JP5254099B2 - Pattern dimension measuring method and charged particle beam writing method - Google Patents

Description

  The present invention relates to a pattern dimension measuring method and a charged particle beam writing method.

  With the high integration of semiconductor devices, pattern miniaturization is progressing rapidly. In order to form a fine pattern, it is necessary to form a resist pattern according to the design dimensions. Therefore, the dimension between the resist patterns after the development processing is measured, and the measured dimension is compared with the design dimension. In addition, the dimensional accuracy between resist patterns is improved by feeding back dimension measurement values to an electron beam drawing process, a heat treatment process after drawing, a developing process, and the like. For example, in the drawing process, the irradiation amount of the electron beam is calculated based on the dimension measurement value between the resist patterns fed back. Here, a length measuring scanning electron microscope (also referred to as a length measuring SEM or a CD-SEM, hereinafter referred to as a “length measuring SEM”) is generally used to measure the dimension between resist patterns.

  In the length measurement SEM, the dimension in plan view between the resist patterns is measured. At that time, the dimension at the height at which the resist pattern is in contact with the chromium film as the base film is measured. For this reason, when a taper part exists under a resist pattern, there exists a problem that a measurement error generate | occur | produces depending on the taper part. If the dimension measurement value including such a measurement error is fed back to the drawing process as described above, the irradiation amount of the electron beam cannot be calculated with high accuracy, and there is a possibility that the drawing accuracy cannot be improved. In order to remove the measurement error of the length measurement SEM, SEM simulated waveforms for various pattern shapes are stored by electron beam simulation, the SEM simulated waveform closest to the measured length measurement SEM image is selected, and the pattern dimensions are determined. A measuring method is disclosed (for example, refer to Patent Document 1).

  However, the method described in Patent Document 1 is complicated, and it has been desired to develop a method for easily and accurately measuring the dimension between resist patterns.

JP 2007-218711 A

  In view of the above problems, an object of the present invention is to provide a pattern dimension measuring method and a charged particle beam writing method capable of easily and accurately measuring the dimension between adjacent resist patterns.

  In order to solve the above-mentioned problems, a first aspect of the present invention is a pattern dimension measuring method for measuring a dimension between two adjacent resist patterns, and two resist patterns measured using a length-measuring scanning electron microscope. The correlation between the dimension at the first height between the two and the dimension at the second height between the two resist patterns obtained from the cross-sectional image is acquired in advance, and two resist patterns to be measured are obtained. A dimension at a first height is measured with a length-measuring scanning electron microscope, and the measured dimension is calibrated to a dimension at a second height based on the correlation.

  In the first aspect, the resist pattern is formed on the base film, the first height is the height of the lower surface of the resist pattern in contact with the base film, and the second height is a taper portion of the resist pattern. It is preferable that the height does not exist.

  In the first aspect, the design dimensions between the resist patterns are changed to obtain the dimensions at the first and second heights, respectively, and the discrete values of the dimensions at the first and second heights are interpolated. It is preferable to acquire the correlation.

  In the first aspect, the resist pattern is preferably formed by performing development processing after irradiation with a charged particle beam.

  According to a second aspect of the present invention, there is provided a charged particle beam writing method for irradiating a specimen with a charged particle beam, wherein the resist on the specimen is applied to the resist on the basis of the dimension obtained by using the pattern dimension measuring method according to the first aspect. In this case, the irradiation amount of the charged particle beam irradiated on the surface is calculated.

  According to the present invention, the second height between the two resist patterns obtained from the dimension and the cross-sectional image at the first height between the two resist patterns measured using the length-measuring scanning electron microscope. The correlation at the first height is obtained in advance, and the dimension at the first height between the resist patterns measured by the length-measuring scanning electron microscope is changed to the dimension at the second height based on this correlation. Since calibration is performed, the dimension between two adjacent resist patterns can be measured easily and accurately.

It is the schematic which shows the structure of the electron beam drawing apparatus in this Embodiment. (A) is sectional drawing which shows the design dimension CDa between resist patterns, (b) is calculated | required from dimension CDs in 1st height h1 between the resist patterns measured by length measurement SEM, and a cross-sectional image. FIG. 10 is a cross-sectional view showing a dimension CDd at a second height h2 between resist patterns. It is a figure which shows the calibration table by which the correlation of two dimensions CDs and CDd was prescribed | regulated when the design dimension CDa was changed. It is a figure which shows correlation with dimension CDs and dimension CDd. It is a figure which shows the table for a calibration by which the correlation of two dimensions CDs and CDd when changing pattern density (rho) was prescribed | regulated. It is a figure which shows the relationship between pattern density (rho) and the difference of the dimension CDd and the dimension CDs.

  FIG. 1 is a schematic diagram showing a configuration of an electron beam drawing apparatus in the present embodiment. In FIG. 1, a stage 3 on which a mask 2 as a sample is placed is accommodated in a sample chamber 1. The stage 3 is driven by the stage drive circuit 4 in the x direction (direction parallel to the paper surface) and the y direction (direction perpendicular to the paper surface). The moving position of the stage 3 is measured by a position circuit 5 using a laser side gauge or the like.

  An electron beam optical system 10 is disposed on the sample chamber 1. The electron beam optical system 10 includes an electron gun 6, various lenses 7, 8, 9, 11, 12, a blanking deflector 13, a beam size variable deflector 14, and a beam scanning main deflector 15. And a sub-deflector 16 for beam scanning and apertures 17 and 18 for beam shaping. The main deflector 15 positions the electron beam in a predetermined subfield (subdeflection area). On the other hand, the sub deflector 16 performs positioning of a figure drawing unit in the subfield. The beam size changing deflector 14 and the beam shaping apertures 17 and 18 serve to control the beam shape. Further, the blanking deflector 13 serves to control the irradiation of the electron beam on the surface of the mask 2.

  In the drawing process using an electron beam, first, the stage 3 is continuously moved in the x direction, and a drawing process is performed on a frame area in which the drawing area is divided into strips according to the deflection width of the main deflection beam. Next, the stage 3 is stepped in the y direction orthogonal to the x direction, and the drawing process is performed in the same manner. By repeating this, each frame region is sequentially drawn.

  In FIG. 1, a magnetic disk 21 is connected to the control computer 20. Here, the magnetic disk 21 stores drawing data of a semiconductor device (LSI). The drawing data read from the magnetic disk 21 is temporarily stored in the pattern memory 22 for each frame area. Pattern data for each frame area stored in the pattern memory 22, that is, frame information including drawing positions, graphic data, and the like is analyzed by a pattern data decoder 23 and a drawing data decoder 24, which are data analysis units, and then blanked. The signal is sent to the amplifier 25, the beam shaping amplifier 26, the main deflection amplifier 27 and the sub deflection amplifier 28.

  In the pattern data decoder 23, drawing data is input, and if necessary, the graphic data included in the frame area is inverted to generate inverted pattern data. Next, the graphic data defined as frame data is divided into unit drawing graphic groups that can be formed by the combination of the beam shaping apertures 17 and 18. Then, the irradiation amount is calculated based on the data obtained by the figure division, and blanking data is created according to the irradiation amount and sent to the blanking amplifier 25. Further, desired beam size data is created and sent to the beam shaping amplifier 26. Next, a predetermined deflection signal is sent from the beam shaping amplifier 26 to the beam size varying deflector 14 of the electron beam optical system 10, thereby controlling the size of the electron beam.

  The drawing data decoder 24 creates subfield positioning data based on the frame data. The obtained data is sent to the main deflection amplifier 27. Then, a predetermined signal is sent from the main deflection amplifier 27 to the main deflector 15 of the electron beam optical system 10, and the electron beam is deflected and scanned at the designated subfield position. In the drawing data decoder 24, a control signal for scanning the sub deflector is generated and sent to the sub deflection amplifier 28. Next, a predetermined sub-deflection signal is sent from the sub-deflection amplifier 28 to the sub-deflector 16, thereby drawing for each sub-field.

  In FIG. 1, the sub deflector 16 is used to control the position of the electron beam at high speed and with high accuracy. For this reason, the deflection range of the sub deflector 16 is limited to the subfield on the mask 2. When the deflection exceeds this range, it is necessary to move the position of the subfield by the main deflector 15. In this way, the main deflector 15 is used to control the position of the subfield, and can move the subfield within the frame (main deflection area). Here, the frame is an area that can be deflected by the main deflector 15. During drawing, since the stage 3 is continuously moving in one direction, the drawing origin of the subfield is tracked by the main deflector 15 so that the drawing origin follows the movement of the stage 3.

  Incidentally, conventionally, the dimension CDs between two adjacent resist patterns drawn by the electron beam drawing apparatus is measured using an external length measuring SEM. Then, in order to improve the drawing accuracy, the dimension measurement value CDs is input to the pattern data decoder 23 via the control computer 20, and the electron beam irradiation amount is determined by the pattern data decoder 23 based on the dimension measurement value CDs. The blanking data was created based on the calculated dose.

  However, the dimension CDs measured by the length measurement SEM is a dimension in a plan view between the resist patterns. Specifically, as shown in FIG. 2B, the dimension CDs at the height h1 of the lower surface of the resist pattern 102 in contact with the Cr film or the MoSi film 100 as the light shielding film (underlying film) is measured. When the resist pattern 102 has a tapered portion (also referred to as a “bottom portion”) 102a at the height h1, the dimension CDs between the tapered portions 102a is measured. The dimension CDs is not a dimension corresponding to the design dimension CDa shown in FIG. 2A, that is, the dimension CDd at the height h2 other than the tapered portion 102a between the resist patterns 102. For this reason, it is not possible to accurately calculate the irradiation amount of the electron beam by simply feeding back the dimension CDs measured by the length measurement SEM to the electron beam drawing apparatus, and the drawing accuracy cannot be improved. . The Cr film 100 is formed on an unillustrated base substrate (transparent glass substrate).

  Therefore, in the present embodiment, as described below, the planar dimension CDs at the height h1 between the two resist patterns measured by the length measurement SEM and the distance between the two resist patterns obtained from the cross-sectional image. The correlation with the cross-sectional dimension CDd at the height h2 is obtained in advance as a calibration table, and the dimension CDm at the height h1 actually measured by the length measuring SEM is used as the height h2 using the calibration table. Is calibrated to the dimension CDc at (5), and the irradiation amount of the electron beam is calculated based on the calibrated dimension CDc.

  FIG. 3 is a diagram showing a calibration table used for calibrating the dimension measured by the length measurement SEM in the present embodiment. A method for creating the calibration table will be described. First, the design dimension CDa between the resist patterns 102 shown in FIG. 2A is changed to 90 nm, 128 nm, and 180 nm, and the resist pattern 102 corresponding to each design dimension CDa is subjected to electron beam lithography and heat treatment ( PEB) and development processing. These resist patterns 102 may be formed at different positions on the same mask, or may be formed at corresponding positions on different masks.

  And about each resist pattern 102, dimension CDs in planar view between the resist patterns 102 is measured by length measurement SEM. That is, as shown in FIG. 2B, the dimension CDs between the tapered portions 102a of the two resist patterns 102 at the height h1 is measured. Further, a cross-sectional image as shown in FIG. 2B is acquired for these two resist patterns 102, and a dimension CDd between the resist patterns 102 at the height h2 is obtained from the cross-sectional images. That is, the dimension CDd between the side walls of the resist pattern 102 without the tapered portion 102a is obtained. In addition, a cross-sectional image can be acquired with a cross-sectional scanning electron microscope (cross-section SEM). And about each design dimension 90nm, 128nm, 180nm, the planar view dimension (CDs) in height h1 measured by length measurement SEM = 91nm, 127nm, 178nm, and the dimension in height h2 calculated | required from the cross-sectional image By associating (CDd) = 105 nm, 148 nm, and 208 nm, a calibration table as shown in FIG. 3 is obtained. The correlation between the two dimensions CDs and CDd defined in the calibration table shown in FIG. 3 can be plotted as shown in FIG. Interpolation processing may be performed between plots. That is, the correlation between the planar view size at the height h1 and the cross-sectional view size at the height h2 is to obtain the dimensions CDs and CDd by changing the design dimension CDa between two adjacent resist patterns. The discrete values of CDs and CDd can be obtained by performing interpolation processing.

  The obtained calibration table is stored in the control computer 20. The control computer 20 receives a planar view dimension CDm at the height h1 between resist patterns to be measured, which is actually measured by an external length measuring SEM. In the control computer 20, the dimension CDc at the height h2 is obtained by calibrating the dimension CDm externally input using the calibration table. The dimension CDc after calibration can be obtained by linear interpolation according to the following equation (1). The dimension CDc after calibration may be obtained by non-linear interpolation instead of linear interpolation.

_{CDc = CDd N + (CDd N} + 1 -CDd N) × (CDm-CDs N)
/ (CDs _{N + 1} -CDs _N ) (1)

For example, when the dimension CDm actually measured by the length measuring SEM is 100 nm, N = 1 in the above equation (1). That is, N = 1 CDs ₁ = 91, CDd ₁ = 105, N = 2 CDs ₂ = 127, and CDd ₂ = 148 defined in the calibration table shown in FIG. 3 are substituted into the above equation (1). By doing so, the dimension CDc = 11.5 nm after calibration is obtained as in the following equation.
CDc = 105 + (148-105) × (100-91) / (127-91)
= 115.8 nm

For example, when the actually measured dimension CDm is 150 nm, N = 2 in the above equation (1). That is, N = 2 CDs ₂ = 127, CDd ₂ = 148, N = 3 CDs ₃ = 178, and CDd ₃ = 208 defined in the calibration table shown in FIG. 3 are substituted into the above equation (1). By doing so, the dimension CDc = 175.1 nm after calibration is obtained as follows.
CDc = 148 + (208-148) × (150-127) / (178-127)
= 175.1 nm

  Thus, the dimension CDc calibrated in the control computer 20 is input to the pattern data decoder 23. Then, the pattern data decoder 23 calculates a dimensional correction amount based on the data obtained by analyzing the frame data and the calibrated dimension CDc, and calculates the electron beam irradiation amount based on the dimensional correction amount. Then, blanking data is created according to the irradiation amount.

  As described above, in the present embodiment, the resist pattern obtained from the planar view dimension CDs at the height h1 between the two resist patterns measured by the length measurement SEM and the cross-sectional images of these two resist patterns. The correlation with the sectional view dimension CDd at the height h2 is obtained in advance as a calibration table. Then, the dimension CDm at the height h1 between the resist patterns actually measured by the length measurement SEM is calibrated to the dimension CDc at the height h2 based on the correlation defined in the calibration table. In the present embodiment, a complicated electron beam simulation as described in Patent Document 1 is not used in order to remove measurement errors due to the length measurement SEM. Therefore, according to the present embodiment, the dimension CDc between the resist patterns 102 can be measured easily and accurately. Furthermore, since the electron beam irradiation amount is calculated based on the dimension CDc after calibration, the irradiation amount is calculated with high accuracy. Thereby, the drawing accuracy can be improved and the demand for pattern miniaturization can be met.

  The present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the spirit of the present invention. For example, although the electron beam is used in the above embodiment, the present invention is not limited to this, and the present invention can also be applied to the case where another charged particle beam such as an ion beam is used.

  Further, the calibration table used for calibrating the dimension between resist patterns is not limited to the one obtained by changing the design dimension CDa, but the pattern type (line pattern, hole pattern), pattern density ρ, You may obtain | require by changing drawing conditions (an irradiation amount, a focus), etc. Here, when the material (resist material, base film material, base substrate material) changes, process conditions such as heat treatment (PEB) and development processing performed after drawing also change, and the values in the calibration table also vary greatly. Therefore, while fixing the material (and process conditions) and changing the design dimension CDa, the pattern type, the pattern density ρ, and the drawing conditions, which are parameters other than the material, the planar dimension CDs and the height at the height h1 are changed. The cross sectional view dimension CDd at the height h2 is obtained, the correlation between these two dimensions CDs and CDd is obtained, and this is defined in each calibration table.

  FIG. 5 is a diagram illustrating an example of a calibration table that defines the correlation between two dimensions CDs and CDd when the pattern density ρ is changed. A method for creating the calibration table shown in FIG. 5 will be described. First, the resist pattern 102 is formed by changing the pattern density ρ of the resist pattern 102 in a predetermined region to 5%, 50%, and 95%. Then, for each resist pattern 102, a planar view dimension CDs at a height h1 between the resist patterns 102 is measured by a length measurement SEM. Here, the dose is determined so that the dimension CDs measured by the length measurement SEM is constant (for example, 180 nm) even if the pattern density ρ changes. Further, a cross-sectional image is acquired for each resist pattern 102, and a dimension CDd at the height h2 is obtained from the cross-sectional image. Then, for each pattern density of 5%, 50%, and 95%, dimensions (CDs) measured by a length measurement SEM = 180 nm, 180 nm, and 180 nm, and dimensions (CDd) obtained from a cross-sectional image = 189 nm, 185 nm, and 181 nm Can be obtained, a calibration table as shown in FIG. 5 can be obtained. A calibration table as shown in FIG. 5 is prepared for each planar dimension CDs at the height h1 measured by the length measurement SEM. The relationship between the pattern density ρ defined in the calibration table shown in FIG. 5 and the difference between CDd and CDs can be plotted and linearly approximated as shown in FIG.

A method for calibrating the dimension CDm between resist patterns will be described using the calibration table shown in FIG. When the dimension CDm in plan view actually measured by the length measurement SEM is 180 nm, the dimension CDc after calibration corresponding to the pattern density ρ can be obtained according to the following equation (2).
CDc = 180−0.09 × ρ + 9.44 = −0.09ρ + 189.44 (2)

  In the above-described embodiment, the calibrated dimension CDc is used when calculating the electron beam irradiation amount in the electron beam lithography apparatus. However, the heat treatment (PEB) or development processing performed after the electron beam lithography is used. The dimension CDc after calibration may be used when determining the processing conditions of the process.

20 control computer 23 pattern data decoder 25 blanking amplifier 100 light shielding film 102 resist pattern 102a taper portion

Claims (5)

In a pattern dimension measuring method for measuring a dimension between two adjacent resist patterns,
A dimension at a first height between two resist patterns measured using a length-measuring scanning electron microscope, and a dimension at a second height between the two resist patterns determined from a cross-sectional image; , The dimension at the first height between the two resist patterns to be measured is measured with the length-measuring scanning electron microscope, and the measured dimension is calculated based on the correlation. 2. A pattern dimension measuring method, characterized by calibrating to a dimension at a height of 2.   The resist pattern is formed on a base film, the first height is a height of a lower surface of the resist pattern in contact with the base film, and the second height is a height at which no taper portion of the resist pattern exists. The pattern dimension measuring method according to claim 1, wherein:   The correlation between the resist patterns is changed to obtain the dimensions at the first and second heights, and the correlation between the discrete values of the dimensions at the first and second heights is interpolated. The pattern dimension measuring method according to claim 1 or 2, wherein:   4. The pattern dimension measuring method according to claim 1, wherein the resist pattern is formed by performing development processing after irradiating a charged particle beam. 5. In a charged particle beam writing method for irradiating a specimen with a charged particle beam,
The dose of the charged particle beam irradiated to the resist on the sample is calculated based on the dimension obtained by using the pattern dimension measuring method according to any one of claims 1 to 4. The charged particle beam drawing method characterized by the above-mentioned.

Images