JP2010225811A - Charged particle beam lithography method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a charged particle beam lithography method capable of reducing dimensional variation by improving accuracy of a correction process. <P>SOLUTION: A plurality of patterns different in area density are drawn by changing a proximity effect correction coefficient for every reference emission amount of a charged particle beam, and dimensions of the patterns after the drawing are measured. Then, an optimum proximity effect correction coefficient and deviations of the patterns from design dimensions when drawn by using the proximity effect correction coefficient are calculated for every reference emission amount. Discrete values of the calculated proximity effect correction coefficients and deviations from the design dimensions are fitted using an expression: η=η<SB>0</SB>/erfc(ΔCD/2/σ<SB>F</SB>)/2, wherein: erfc, η and ΔCD are an error function, a proximity effect correction coefficient, and a deviation of a dimension from a design dimension, respectively; and η<SB>0</SB>and σ<SB>F</SB>are each a fitting parameter, and represent a standard proximity effect correction coefficient and an extent of the influence of forward scattering, respectively. The proximity effect correction coefficient and the deviation from the design dimension are interpolated and/or extrapolated using the fitting result. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a charged particle beam writing method.

  In recent years, with the increase in the degree of integration of semiconductor devices, the dimensions of individual elements have been reduced, and the widths of wirings and gates constituting each element have also been reduced.

  The photolithography technology that supports this miniaturization includes a step of applying a resist composition to the surface of a substrate to be processed or processed to form a resist film, and exposing a predetermined resist pattern by irradiating light. A step of forming a resist pattern latent image, a step of heat treatment as necessary, a step of developing this to form a desired fine pattern, and a process such as etching on the substrate using this fine pattern as a mask The process to perform is included.

  In the photolithography technology, the wavelength of exposure light is proportional to the width of a resolvable wiring pattern or the like. Accordingly, as one of means for reducing the pattern size, the wavelength of exposure light used for forming the resist pattern latent image is being shortened.

  In addition, development of an electron beam lithography technique is being promoted as a higher resolution exposure technique. The electron beam lithography technique has an essentially excellent resolution because the electron beam used is a charged particle beam. Further, since a large depth of focus can be secured, there is an advantage that dimensional variation can be suppressed even on a high level difference. For this reason, in addition to being applied to the development of state-of-the-art devices typified by DRAM, some are also used in the production of ASICs. Furthermore, electron beam lithography technology is widely used in the field of manufacturing masks or reticles that serve as masters for transferring LSI patterns onto wafers.

  Patent Document 1 discloses a variable shaping type electron beam drawing apparatus used in an electron beam lithography technique. The drawing data in such an apparatus is created by performing processing such as correction and graphic pattern division on design data (CAD data) such as a semiconductor integrated circuit designed using a CAD system.

  For example, the graphic pattern division processing is performed in units of the maximum shot size defined by the size of the electron beam, and the coordinate position, size, and irradiation time of each divided shot are set. Then, drawing data is created so that a shot is formed according to the shape and size of the graphic pattern to be drawn. The drawing data is divided into strip-shaped frames (main deflection areas) and further divided into sub-deflection areas. That is, the drawing data of the entire chip has a data hierarchical structure including a plurality of strip-shaped frame data according to the size of the main deflection area and a plurality of sub deflection area units smaller than the main deflection area in the frame. .

  The sub-deflection area is an area where an electron beam is scanned by the sub-deflector at a speed higher than that of the main deflection area, and is generally a minimum drawing unit. When drawing in the sub-deflection area, a shot having a size and shape prepared according to the pattern figure is formed by the shaping deflector. Specifically, after the electron beam emitted from the electron gun is shaped into a rectangular shape by the first aperture, it is projected onto the second aperture by the shaping deflector to change the beam shape and dimensions. . Thereafter, as described above, the light is deflected by the sub-deflector and the main deflector and is irradiated onto the mask placed on the stage.

In the electron beam drawing apparatus, a correction process for changing the beam irradiation amount is required so that the pattern dimension becomes the same as the dimension of the design data. This process is performed for factors that cause dimensional variations of the resist pattern, such as proximity effect, fogging effect, and loading effect. Here, the proximity effect refers to a phenomenon in which electrons irradiated on the resist film are reflected inside the glass substrate and re-irradiated on the resist film. On the other hand, the fogging effect is a phenomenon in which electrons irradiated on the resist film are reflected on the surface thereof and further reflected on the optical components of the electron beam drawing apparatus, and then re-irradiated over a wide range. This phenomenon is also caused by secondary electrons generated by irradiating the resist film with electrons. The loading effect refers to a dimensional variation caused by a difference in the area of the resist film or the light shielding film in the surface when the lower light shielding film or the like is etched using the resist pattern as a mask. While the influence radius σ of the proximity effect is about 10 and several μm, the influence radius σ _F of the fogging effect is about 10 mm, and the influence radius σ _L of the loading effect ranges from 10 mm to several tens mm.

Japanese Patent Application Laid-Open No. 2004-228561 discloses a method for obtaining an irradiation amount by simultaneously correcting dimensional variations due to a proximity effect, a fogging effect, and a loading effect. In this method, it is necessary to obtain the relationship among the proximity effect correction coefficient η, the reference dose D _base, and the pattern dimension CD. Specifically, a plurality of line patterns having different pattern area densities U are arranged, and patterns are drawn by changing each value of the proximity effect correction coefficient η, the reference dose D _base, and the proximity effect influence range σ _B. Next, the dimension CD of the pattern after drawing is measured, and the proximity effect correction coefficient η when the difference in the dimension CD of each pattern is minimized is obtained. Next, an optimum combination of the proximity effect correction coefficient η and the reference dose D _base satisfying the proximity effect condition is obtained, and in each combination, the influence range in which the difference in the dimension CD of each pattern becomes the smallest is the optimum influence range σ _B. To do. Based on the above results, linear interpolation is performed so that the optimum proximity effect correction coefficient η, the reference dose D _base, and the dimension CD have continuous correlations.

JP-A-9-293670 JP 2007-150243 A

Thus, the method of Patent Document 2, an optimum proximity correction coefficient relation between η and the base dose D _base and dimension CD is optimum set of η and CD at that time the D _base was subjected to drawing experiments Is obtained by interpolation processing. Here, it is preferable to perform as many drawing experiments as possible and obtain as many (D _base , η, CD) pairs as possible. However, in reality, there are limitations on the measurement time and the like, so only the minimum set necessary for the interpolation process is obtained. As a result, there has been a problem that the error when performing the linear interpolation process becomes large, and the deviation from the target correction dimension becomes large.

  In recent years, as the design rules of semiconductor devices are continually miniaturized and highly accurate, the requirements for dimensional accuracy (CD accuracy) for lithography technology are becoming stricter. In particular, the requirements for masks are very strict. Accordingly, an object of the present invention is to provide a charged particle beam drawing method capable of increasing the accuracy of correction processing and reducing dimensional fluctuations in view of these points.

  Other objects and advantages of the present invention will become apparent from the following description.

The first aspect of the present invention is to draw a plurality of patterns with different area densities by changing the proximity effect correction coefficient for each reference irradiation amount of the charged particle beam, and measure the dimension of the pattern after the drawing;
For each reference irradiation amount, calculating an optimal proximity effect correction coefficient and a deviation from the design dimension of the pattern when drawn using this proximity effect correction coefficient;
Fitting the calculated proximity effect correction coefficient and the discrete value of the deviation from the design dimension by the following equation;
η = η ₀ / erfc (ΔCD / 2 / σ _F ) / 2
(Where erfc is the error function, η is the proximity effect correction coefficient, ΔCD is the deviation of the dimension from the design dimension, η ₀ and σ _F are the fitting parameters, and the influence range of the standard proximity effect correction coefficient and the forward scattering respectively. Represents.)
The present invention relates to a charged particle beam drawing method including a step of interpolating and / or extrapolating a proximity effect correction coefficient and a deviation from a design dimension using a fitting result.

  The first aspect of the present invention preferably further includes a step of performing spline interpolation between the calculated proximity effect correction coefficient and a discrete value of deviation from the design dimension.

The second aspect of the present invention is a step of drawing a plurality of patterns having different area densities by changing the proximity effect correction coefficient for each reference irradiation amount of the charged particle beam, and measuring the dimension of the pattern after drawing,
For each proximity effect correction coefficient, calculating a deviation from the design dimension of the pattern when drawn using this proximity effect correction coefficient and a reference irradiation amount,
Fitting the deviation from the calculated design dimension and the discrete value of the reference dose with a polynomial;
The present invention relates to a charged particle beam writing method characterized by having a step of interpolating and / or extrapolating a deviation from a design dimension and a reference irradiation amount using a fitting result.

  The second aspect of the present invention can further include a step of performing spline interpolation on the deviation from the calculated design dimension and the discrete value of the reference dose.

In the first aspect or the second aspect of the present invention, it is preferable that the calculated discrete values are weighted for fitting.
In particular, fitting is performed by weighting the difference between the calculated proximity effect correction coefficient and the design dimension, or the calculated reference dose and deviation from the design dimension, where the deviation from the design dimension is the maximum or minimum. It is preferable to do.

  According to the present invention, it is possible to improve the accuracy of correction processing and reduce dimensional fluctuation.

It is an example of the flowchart which shows the electron beam drawing method of this Embodiment. It is an example of the method of deriving | leading-out S101 of FIG. In this Embodiment, it is an example which showed the relationship between a proximity effect correction coefficient and a dimension for every some pattern from which an area density differs. In this Embodiment, it is an example of the graph which shows the correlation of the proximity effect correction coefficient with respect to a dimension, and reference | standard irradiation amount. It is an example which showed the shift | offset | difference of the pattern dimension at the time of drawing, and a design dimension for every area density. It is an example of the flowchart which shows the electron beam drawing method of this Embodiment. It is a block diagram of the electron beam drawing apparatus in this Embodiment. It is explanatory drawing of the drawing method by the electron beam of this Embodiment. It is a conceptual diagram which shows the irradiation amount calculation method in this Embodiment.

  FIG. 1 is a flowchart showing an electron beam writing method according to the present embodiment.

As shown in FIG. 1, in the electron beam drawing method, for each area density U, the relationship between the dose D _i and the pattern dimension CD _i is obtained by a drawing experiment (S101), and for each reference dose obtained in S101. A step of deriving CD (η, D _base ) by obtaining an optimum proximity effect correction coefficient (S102), a step of fitting the obtained discrete CD (η, D _base ) value (S103), and fitting A step of interpolating and / or extrapolating CD (η, D _base ) from the result (S104), a step of performing spline interpolation of the discrete CD (η, D _base ) obtained in S102 (S105), and S104. from the results of S105 and CD _{(η, D base)} step of defining a (S106), and the step of calculating the dose _{D 50} in area density _{U = 50% (S107),} D base eta is a step (S108) that evaluates to or less than a predetermined value, iso-focul a step (S109) for calculating a _{D base} and eta with _dose, assessing the relationship between _{D base} and eta ( S110), a step of evaluating proximity effect correction error for each pattern difference ΔCD (difference between design dimension and actual drawing result) in a predetermined drawing range (S111), and a series of steps of drawing step (S112) Have

In S101, the relationship between the dose D _i and the pattern dimension CD _i is obtained for each area density U. Specifically, first, a plurality of patterns having different area densities U are drawn while changing the electron beam irradiation amount D _i, and the pattern dimension CD _i after drawing is measured to obtain the relationship with the irradiation amount D _i . Here, using a set of a plurality of D _bases and the proximity effect correction coefficient η, D = D _base D _p (η, U) is changed for rendering. For example, as shown in FIG. 2, a pattern set in which a line pattern having an area density U of about 0%, a line pattern of 50%, and a line pattern of 100% is arranged, and the irradiation amount of the electron beam is arranged. by changing the value of D _i is drawn to the mask. Next, the line width dimension CD _i of each pattern after drawing is measured.

In S102, the optimum η is calculated for each D _base used in the experiment. Specifically, as shown in FIG. 3, the relationship between η and dimension CD in a certain D _base is determined by linear interpolation for each area density, and η that minimizes the dimensional difference between the area densities is set as the optimum value. . In this way, the correlation CD (η, D _base ) between the proximity effect correction coefficient η and the reference dose D _base for the dimension CD is calculated.

In S103, the correlation CD (η, D _base ) obtained in S102 is fitted. That is, a continuous value is obtained by fitting a discrete set of dimensions CD and D _base or η with an appropriate approximate expression. Here, fitting in the present invention is understood as a mathematical optimization method for determining the best fitting parameters or coefficients of a continuous function for discrete values. The term generally encompasses all mathematical methods for curve fitting calculations. The purpose of such curve fitting calculations is to derive the function that best fits the data.

As a result of diligent research, the present inventor has found that the relationship between the dimension CD obtained in S102 and the proximity effect correction coefficient η can be expressed well by selecting the following equation (1) as a fitting function.

η = η ₀ / erfc (ΔCD / 2 / σ _F ) / 2 (1)

Here, erfc represents an error function. Further, η is a proximity effect correction coefficient, and ΔCD is a deviation of the pattern dimension from the design dimension. Furthermore, η ₀ and σ _F are fitting parameters, which respectively represent a standard proximity effect correction coefficient and an influence range of forward scattering.

  For example, a mask in which a chromium film serving as a light shielding layer is formed on a transparent glass substrate such as quartz and a resist FEP-171 (trade name) manufactured by Fuji Film Co., Ltd. is formed as a resist film on the chromium film is prepared. . And the process of S101 and S102 is performed with respect to this resist. The relationship between the obtained dimension CD and the proximity effect correction coefficient η is well expressed by the fitting function of Expression (1).

Further, the present inventor has found that the relationship between the dimension CD and the reference dose D _base can be expressed well by a cubic polynomial. Therefore, in the steps S103 and S104, (a) fitting the calculated proximity effect correction coefficient and the discrete value of the deviation from the design dimension with the equation (1), and (b) deviation from the calculated design dimension. Either or both of fitting a discrete value of the reference dose with a polynomial can be performed.

The fitting function of Expression (1) has only parameters η ₀ and σ, and CD (D _base ) can be fitted with a cubic polynomial. Therefore, if fitting is used when there is an error in the relationship between CD (η) and CD (D _base ), the possibility of interpolating by picking up this error can be reduced. Further, since it is considered that the number of parameters is sufficiently smaller than the number of sets of CD (η) and CD (D _base ), errors included in these sets can be reduced or averaged during the fitting process. Furthermore, when extrapolating CD (η) outside the range drawn in S101, the error can be reduced as compared with a simple extrapolation such as linear. In consideration of the fact that the CD (η, D _base ) obtained in S102 includes an error, using the value obtained by fitting results in several pieces of data near the extrapolation point being 1 It can be said that the reliability is higher than using a value obtained by extrapolation by the following equation or the like.

In S104, CD (η, D _base ) is interpolated and / or extrapolated from the fitting result in S103. In the present embodiment, instead of S103 and S104, the discrete CD (η, D _base ) obtained in S102 may be spline-interpolated to obtain the relationship of CD (η, D _base ). (S105).

Next, CD (η, D _base ) is determined from the result of S104 or S105 (S106). Here, in the present embodiment, the CD (η, D _base ) obtained in S102 is interpolated from the fitting result such as the spline function or Expression (1), and the result of fitting using the function of Expression (1). It is preferable to extrapolate using. When combining interpolation by spline function and extrapolation by fitting, it is preferable to perform fitting with an appropriate weight in order to increase the correlation between the two. For example, fitting is performed by increasing the weight of the data corresponding to the maximum value and the minimum value of the dimension difference ΔCD in the CD (η, D _base ) obtained in S102. This can prevent the data from fluctuating greatly at the boundary between interpolation and extrapolation.

In the present embodiment, fitting by the least square method or fitting by another error evaluation function may be performed. The coefficient of the function is determined so that the sum of the squares of the deviations of all value pairs is minimized. For example,

y = a ₀ + a ₁ x + a ₂ x ² + a ₃ x ³

If the third order polynomial function is selected as the fitting function, the coefficients a ₀ , a ₁ , a ₂ and a ₃ are determined so that the curve of this function fits as closely as possible to the value pairs. In the case of the resist PRL-009 (trade name) manufactured by Fuji Film Co., Ltd., the CD (η, D _base ) obtained in S102 can be expressed well by using a fourth-order polynomial as the fitting function. In order to prevent data errors from being picked up by fitting, it is preferable that the order of the formula used for fitting is as low as possible. Therefore, it is preferable that a maximum value of the order is determined in advance, and if the fitting cannot be performed below this order, it is determined as an error and the drawing process is stopped.

In S107, CD obtained in S106 _{(η, D base)} is used to calculate the dose _{D 50} of the pattern area density of 50% of the line pattern. Specifically, the relationship of D ₅₀ = D _base D _p (η, U) is obtained, and the relationship of (D _base , η, CD, D ₅₀ ) is obtained. In the above equation, _Dp is the proximity effect correction dose.

In S108, from the relationship (D _base , η, CD, D ₅₀ ) obtained in S107, a pair of line widths having an area density of 50% of the proximity effect correction coefficient on the correlation continuous line with respect to the dimension and the reference irradiation amount. A combination (D _base , η) of the proximity effect correction coefficient and the reference irradiation amount at which the irradiation amount in one line pattern matches the iso-focal dose is obtained.

If it is determined in S108 that (D _base , η) is equal to or smaller than a predetermined value, that is, within the limits of the drawing conditions, the process proceeds to S109, and (D _base , η) is calculated using iso-focal dose. (S109). Then, in S110, obtained in S109 _{(D base,} η) to predict the pattern dimension CD _U per area density U when _using, to evaluate the relationship of _{(D base, η).} For example, CD _U is obtained by appropriately interpolating the relationship between D _i and dimension CD _i obtained in the experiment. If any of D _base , η, and irradiation dose exceeds a predetermined value, it is determined as an error and the drawing process is stopped to prevent drawing with an incorrect parameter.

FIG. 5 is an example showing a deviation between the pattern dimension and the design dimension when drawing for each area density U. FIG. In this graph, the dose D for each area density in each dimension CD is calculated from the relationship between D _base and η for each dimension CD shown in FIG. 4, and this dose D and the relationship obtained in S106 are used. Desired. In this example, a mask is assumed in which a chromium film serving as a light shielding layer is formed on a transparent glass substrate such as quartz and a resist film is formed on the chromium film. Moreover, the resist FEP-171 (brand name) by Fuji Film Co., Ltd. is assumed as a resist film. In FIG. 5, the horizontal axis is a shift width from the dimension under the reference irradiation condition, and the vertical axis is a value obtained by subtracting the actual dimension from the target dimension after the proximity effect correction. However, FIG. 5 does not take into account dimensional variations according to the type of resist.

Using FIG. 5, the relationship of (D _base , η) can be evaluated. In S111, when it is determined that the deviation from the design dimension is equal to or less than the predetermined value and within the allowable range, the drawing process is performed using the relationship of (D _base , η) (S112). Specifically, a reference dose map D _base (x, y) and a proximity effect correction coefficient map η (x, y) are created to determine a proximity effect correction dose D _p (x, y). The electron beam dose is obtained from the above. Then, the drawing process is performed by calculating the irradiation time. In S111, when the deviation from the target dimension or the proximity error exceeds a predetermined value, it is determined as an error and the drawing process is stopped to prevent drawing with an incorrect parameter.

  FIG. 1 is a flow when only proximity effect correction is performed without performing loading correction. Next, a flow when proximity effect correction and loading correction are performed will be described with reference to FIG.

  In FIG. 6, S201 to S207 are the same as S101 to S107 of FIG.

In S208, from the relationship of (D _base , η, CD, D ₅₀ ) obtained in S207, a pair of line widths that has an area density of 50% of the proximity effect correction coefficient on the correlation continuous line with respect to the dimension and the reference irradiation amount. The combination of the proximity effect correction coefficient and the reference irradiation amount that coincides with the iso-focal dose in one line pattern, and the pattern dimension CD _{iso at} this time are obtained. Specifically, as _{ΔCD = CD-CD iso, (} D base, η, CD) from the relationship of _{(D base, η, ΔCD)} determine the relationship. At this time, CD may be replaced by ΔCD, but the relationship (D _base , η, ΔCD) may be obtained by interpolating the relationship (D _base , η, CD).

In S208, when it is determined that the CD (η, D _base ) and the irradiation amount are equal to or less than the predetermined value, that is, within the limits of the drawing conditions, the process proceeds to S209, and the CD (D _base , CD using the iso-focal dose η) is calculated (S209). Next, in S210, the resulting _{CD (D base, η)} at S209 to predict the pattern dimension CD _U per area density U when using CD _{(η, D base)} to evaluate the relationship. For example, CD _U is obtained by appropriately interpolating the relationship between D _i and dimension CD _i obtained in the experiment. If any of the difference between the CD (η, D _base ) and the predicted dimension value and the irradiation dose exceeds a predetermined value, it is determined as an error and the drawing process is stopped, and drawing is performed with an incorrect parameter. To prevent it.

The evaluation of S210 can be performed using the relationship of FIG. If it is determined in S211 that the deviation from the design dimension is equal to or less than the predetermined value and is within the allowable range, a drawing process is performed using the relationship of CD (η, D _base ) (S212). Specifically, a reference dose map D _base (x, y) and a proximity effect correction coefficient map η (x, y) are created to determine a proximity effect correction dose D _p (x, y). The amount of electron beam irradiation is obtained from the above. Then, the drawing process is performed by calculating the irradiation time. In S211, if the deviation from the design dimension exceeds a predetermined value, it is determined as an error and the drawing process is stopped to prevent drawing with an incorrect parameter.

  By drawing the sample by the drawing method as described above, it is possible to improve the accuracy of the correction process and reduce the dimensional variation.

  FIG. 7 is a configuration diagram of the electron beam drawing apparatus according to the present embodiment.

  As shown in FIG. 7, the electron beam drawing apparatus includes a drawing unit that draws an electron beam on a sample and a control unit that controls drawing. In the sample chamber 1, a stage 3 on which a mask 2 as a sample is installed is provided. The mask 2 is formed by forming a chromium film as a light shielding film on a transparent glass substrate such as quartz, and further forming a resist film thereon. In this embodiment mode, writing is performed on the resist film with an electron beam. The stage 3 is driven by the stage drive circuit 4 in the X direction (left and right direction on the paper surface) and the Y direction (vertical direction on the paper surface). The moving position of the stage 3 is measured by a position circuit 5 using a laser length meter or the like.

  An electron beam optical system 10 is installed above the sample chamber 1. The optical system 10 includes an electron gun 6, various lenses 7, 8, 9, 11, 12, a blanking deflector 13, a shaping deflector 14, a beam scanning main deflector 15, and a beam scanning sub deflector. 16 and two beam shaping apertures 17, 18 and the like.

  FIG. 8 is an explanatory diagram of a drawing method using an electron beam. As shown in this figure, the pattern 51 drawn on the mask 2 is divided into strip-shaped frame regions 52. Drawing with the electron beam 54 is performed for each frame region 52 while the stage 3 continuously moves in one direction (for example, the X direction). The frame area 52 is further divided into sub-deflection areas 53, and the electron beam 54 draws only necessary portions in the sub-deflection areas 53. The frame area 52 is a strip-shaped drawing area determined by the deflection width of the main deflector 15, and the sub-deflection area 53 is a unit drawing area determined by the deflection width of the sub-deflector 16.

  Positioning of the reference position of the sub deflection area is performed by the main deflector 15, and drawing in the sub deflection area 53 is controlled by the sub deflector 16. That is, the main deflector 15 positions the electron beam 54 in a predetermined sub-deflection region 53, and the sub-deflector 16 determines the drawing position in the sub-deflection region 53. Further, the shape and size of the electron beam 54 are determined by the shaping deflector 14 and the beam shaping apertures 17 and 18. Then, the sub-deflection area 53 is drawn while continuously moving the stage 3 in one direction. When drawing of one sub-deflection area 53 is completed, the next sub-deflection area 53 is drawn. When drawing of all the sub-deflection areas 53 in the frame area 52 is completed, the stage 3 is stepped in a direction orthogonal to the direction in which the stage 3 is continuously moved (for example, the Y direction). Thereafter, the same processing is repeated, and the frame area 52 is sequentially drawn.

In FIG. 7, reference numeral 20 denotes an input unit, which is a part where drawing data of the mask 2 is input to the electron beam drawing apparatus through a magnetic disk as a storage medium. In the input unit 20, the results of drawing the first pattern and the second pattern described above while changing the irradiation amount of the electron beam, the data of the loading effect correction coefficient γ and the loading effect influence range σ _L , and the fogging effect correction Data of the coefficient θ and the influence range σ _i of the fogging effect is also input.

  Based on the information read from the input unit 20, the first calculation unit of the control computer 19 fits the first pattern dimension and the discrete value of the dose to obtain a continuous value. Next, in the second calculation unit of the control computer 19, the relationship between the proximity effect correction coefficient and the dimension for each of the plurality of first patterns having different area densities from the continuous values obtained by the first calculation unit. Is required. From this relationship, the proximity effect correction coefficient that minimizes the dimensional difference between the first patterns and the average dimension of the pattern when using this proximity effect correction coefficient are obtained.

  Further, in the third calculation unit of the control computer 19, a continuous value is obtained by fitting the dimension of the second pattern and the discrete value of the dose. Next, in the fourth calculation unit of the control computer 19, the relationship between the fogging effect correction coefficient and the dimension for each of a plurality of second patterns having different area densities from the continuous values obtained by the third calculation unit. From this relationship, the fogging effect correction coefficient that minimizes the dimensional difference between the second patterns is obtained. Then, the irradiation amount calculation unit 31c described later calculates the electron beam irradiation amount using the proximity effect correction coefficient and the fogging effect correction coefficient.

  The drawing data read from the input unit 20 is temporarily stored in the pattern memory 21 for each frame area 52. The pattern data for each frame area 52 stored in the pattern memory 21, that is, the frame information composed of the drawing position, the drawing graphic data, etc. is corrected by the drawing data correction unit 31 and then the pattern data which is the data analysis unit The data is sent to the decoder 22 and the drawing data decoder 23.

  The drawing data correction unit 31 includes a loading effect correction dimension value calculation unit 31a, a proximity effect correction dose calculation unit 31b, a dose calculation unit 31c, an irradiation time calculation unit 31d, and a fogging effect correction dose calculation unit 31e. Have

The fogging effect correction dose calculation unit 31e calculates an electron beam fogging effect correction relative dose _Dk (x, y) for correcting the fogging effect for each unit area for fogging effect correction. Here, the fogging effect correction relative dose D _k (x, y) is a pattern density dependent value D _f (x, y) depending on the pattern area density U in the fogging effect correction unit region, and the fogging effect correction unit. It is affected by the position-dependent value D _r (x, y) depending on the position of the area in the mask plane. The pattern density dependent value D _f (x, y) is obtained by a distribution function based on the mask in-plane coordinates x, which becomes the pattern area density U (x, y), the fogging effect correction coefficient θ, and the proximity effect correction coefficient η. Further, the pattern density dependent value D _f (x, y) is obtained as follows. For example, as a sample, a mask is prepared in which a chromium film that serves as a light shielding layer and a resist film are formed in this order on a transparent glass substrate such as quartz. A pattern set in which a line pattern having a pattern area density of about 0%, a line pattern of 50%, and a line pattern of 100% is formed for each fogging effect correction unit region with respect to the drawing region of the mask. Is drawn with an electron beam. Next, development processing is performed, and the line width dimension CD of the obtained resist pattern is measured (Measurement 1). Next, the chromium film is patterned by etching the chromium film using the resist pattern as a mask, and the line width dimension CD of the obtained chromium pattern is measured (measurement 2). The relative dose for correcting the difference in the line width dimension CD of each area density line pattern for each pattern set position in the measurement 1 is defined as a mask effect dependent fogging effect corrected relative dose (position-dependent relative value) D _r ( x, y). Then, the product of the pattern density dependent value D _f (x, y) and the position dependent value D _r (x, y) for each fog effect correction unit area is calculated as the fog effect correction relative irradiation dose D in each fog effect correction unit area. Let _k (x, y).

Further, the fogging effect correction dose calculator 31e creates the fogging effect correction relative dose D _k (x, y) in each fogging effect correction unit region as a fogging effect correction relative dose map for each fogging effect correction unit region. To do.

The loading effect correction dimension value calculation unit 31a calculates a loading effect correction dimension value CD (x, y) that corrects a shift in the pattern line width due to the loading effect in the loading effect correction unit area (second area). . Next, an electron beam reference dose map in the second region is created based on the loading effect correction dimension value CD (x, y). This map is created by _obtaining a reference dose map D _base (x, y) corresponding to the loading effect correction dimension value CD (x, y) from the relationship of CD (η, D _base ) shown in FIG. The

In addition, the loading effect correction dimension value calculation unit 31a generates the proximity effect correction coefficient map in the second area based on the loading effect correction dimension value CD (x, y) in the loading effect correction unit area (second area). Created. This map is created by obtaining the proximity effect correction coefficient map η (x, y) corresponding to the loading effect correction dimension value CD (x, y) from the relationship of CD (η, D _base ) shown in FIG. The

By obtaining the reference dose map D _base (x, y) and the proximity effect correction coefficient map η (x, y) from the relationship of CD (η, D _base ) shown in FIG. It is possible to realize a correction with a dimension value of.

In the proximity effect correction dose calculation unit 31b, the proximity effect correction unit is calculated from the reference dose map D _base (x, y) obtained by the loading effect correction dimension value calculation unit 31a and the proximity effect correction coefficient map η (x, y). proximity effect correction dose D _p is calculated in the region (first region).

In the irradiation amount calculation unit 31c, the irradiation amount of the electron beam at the actual irradiation position is calculated. That is, the electron beam dose D is obtained based on the fogging effect corrected relative dose D _k (x, y) and the proximity effect corrected dose D _p .

  In the irradiation time calculation unit 31d, the irradiation time T of the electron beam at each position in the drawing region is calculated. Here, since the irradiation amount D is the product of the irradiation time T and the current density J, the irradiation time T can be obtained by gradually decreasing the irradiation amount D by the current density J.

  Information from the pattern data decoder 22 is sent to a blanking circuit 24 and a beam shaper driver 25. Specifically, blanking data based on the data is created by the pattern data decoder 22 and sent to the blanking circuit 24. Desired beam size data is also created and sent to the beam shaper driver 25. Then, a predetermined deflection signal is applied from the beam shaper driver 25 to the shaping deflector 14 of the electron optical system 10 to control the size of the electron beam 54.

  7 is connected to a settling time determination unit 29. The settling time determination unit 29 is connected to a sub deflection region deflection amount calculation unit 28. The sub deflection region deflection amount calculation unit 28 is a pattern data decoder. 22 is connected. The deflection control unit 30 is connected to the blanking circuit 24, the beam shaper driver 25, the main deflector driver 26, and the sub deflector driver 27.

  The output of the drawing data decoder 23 is sent to the main deflector driver 26 and the sub deflector driver 27. Then, a predetermined deflection signal is applied from the main deflector driver 26 to the main deflection unit 15 of the electron optical system 10, and the electron beam 54 is deflected and scanned to a predetermined main deflection position. Further, a predetermined sub deflection signal is applied from the sub deflector driver 27 to the sub deflector 16, and drawing in the sub deflection region 53 is performed.

  Next, a drawing method by the electron beam drawing apparatus will be described.

  First, the mask 2 is placed on the stage 3 in the sample chamber 1. Next, the position of the stage 3 is detected by the position circuit 5, and the stage 3 is moved to a position where drawing can be performed by the stage drive circuit 4 based on a signal from the control computer 19.

  Next, an electron beam 54 is emitted from the electron gun 6. The emitted electron beam 54 is collected by the illumination lens 7. Then, the blanking deflector 13 performs an operation for irradiating the mask 2 with the electron beam 54.

  The electron beam 54 incident on the first aperture 17 passes through the opening of the first aperture 17 and is then deflected by the shaping deflector 14 controlled by the beam shaper driver 25. And it passes through the opening part provided in the 2nd aperture 18, and becomes a beam shape which has a desired shape and a dimension. This beam shape is a drawing unit of the electron beam 54 applied to the mask 2.

  The electron beam 54 is shaped into a beam shape and then reduced by the reduction lens 11. The irradiation position of the electron beam 54 on the mask 2 is controlled by the main deflector 15 controlled by the main deflector driver 26 and the sub deflector 16 controlled by the sub deflector driver 27. The main deflector 15 positions the electron beam 54 in the sub deflection region 53 on the mask 2. The sub deflector 16 positions the drawing position in the sub deflection region 53.

  Drawing with the electron beam 54 on the mask 2 is performed by scanning the electron beam 54 while moving the stage 3 in one direction. Specifically, the pattern is drawn in each sub deflection region 53 while moving the stage 3 in one direction. After all the sub-deflection areas 53 in one frame area 52 have been drawn, the stage 3 is moved to a new frame area 52 and drawn similarly.

  After drawing all the frame regions 52 of the mask 2 as described above, the drawing is replaced with a new mask and drawing by the same method as described above is repeated.

  Next, drawing control by the control computer 19 will be described.

  The control computer 19 reads the drawing data of the mask recorded on the magnetic disk by the input unit 20. The read drawing data is temporarily stored in the pattern memory 21 for each frame area 52.

  The drawing data for each frame area 52 stored in the pattern memory 21, that is, the frame information composed of the drawing position and drawing graphic data, etc. is corrected by the drawing data correction unit 31 as described above, and then subjected to data analysis. The data is sent to the sub deflection area deflection amount calculation unit 28, the blanking circuit 24, the beam shaper driver 25, the main deflector driver 26, and the sub deflector driver 27 via the pattern data decoder 22 and the drawing data decoder 23. .

  The pattern data decoder 22 generates blanking data based on the drawing data and sends it to the blanking circuit 24. Further, desired beam shape data is created based on the drawing data and is sent to the sub deflection region deflection amount calculation unit 28 and the beam shaper driver 25.

  The sub deflection region deflection amount calculation unit 28 calculates the deflection amount (movement distance) of the electron beam for each shot in the sub deflection region 53 from the beam shape data created by the pattern data decoder 22. The calculated information is sent to the settling time determination unit 29, and the settling time corresponding to the movement distance by the sub deflection is determined.

  The settling time determined by the settling time determination unit 29 is sent to the deflection control unit 30, and then the deflection control unit 30 measures the blanking circuit 24, the beam shaper driver 25, It is appropriately sent to either the deflector driver 26 or the sub deflector driver 27.

  In the beam shaper driver 25, a predetermined deflection signal is applied to the shaping deflector 14 of the optical system 10, and the shape and size of the electron beam 54 are controlled.

  The drawing data decoder 23 generates positioning data for the sub deflection region 53 based on the drawing data, and this data is sent to the main deflector driver 26. Next, a predetermined deflection signal is applied from the main deflector driver 26 to the main deflector 15, and the electron beam 54 is deflected and scanned to a predetermined position in the sub deflection region 53.

  The drawing data decoder 23 generates a control signal for scanning the sub deflector 16 based on the drawing data. After the control signal is sent to the sub deflector driver 27, a predetermined sub deflection signal is applied from the sub deflector driver 27 to the sub deflector 16. Drawing in the sub deflection region 53 is performed by repeatedly irradiating the electron beam 54 after the settling time has elapsed.

  FIG. 9 is a conceptual diagram showing an example of the flow of the dose calculation method in the present embodiment. This figure particularly shows a flow until the electron beam irradiation amount is calculated in the electron beam drawing method.

  First, as a global region, a unit region (first mesh region) for fogging effect correction divided in a mesh shape with a size of μm to mm, for example, 0.5 mm to 1 mm is defined. Similarly, a loading effect correction unit region (second mesh region) divided into a mesh shape with a size of μm to mm, for example, 0.5 mm to 1 mm is defined. Here, as an example, a description will be given of an example in which a unit region for fogging effect correction and a unit region for correction of loading effect are defined as regions having the same dimensions. That is, a unit region for fogging / loading effect correction divided into a mesh shape with a size of μm to mm order, for example, 0.5 to 1 mm is defined as a global region. However, the present invention is not limited to this, and the unit area for fogging effect correction and the unit area for loading effect correction may be defined as areas having different dimensions. Further, as a local region, for proximity effect correction divided in a mesh shape in the order of μm smaller than the size of the unit region for fogging effect correction and the unit region for loading effect correction, for example, a size of 1 μm or less. A unit small region (third mesh region) is defined.

As described above, the fogging effect correction dose calculation unit 31e in FIG. 7 performs the fogging effect correction relative dose _Dk (x, y) of the charged particle beam for correcting the fogging effect for each fogging / loading effect correction unit region. Is calculated. Specifically, first, the fog correction data V (x, y) in the fog / loading effect correction unit area is calculated. V (x, y) is obtained by the following equation as shown in FIG.

ｇ（ｘ，ｙ）は、かぶり効果の分布関数であり、かぶり効果の影響範囲σ_ｆのガウス分布で近似できる。かぶり効果散乱半径はｃｍのオーダーであり、かぶり・ローディング効果補正単位領域をかぶり効果散乱半径の１０分の１以下とすることで、Ｖ（ｘ，ｙ）は下記式から求められる。

g (x, y) is a distribution function of the fogging effect, and can be approximated by a Gaussian distribution of the influence range σ _f of the fogging effect. The fogging effect scattering radius is in the order of cm, and V (x, y) can be obtained from the following equation by setting the fogging / loading effect correction unit region to 1/10 or less of the fogging effect scattering radius.

ρ（ｘ，ｙ）は、各かぶり・ローディング効果単位領域のパターン面積密度、Ｓ_ｍｅｓｈは、かぶり・ローディング効果単位領域の面積である。ここで、かぶり補正用単位領域とローディング補正用単位領域とが異なる場合には、それぞれρ（ｘ，ｙ）を計算すればよい。

ρ (x, y) is the pattern area density of each fog / loading effect unit area, and S _mesh is the area of the fog / loading effect unit area. Here, if the fog correction unit area and the loading correction unit area are different, ρ (x, y) may be calculated.

As described above, the fogging effect correction relative dose D _k (x, y) is equal to the pattern density dependent value D _f (x, y) depending on the pattern area density U in the unit area for fogging effect correction, and the fogging effect correction. _Is affected by the position-dependent value D _r (x, y) depending on the position of the unit area in the mask plane. Therefore, the product of the pattern density dependent value D _f (x, y) and the position dependent value D _r (x, y) for each fog effect correction unit region is the fog effect correction relative irradiation dose D in each fog effect correction unit region. Let _k (x, y).

The fogging effect correction dose calculation unit 31e creates a fogging effect correction relative dose map for each fogging effect correction unit area based on the fogging effect correction relative dosage _Dk (x, y) in each fogging effect correction unit area. To do.

7 calculates the correction dimension value CD (x, y) for correcting the shift of the pattern line width dimension due to the loading effect in each fogging / loading effect correction unit area. Here, the loading effect correction dimension value CD (x, y) includes the pattern density dependency value L (x, y) depending on the pattern area density in the fogging / loading effect correction unit area and the fogging / loading effect correction unit area. It is affected by the position dependent value P (x, y) depending on the position in the mask plane. Therefore, first, a pattern density dependent loadig effect correction dimension value (pattern density dependent value) L (x, y) is calculated. This value is obtained from the following equation.

The loading effect correction coefficient γ, g (x, y) is a distribution function of the loading effect. Here, g (x, y) can be approximated by a Gaussian distribution having a loading effect influence range (effect radius) σ _L. By setting the loading effect radius to the order of cm and setting the fogging / loading effect correction unit area to 1/10 or less of the loading effect radius, L (x, y) can be obtained from the following equation.

ρ（ｘ，ｙ）は、各かぶり・ローディング単位領域のパターン面積密度、Ｓ_ｍｅｓｈは、かぶり・ローディング効果単位領域の面積である。

ρ (x, y) is the pattern area density of each fog / loading unit area, and S _mesh is the area of the fog / loading effect unit area.

Next, a position-dependent loadig effect correction dimension value (position-dependent value) P (x, y) is obtained by experiment. With reference to the technique for obtaining the fogging effect corrected relative dose (position-dependent relative value) D _r (x, y), the line width at each pattern set position in measurement 1 from the line width dimension CD at each pattern set position in measurement 2 The difference value obtained by subtracting the dimension CD may be the position-dependent loading effect correction dimension value (position-dependent value) P (x, y). Further, this loading effect correction dimension value is a value for each pattern set position, but if it is interpolated to a value for each fogging / loading effect unit area, a more accurate loading effect correction dimension value P (x , Y).

  The fogging effect correction dimension value (pattern density dependent value) L (x, y) depending on the pattern area density and the fogging effect correction dimension value (position dependent value) P (x, y) depending on the position in the mask surface. The sum for each loading effect unit area is defined as a loading effect correction dimension value CD (x, y) in each fogging / loading effect correction unit area. By considering not only the pattern density dependent value but also the position dependent value, a more accurate loadig effect correction dimension value can be obtained.

  Here, the loading effect correction dimension value in the fogging / loading effect correction unit area is suitable not only for the dimension caused by the loading effect but also by adding a dimension for correcting non-uniformity in the mask surface other than that caused by etching. is there. For example, dimensional non-uniformity due to uneven development in the developing device can be mentioned.

The loading effect correction dimension value calculation unit 31a in FIG. 7 performs the electron beam 54 in each fogging / loading effect correction unit area based on the loading effect correction dimension value CD (x, y) in each fogging / loading effect correction unit area. Create a reference dose map. This map obtains the reference dose map D _base (x, y) corresponding to the loading effect correction dimension value CD (x, y) from the relationship of CD (η, D _base ) obtained in the present embodiment. Created.

The loadig effect correction dimension value calculation unit 31a creates a proximity effect correction coefficient map in each fogging / loading effect correction unit area based on the loadig effect correction dimension value CD (x, y) in each fogging / loading effect correction unit area. To do. This map obtains the proximity effect correction coefficient η (x, y) corresponding to the loadig effect correction dimension value CD (x, y) from the relationship of CD (η, D _base ) obtained in the present embodiment. Created. Then, by obtaining a reference dose map D _base (x, y) and a proximity effect correction coefficient map η (x, y) from the relationship of CD (η, D _base ), the same dimension value in any pattern category Correction can be realized.

The proximity effect correction dose calculation unit 31b in FIG. 7 corrects the proximity effect correction irradiation of the electron beam 54 that corrects the proximity effect in the unit small region for proximity effect correction from the reference dose map and the proximity effect correction coefficient map. The quantity D _p is calculated. Then, to calculate the proximity effect correction dose D _p in the actual irradiation position (x, y). Here, the proximity effect correction dose D _p is the actual irradiation position (x, y) obtained by the interpolation calculation using the proximity effect correction dose D _p in the four proximity effect correction unit area around surrounding the.

  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, the charged particle beam drawing apparatus described in this embodiment may include an evaluation unit that evaluates the accuracy of fitting. For example, a pattern dimension and a design dimension when drawing using the obtained parameters are input to the input unit 20 of FIG. Next, the evaluation unit obtains a relationship as shown in FIG. If it is determined that there is a problem with the fitting accuracy based on this relationship, the drawing process is stopped. Thereby, the drawing accuracy of the electron beam can be determined in a preliminary stage.

  In the above embodiment, an electron beam is used. However, the present invention is not limited to this, and the present invention can also be applied to cases where other charged particle beams such as an ion beam are used.

DESCRIPTION OF SYMBOLS 1 Sample chamber 2 Mask 3 Stage 4 Stage drive circuit 5 Position circuit 6 Electron gun 7, 8, 9, 11, 12 Various lenses 10 Optical system 13 Blanking deflector 14 Molding deflector 15 Main deflector 16 Sub deflector 17 First aperture 18 Second aperture 19 Control computer 20 Input unit 21 Pattern memory 22 Pattern data decoder 23 Drawing data decoder 24 Blanking circuit 25 Beam shaper driver 26 Main deflector driver 27 Sub deflector driver 28 Sub deflection area deflection Amount calculation unit 29 Settling time determination unit 30 Deflection control unit 31 Drawing data correction unit 31a Loading effect correction dimension value calculation unit 31b Proximity effect correction irradiation amount calculation unit 31c Irradiation amount calculation unit 31d Irradiation time calculation unit 31e Fog effect correction irradiation amount calculation Part 51 Pattern to be drawn 52 Frame region 53 Sub deflection region 54 Electron beam

Claims (5)

Drawing a plurality of patterns with different area densities by changing the proximity effect correction coefficient for each reference irradiation amount of the charged particle beam, and measuring the dimension of the pattern after drawing;
Calculating an optimum proximity effect correction coefficient for each reference dose and a deviation from the design dimension of the pattern when drawn using the proximity effect correction coefficient;
Fitting the calculated proximity effect correction coefficient and a discrete value of deviation from the design dimension by the following equation:
η = η ₀ / erfc (ΔCD / 2 / σ _F ) / 2
(Where erfc is the error function, η is the proximity effect correction coefficient, ΔCD is the deviation of the dimension from the design dimension, η ₀ and σ _F are the fitting parameters, and the influence range of the standard proximity effect correction coefficient and the forward scattering respectively. Represents.)
A charged particle beam drawing method comprising: interpolating and / or extrapolating the proximity effect correction coefficient and a deviation from the design dimension using the fitting result.   The charged particle beam drawing method according to claim 1, further comprising a step of performing spline interpolation between the calculated proximity effect correction coefficient and a discrete value of deviation from the design dimension. Drawing a plurality of patterns with different area densities by changing the proximity effect correction coefficient for each reference irradiation amount of the charged particle beam, and measuring the dimension of the pattern after drawing;
For each of the proximity effect correction coefficients, calculating a deviation from the design dimension of the pattern and a reference dose when drawing using the proximity effect correction coefficient; and
Fitting the calculated deviation from the design dimension and a discrete value of the reference dose with a polynomial;
A charged particle beam drawing method comprising: a step of interpolating and / or extrapolating the deviation from the design dimension and the reference irradiation amount using the fitting result.   The charged particle beam drawing method according to claim 3, further comprising a step of performing spline interpolation between the calculated deviation from the design dimension and the discrete value of the reference dose. 5. The charged particle beam drawing method according to claim 2, wherein the calculated discrete values are weighted for fitting.

Images