1 to PN constitute a reference pattern. Each concave-convex periodic structure constituting the reference pattern must have concave portions having the same width with at least one convex portion interposed therebetween. This is because the shadowing effect in the adjacent concave portion affects the measurement of the width of the convex portion by the CD-SEM.

Further, among the N kinds of irregular periodic structures constituting the reference pattern, the irregular periodic structure having the smallest convex pitch has the same width of the opening of the concave portion and the width of the base of the convex portion, and the bottom portion of the concave portion. It is preferable that the width of the projection and the width of the top of the projection are the same. In this case, it is preferable that the width of the concave portion in the concave-convex periodic structure is equal to or smaller than the width of the concave portion having the narrowest width among the concave portions constituting the measurement target pattern. This makes it possible to create a correction curve, which will be described later, based on a plot in a fine recess to be measured, rather than extrapolation based on a plot with a large recess width, and the accuracy of the correction curve is higher. . Note that the measurement of the width of the convex portion by CD-SEM depends on the inclination angle of the side wall portion, and it is difficult to define the width of the base portion of the convex portion when a round exists in the side wall portion. In the uneven periodic structure having the smallest convex pitch among the N types of irregular periodic structures constituting the pattern, the width of the opening portion of the concave portion and the width of the top portion of the convex portion may be the same.
Next, while measuring the convex part width | variety of each uneven | corrugated periodic structure which comprises a reference pattern with CD-SEM, one type of uneven | corrugated periodic structure is set as a reference | standard uneven | corrugated periodic structure from N types of uneven | corrugated periodic structures.
In the measurement of the convex portion by the CD-SEM, the distance of the portion where the contrast distribution has the maximum inclination can be set as the width of the convex portion based on the contrast distribution of the electron beam obtained by the measurement. Further, the distance between the peaks of the contrast distribution may be set as the width of the convex portion, and further, the distance at a location that is 60% of the peak value of the contrast distribution may be set as the width of the convex portion.

  The standard concavo-convex periodic structure from among the N concavo-convex periodic structures is normally set by selecting the concavo-convex periodic structure having the largest convex pitch among the N concavo-convex periodic structures constituting the reference pattern. It is preferable. As described above with reference to FIG. 13, the concave / convex periodic structure with the largest convex pitch, that is, the concave / convex periodic structure with the largest concave width is used as a reference concave / convex periodic structure as described later. This is because the accuracy of the correction curve to be created is higher. Of course, in the same manner as described above with reference to FIG. 13, depending on the inclination angle of the side wall of the concave portion, even if the concave / convex periodic structure that is not the concave / convex periodic structure having the largest convex pitch is used as the reference concave / convex periodic structure, highly accurate correction is possible. In some cases, a curve may be obtained, and the present invention is not limited to setting the concave-convex periodic structure having the largest concave portion width as the reference concave-convex periodic structure.

Next, with respect to the other irregular periodic structures excluding the reference irregular periodic structure, the difference between the measured value by the CD-SEM of the convex portion and the measured value by the CD-SEM of the convex portion of the reference irregular periodic structure is calculated. For example, when the concave / convex periodic structure having the largest convex pitch is set as the standard concave / convex periodic structure among N types of concave / convex periodic structures constituting the reference pattern, (N-1) species. Therefore, in this case, with respect to the measured values Lm1 to Lm (N-1) by the CD-SEM of the convex portions in the (N-1) types of concave and convex periodic structures excluding the reference concave and convex periodic structures, the CD of the convex portions of the standard concave and convex periodic structures. -Errors (LmN-Lm1) to (LmN-Lm (N-1)) from the measured value LmN by SEM are calculated. This error is obtained as (N-1) values.
Next, in the N types of uneven periodic structures of the reference pattern, values (P1−Lm1) to (PN−LmN) obtained by subtracting the measured values Lm of the corresponding protrusions from the pitch P of the respective protrusions are used as the uneven period. Calculated as the recess width of the structure.

Next, the recess widths (P1−Lm1) to (PN−LmN) in the N types of periodic patterns of the reference pattern and the errors (LmN−Lm1) to (LmN−Lm (N−1) calculated as described above. ) And (0), a correction curve indicating an error corresponding to the recess width is created as shown in FIG. In FIG. 16, the horizontal axis represents the recess width (P-Lm), the vertical axis represents the error (LmN-LmK) (K is an integer from 1 to (N-1)), and a minimum of two from the N plots. An approximate curve is obtained using multiplication, and this is used as a correction curve. In FIG. 16, the correction curve is indicated by a two-dot chain line for convenience, and the number of plots shown in FIG. 16 is for convenience.
Next, the concave portion of the pattern to be measured is measured with a CD-SEM, the obtained measurement value (L′ m) is applied to the correction curve, the corresponding error (ΔL′m) is specified, and the correction amount And The correction amount (ΔL′m) specified in this way is subtracted (L′ m−ΔL′m) from the measured value (L′ m) of the concave portion by CD-SEM, thereby obtaining the width of the bottom portion of the concave portion. An approximate measurement result is obtained. Note that the conditions for measuring the concave portions of the pattern to be measured by the CD-SEM are the same as the conditions for measuring the convex width of each concave-convex periodic structure constituting the above-described reference pattern by the CD-SEM.

[Imprint mold]
FIG. 17 is a plan view showing an embodiment of an imprint mold of the present invention. In FIG. 17, the imprint mold 11 includes a base material 12, a concavo-convex structure forming region 13 (region surrounded by a one-dot chain line) set on one surface 12 a of the base material 12, and the concavo-convex structure forming region. It has a main pattern located in the set main pattern area 14 and a reference pattern located in the reference pattern area 15 set in the concavo-convex structure forming area 13.
The material of the base material 12 can be appropriately determined according to the use conditions of the imprint mold 11. For example, when the molding resin material used for imprinting is photocurable, a material that can transmit irradiation light for curing them can be used, such as quartz glass, silicate glass, calcium fluoride. In addition to glass such as magnesium fluoride and acrylic glass, sapphire and gallium nitride, resin such as polycarbonate, polystyrene, acrylic, and polypropylene, or any laminated material thereof can be used. In addition, when the molding resin material to be used is not photo-curable, or when it is possible to irradiate light for curing the molding resin material from the transfer substrate side, the mold needs to have optical transparency. Therefore, the substrate 12 may not be a light-transmitting material. In addition to the above materials, for example, metals such as silicon, nickel, titanium, and aluminum, and alloys thereof, oxides, nitrides, or These arbitrary laminated materials can be used.

Further, the base material 12 may have a so-called mesa structure in which a flat surface is provided with a convex structure portion formed on the one surface through a periphery and a step. When the base material 12 has a mesa structure, in the example shown in FIG. 17, the concavo-convex structure forming region 13 is set on the surface of the convex structure portion.
In the illustrated example, the main pattern regions 14 are set at three locations of the concavo-convex structure forming region 13, but the present invention is not limited to this. The main pattern located in the main pattern region 14 may be a line / space-shaped uneven structure, a hole-shaped recessed structure, or the like, and is not particularly limited. Further, the pattern size of the main pattern is not particularly limited. For example, in the case of a line / space-shaped concavo-convex structure, the aspect ratio (ratio of length to width in plan view) of the concave portion or the convex portion is 3 or more. Is preferred.
In the illustrated example, the reference pattern region 15 set in the concavo-convex structure forming region 13 is at a position separated from the main pattern region 14, but as shown in FIG. 18, within the main pattern region 14, for example, three main pattern regions It may be located in each of the pattern areas 14.
As will be described later, the reference pattern located in the reference pattern region 15 has a plurality of uneven periodic structures in which convex portions and concave portions are alternately arranged, and the widths of the convex portions are the same among various uneven periodic structures. However, the pitch of the convex portions is different.

This reference pattern is preferably formed under the same conditions as the main pattern. This is because the inclination angle of the side wall portion of the concave portion of the reference pattern and the inclination angle of the side wall portion of the concave portion of the main pattern are the same. Therefore, the formation of the reference pattern may be performed simultaneously with the formation of the main pattern, or may be a separate process from the formation of the main pattern.
FIG. 19 is a diagram for explaining an uneven periodic structure included in a reference pattern. In FIG. 19, the concave / convex periodic structure 21 included in the reference pattern includes three types of concave / convex periodic structures 21A, 21B, and 21C, and the width of the convex portion 22 in the three types of concave / convex periodic structures 21A, 21B, and 21C. Are the same. Further, in the uneven periodic structure 21A shown in FIG. 19A, the pitch of the convex portions 22 (concave / convex cycle) is P1, and the concave / convex periodic structure 21B shown in FIG. ) Is P2, and in the concave-convex periodic structure 21C shown in FIG. 19C, the pitch (protrusion period) of the convex portions 22 is P3, and P1 <P2 <P3. Accordingly, the width of the concave portion 23A of the concave / convex periodic structure 21A shown in FIG. 19A, the width of the concave portion 23B of the concave / convex periodic structure 21B shown in FIG. 19B, and the concave / convex periodic structure 21C shown in FIG. The width of the recess 23C is in a relationship of 23A <23B <23C.

Each concave-convex periodic structure 21 is assumed to have concave portions 23 having the same width with at least one convex portion 22 interposed therebetween. That is, in the concave-convex periodic structure 21A shown in FIG. 19A, the concave portions 23A and 23A having the same width are located with the convex portion 22 in between. Similarly, in the concave-convex periodic structure 21B shown in FIG. 19B, concave portions 23B, 23B having the same width are located across the convex portion 22, and in the concave-convex periodic structure 21C shown in FIG. Concave portions 23C and 23C having the same width are located across the convex portion 22. This is because the shadowing effect in the adjacent recesses 23 affects the measurement of the width of the projections 22 by the CD-SEM.
In the description of the concavo-convex periodic structure provided in the reference pattern, three types of concavo-convex periodic structures 21A, 21B, and 21C are cited for convenience. However, the type of the concavo-convex periodic structure 21 constituting the reference pattern, that is, the convex portion 22 The type of pitch affects the accuracy in determining the correction amount of the pattern measurement value by using the reference pattern, and the more types of the uneven periodic structure 21 are more suitable. The types of the concavo-convex periodic structure 21 are, for example, 5 types or more, preferably 10 types or more, more preferably 20 types or more.

Further, among the multiple types of irregular periodic structures 21 constituting the reference pattern, the irregular periodic structure having the smallest pitch of the convex portions 22 has the same width of the opening portion 23c of the concave portion 23 and the width of the base portion 22b of the convex portion 22. It is preferable that the width of the bottom portion 23b of the concave portion 23 and the width of the top portion 22c of the convex portion 22 are the same. In this case, the width of the concave portion 23 in the concave / convex periodic structure is preferably equal to or smaller than the width of the concave portion having the narrowest width among the concave portions constituting the main pattern located in the main pattern region 14. This makes it possible to create a correction curve in the pattern measurement method described later, based on a plot in a fine recess to be measured, rather than extrapolation based on a plot with a large recess width, and the accuracy of the correction curve can be improved. It will be expensive. Note that the measurement of the width of the convex portion by CD-SEM depends on the inclination angle of the side wall portion, and it is difficult to define the width of the base portion of the convex portion when a round exists in the side wall portion. Among the plurality of types of irregular periodic structures 21 constituting the pattern, the irregular periodic structure having the smallest pitch of the convex portions 22 has the same width of the opening portion 23c of the concave portion 23 and the width of the top portion 22c of the convex portion 22. Also good.
On the other hand, the width of the concave portion 23 in the concave / convex periodic structure having the widest pitch of the convex portions 22 among the multiple types of concave / convex periodic structures 21 constituting the reference pattern is the amount of correction of the pattern measurement value by using the reference pattern. In order to improve accuracy in determination, the thickness is preferably set to 200 nm or more, and can be set as appropriate within a range of 200 to 500 nm, for example.

The reference pattern may have a plurality of types of inverted concavo-convex periodic structures in which the concavo-convex structure of the concavo-convex periodic structure 21 is inverted together with the above-described plurality of types of concavo-convex periodic structures 21. In a plurality of types of inverted concavo-convex periodic structures, the widths of the recesses are the same, but the pitches of the recesses are different. Such an inverted concavo-convex periodic structure of the reference pattern is obtained by imprinting using the imprint mold 11 in the reference pattern transferred and formed on the resin material to be formed together with the main pattern located in the main pattern region 14. This is a seed irregular structure 21. Therefore, even in the pattern formed using the imprint mold 11, the pattern measurement value using the reference pattern can be corrected.
In the imprint mold of the present invention, the substrate may have a mesa structure as described above. In this case, a reference pattern region is set on the substrate around the convex structure portion. Also good. FIG. 20 is a plan view showing an example of such an imprint mold. In FIG. 20, an imprint mold 31 includes a base material 32 having a convex structure portion 32 ′ on one surface 32a, an uneven structure forming region 33 set on the surface of the convex structure portion 32 ′, and the uneven structure. A main pattern (not shown) located in the main pattern area 34 set in the formation area 33 and a reference pattern (not shown) located in the reference pattern area 35 set in the base material 32 around the convex structure portion 32 ′. Z)). In such an imprint mold 31 as well, the reference pattern is preferably formed under the same conditions as the main pattern.

  Such imprint molds 11 and 31 can perform high-precision pattern measurement by a pattern measurement method described later. In addition, as described above, when the reference pattern has an inverted concavo-convex periodic structure in which the concavo-convex structure is reversed together with the concavo-convex periodic structure, high-accuracy pattern measurement is possible even with a pattern formed by imprinting by the pattern measurement method described later. It is.

[Pattern measurement method]
Next, the pattern measurement method of the present invention will be described using the above-described imprint mold 11 as an example.
In the pattern measurement method of the present invention, in the reference pattern measurement step, the width of the convex portion 22 of each concavo-convex periodic structure 21 is measured with a CD-SEM for the reference pattern of the imprint mold 11. In the measurement of the convex portion 22 by the CD-SEM, based on the contrast distribution of the electron beam obtained by the measurement, the distance of the portion where the contrast distribution has the maximum inclination can be set as the width of the convex portion. Further, the distance between the peaks of the contrast distribution may be set as the width of the convex portion, and further, the distance at a location that is 60% of the peak value of the contrast distribution may be set as the width of the convex portion.

Next, in the correction curve creating step, first, one type of uneven periodic structure is set as a standard uneven periodic structure from the plurality of types of uneven periodic structures 21 constituting the reference pattern. In general, it is preferable to set the standard uneven periodic structure by selecting an uneven periodic structure having the largest pitch of the convex portions 22 among a plurality of types of uneven periodic structures 21 constituting the reference pattern. Thereby, the accuracy of the correction curve created as described later becomes higher. However, the present invention is not limited to making the concave-convex periodic structure having the largest pitch of the convex portions 22 the reference concave-convex periodic structure.
Next, with respect to the measurement values by the CD-SEM of the convex portions 22 of the other concave / convex periodic structures 21 excluding the reference concave / convex periodic structure, an error from the measured values by the CD-SEM of the convex portions 22 of the reference concave / convex periodic structure is calculated. Further, in each irregular periodic structure 21 of the reference pattern, a value obtained by subtracting the measured value of the convex part 22 from the pitch of the convex part 22 is calculated as the concave part width of the irregular periodic structure. The calculation of the error and the calculation of the recess width can be performed as described in the pattern measurement correction amount determination method described above.

Next, a correction curve indicating an error corresponding to the recess width is created from the relationship between the recess width and the error in each concave and convex periodic structure 21 of the reference pattern. The creation of the correction curve can be performed as described in the method for determining the correction amount of the pattern measurement described above.
On the other hand, in the pattern measurement process, the main pattern, which is the pattern to be measured, is measured with a CD-SEM. The conditions for measuring the main pattern by the CD-SEM are the same as the conditions for measuring the width of the convex portion 22 of each concave-convex periodic structure 21 constituting the reference pattern by the CD-SEM.
Then, in the correction step, the measurement value of the concave portion of the main pattern, which is the pattern to be measured, is applied to the correction curve created as described above, the corresponding error is specified as a correction amount, and this correction amount is used as the correction amount of the concave portion. Subtract from the measured value to obtain the recess width. Thereby, the measurement result approximated to the width of the bottom portion of the concave portion is obtained.
In the pattern measurement method of the present invention, in the correction step, the specified correction amount can be added to the measurement value of the convex portion adjacent to the concave portion to obtain the convex portion width. As a result, a measurement result approximate to the width of the pattern at the same depth, for example, the width of the base of the convex portion is obtained.

Further, as described above, when the reference pattern has a plurality of types of inverted concavo-convex periodic structure 21 in which the concavo-convex structure of the concavo-convex periodic structure 21 is inverted together with the plurality of types of concavo-convex periodic structure 21, the imprint mold 11 is used. Even in the formed pattern, the pattern measurement value using the reference pattern can be corrected. Therefore, the accuracy of the imprint mold and the pattern produced using the imprint mold can be reliably confirmed.
The above-described embodiments of the present invention are examples, and the present invention is not limited to these embodiments.

  It is useful for measuring a pattern having a fine dimension, for example, a fine dimension of several tens of nm or less, and useful for various applications in which a pattern is formed by imprinting using an imprint mold.

DESCRIPTION OF SYMBOLS 1 ... Pattern 2 ... Convex part 2b ... Base part of convex part 2c ... Top part of convex part 3 ... Recessed part 3a ... Side wall part of recessed part 3b ... Bottom part of recessed part 3c ... Opening part of recessed part 11, 31 ... Mold for imprint 12, 32 ... Base material 32 '... Convex structure part 14, 34 ... Main pattern area 15, 35 ... Reference pattern area 21, 21A, 21B, 21C ... Concave and convex periodic structure 22 ... Convex part 22b ... Base part of convex part 22c ... Convex part Top part 23A, 23B, 23C ... Concave part 23a ... Side wall part of concave part 23b ... Bottom part of concave part 23c ... Opening part of concave part

Claims (19)

A reference pattern having a plurality of concave and convex periodic structures in which convex portions and concave portions are alternately arranged, the reference patterns having the same convex portion width but different convex portion pitches among the various concave and convex periodic structures. , Formed in advance under the same conditions as the pattern to be measured,
While measuring the convex part width of each concavo-convex periodic structure of the reference pattern with a scanning electron microscope, setting one type of concavo-convex periodic structure from a plurality of types of concavo-convex periodic structures as a reference concavo-convex periodic structure,
For the measurement value by the scanning electron microscope of the convex portion of the other concave and convex periodic structure excluding the reference concave and convex periodic structure, calculate the error from the measurement value by the scanning electron microscope of the convex portion of the reference concave and convex periodic structure,
In each concave-convex periodic structure of the reference pattern, a value obtained by subtracting the measured value of the convex portion from the pitch of the convex portion is calculated as the concave portion width of the concave-convex periodic structure,
From the relationship between the concave portion width and the error amount in each concave and convex periodic structure of the reference pattern, create a correction curve indicating an error amount corresponding to the concave portion width,
A correction amount determination method characterized by applying a measured value of a concave portion of the measurement target pattern measured with a scanning electron microscope to the correction curve, and specifying a corresponding error amount as a correction amount.   2. The correction amount determination method according to claim 1, wherein each of the concave and convex periodic structures constituting the reference pattern has concave portions having the same width across at least one convex portion.   The concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the base portion of the convex portion. 3. The correction amount determination method according to claim 1, wherein the width and the width of the top of the convex portion are the same.   The concavo-convex periodic structure having the smallest convex pitch among the plurality of types of concavo-convex periodic structures constituting the reference pattern is characterized in that the width of the opening portion of the concave portion and the width of the top portion of the convex portion are the same. The correction amount determination method according to claim 1 or 2. A reference pattern formed under the same conditions as the pattern to be measured, which has a plurality of irregular periodic structures in which convex portions and concave portions are alternately arranged, and the width of the convex portion is the same among various concave and convex periodic structures. For a reference pattern having a different pitch of the convex portion, a reference pattern measuring step of measuring the convex portion width of each concave and convex periodic structure with a scanning electron microscope,
One kind of uneven periodic structure is set as a reference uneven periodic structure from a plurality of kinds of uneven periodic structures, and the reference value is measured with a scanning electron microscope on the convex portions of other uneven periodic structures excluding the reference uneven periodic structure. Calculate the error from the measurement value of the convex portion of the concave / convex periodic structure by the scanning electron microscope, and in each concave / convex periodic structure of the reference pattern, the value obtained by subtracting the measured value of the convex portion from the pitch of the convex portion Calculating as a concave width of the periodic structure, and a correction curve creating step for creating a correction curve indicating an error corresponding to the concave width from the relationship between the concave width and the error in each concave and convex periodic structure of the reference pattern;
A pattern measurement process for measuring the pattern to be measured with a scanning electron microscope;
A step of applying a measurement value of a concave portion of the pattern to be measured to the correction curve, specifying a corresponding error amount as a correction amount, and subtracting the correction amount from the measurement value of the concave portion to obtain a concave portion width. A pattern measurement method characterized by this.   6. The pattern measurement method according to claim 5, wherein, in the correction step, the correction amount is added to a measurement value of a convex portion adjacent to the concave portion to obtain a convex portion width.   7. The pattern measurement method according to claim 5, wherein each concave-convex periodic structure constituting the reference pattern has concave portions having the same width with at least one convex portion interposed therebetween.   The concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the base portion of the convex portion. The pattern measurement method according to claim 5, wherein the width and the width of the top of the convex portion are the same.   A base pattern, a main pattern located in a main pattern area set on the base material, and a reference pattern located in a reference pattern area set on the base material, the reference pattern including a convex portion For imprints, characterized by having a plurality of concave and convex periodic structures in which concave portions are alternately arranged, and the convex portions have the same width but different convex pitches among the various concave and convex periodic structures. mold.   The base material is provided with a convex structure portion having a flat surface formed through a periphery and a step on one surface, and the main pattern region and the reference pattern are set on the surface of the convex structure portion. The imprint mold according to claim 9, wherein the mold is an imprint mold.   The imprint mold according to claim 9 or 10, wherein the main pattern region and the reference pattern region are located apart from each other.   11. The imprint mold according to claim 9, wherein the reference pattern region is located in the main pattern region.   The base material has a convex structure part having a flat surface formed through a periphery and a step on one surface, the main pattern region is set on the surface of the convex structure part, and the reference pattern is the convex structure. The imprint mold according to claim 9, wherein the mold is set on the base material around a portion.   The imprint mold according to any one of claims 9 to 13, wherein each of the concave and convex periodic structures constituting the reference pattern has concave portions having the same width with at least one convex portion interposed therebetween. .   The concave-convex periodic structure having the smallest convex pitch among the multiple types of concave-convex periodic structures constituting the reference pattern has the same width of the opening portion of the concave portion and the width of the base portion of the convex portion. The mold for imprinting according to any one of claims 9 to 14, wherein the width and the width of the top of the convex portion are the same.   The concavo-convex periodic structure having the smallest convex pitch among the plurality of types of concavo-convex periodic structures constituting the reference pattern is characterized in that the width of the opening portion of the concave portion and the width of the top portion of the convex portion are the same. Item 15. The imprint mold according to any one of items 9 to 14.   The width of the concave portion in the concave / convex periodic structure having the smallest convex pitch among the plural types of concave / convex periodic structures constituting the reference pattern is equal to or smaller than the narrowest concave portion among the concave portions constituting the main pattern. The imprint mold according to any one of claims 9 to 16, characterized in that   The width of the concave portion in the concave / convex periodic structure having the widest convex pitch among the plural types of concave / convex periodic structures constituting the reference pattern is 200 nm or more. The mold for imprint in any one.   The reference pattern has an inverted concavo-convex periodic structure in which the concavo-convex structure of the plurality of concavo-convex periodic structures is inverted, and in the inverted concavo-convex periodic structures, the width of the recesses is the same but the pitch of the recesses is different. The imprint mold according to any one of claims 9 to 18, wherein the mold is an imprint mold.

Images