JP5416998B2 - Charged particle beam drawing method and charged particle beam drawing apparatus - Google Patents

Description

  The present invention relates to a charged particle beam drawing method and a charged particle beam drawing apparatus.

  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, correction processing for changing the beam irradiation amount is necessary so that the pattern dimension is 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

As described above, in the method of Patent Document 2, several types of proximity effect correction coefficient η, reference irradiation amount D _base, and proximity effect influence range σ _B are determined in advance, and a plurality of patterns are drawn according to each value. Work is required. The relationship between the optimal proximity effect correction coefficient η, the reference dose D _base, and the dimension CD is obtained by interpolating the D _{base on} which the drawing experiment was performed and the optimal combination of η and CD at that time. However, the conventional method has a problem that a deviation from a 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. In view of the above, an object of the present invention is to provide a charged particle beam drawing method and a charged particle beam drawing apparatus that can improve the accuracy of correction processing and reduce dimensional variation.

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

The first aspect of the present invention is a first step in which a plurality of patterns having different area densities are drawn while changing the irradiation amount of the charged particle beam, and the dimension of the pattern after drawing is measured to obtain the relationship with the irradiation amount. When,
From the continuous values obtained by fitting the discrete values of the dimensions and dose, the relationship between the proximity effect correction coefficient and the dimensions is determined for each of the multiple patterns with different area densities, and the difference in dimensions between the patterns is determined from this relationship. And a second step of obtaining a minimum proximity effect correction coefficient.

The first aspect of the present invention includes a third step of obtaining an irradiation amount of the charged particle beam using the proximity effect correction coefficient obtained in the second step,
And a fourth step of drawing with a charged particle beam on a predetermined resist based on the dose determined in the third step.
In this case, an approximate expression corresponding to the type of resist is used for fitting.

  In the first aspect of the present invention, a threshold model expression or a cubic polynomial can be used as the approximate expression.

According to a second aspect of the present invention, an input unit to which a result of drawing a plurality of patterns having different area densities by changing the irradiation amount of the charged particle beam is input;
A first calculation unit that obtains a continuous value by fitting a discrete value of a pattern dimension and an irradiation amount;
The relationship between the proximity effect correction coefficient and the dimension is obtained for each of a plurality of patterns having different area densities from the continuous values obtained by the first calculation unit, and the proximity that minimizes the dimension difference between the patterns is obtained from this relationship. A second calculation unit for obtaining an effect correction coefficient;
An irradiation amount calculation unit for determining an irradiation amount of the charged particle beam using the proximity effect correction coefficient obtained by the second calculation unit;
The present invention relates to a charged particle beam drawing apparatus having a drawing unit for drawing with a charged particle beam based on an irradiation amount obtained by an irradiation amount calculating unit.

  In the second aspect of the present invention, an equation of a threshold model or a cubic polynomial can be used for fitting.

  According to the first aspect of the present invention, it is possible to improve the accuracy of the correction process and reduce the dimensional variation.

  According to the second aspect of the present invention, there is provided a charged particle beam drawing apparatus capable of improving the accuracy of correction processing and reducing dimensional fluctuations.

It is a 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 made the dimension in each irradiation amount into a graph. It is an example which shows the procedure which performs fitting in this Embodiment. 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 made the dimension in each reference irradiation amount by a conventional method a graph. FIG. 7 is an example in which the relationship between the proximity effect correction coefficient and the dimension is obtained for each of a plurality of patterns having different area densities by linear interpolation. It is the figure which showed the variation | change_quantity of the dimension when a proximity effect correction coefficient and a reference irradiation amount were changed. This is an example in which the deviation between the pattern dimension and the design dimension is compared between the present embodiment (a) and the conventional method (b). 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.

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

As shown in FIG. 1, the electron beam writing method includes a step (S101) of obtaining a correlation CD (η, D _base ) between a proximity effect correction coefficient η and a reference dose D _base with respect to a pattern dimension CD, and a loading effect correction coefficient. A step (S102) for obtaining γ and an influence range σ _{L of} the loading effect, a step (S103) for obtaining a loading effect correction dimension value CD (x, y), and a reference dose map D _base (x, y) are created. A step (S104), a step (S105) for creating a proximity effect correction coefficient map η (x, y), a step (S106) for obtaining a proximity effect correction dose D _p (x, y), an electron beam A series of processes including an irradiation amount calculation step (S107) at a figure irradiation position, an irradiation time calculation step (S108), and a drawing step (S109) as an example of a step for obtaining an irradiation amount Having.

In S101, the correlation effect correction coefficient η with respect to the pattern dimension CD and the correlation CD (η, D _base ) between the reference dose D _base are obtained. Specifically, first, a plurality of patterns having different area densities U are drawn while changing the electron beam irradiation amount D, and the dimension CD of the pattern after drawing is measured to obtain the relationship with the irradiation amount D. Here, it may be drawn by changing the _{D base} as _{D = D base} without using the proximity effect correction coefficient D. Alternatively, using a set of a plurality of _{D base} and the proximity effect correction coefficient _{η, D = D base D p} (η, U) may be drawn by changing the. Next, a relationship between the proximity effect correction coefficient η and the dimension CD is obtained for each of a plurality of patterns having different area densities U from continuous values obtained by fitting the discrete values of the dimension CD and the irradiation amount D. From this relationship, A proximity effect correction coefficient η that minimizes the difference in dimension CD between patterns is obtained.

  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. Drawing on the mask while changing the value of D. Next, the line width dimension CD of each pattern after drawing is measured.

FIG. 3 is an example in which the dimensions at each dose are graphed. In the present embodiment, the discrete values (● in FIG. 3) of the dimension CD and the dose D are fitted with an appropriate approximate expression. Thereby, a continuous value as shown by the curve of FIG. 3 is obtained. 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. For example, fitting by the least square method or fitting by another error evaluation function is used, and 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 drawing an electron beam on a resist by the series of steps shown in FIG. 1, it is preferable to use an approximate expression corresponding to the type of resist for fitting. For example, in the case of resist FEP-171 (trade name) manufactured by Fuji Film Co., Ltd., it is preferable to perform fitting using a threshold model equation. In the case of the resist PRL-009 (trade name) manufactured by the same company, it is preferable to perform fitting using a cubic polynomial.

  FIG. 4 is an example when fitting is performed using the threshold model equation.

In FIG. 4, first, (σ _F , D _th , η) is independently taken for each area density, and the obtained data is fitted by the following threshold model formula (S201). G (ξ) is a function representing the dose distribution due to forward scattering at the pattern edge, and an error function is used here as an example. η is the proximity effect correction coefficient, U is the pattern area density, D _th is the development threshold, D is the dose, σ _F is the range of influence of forward scattering, and CD is the pattern dimension. CD ₀ is a design dimension or a dimension obtained by adding a change due to development independent of the irradiation amount to the design dimension.
G (ξ) + ηU = D _th / D
G (ξ) = erfc (ξ / σ _F ) / 2
ξ = (CD−CD ₀ ) / 2

Next, an appropriate set of CD (η, D _base ) is obtained from the curve obtained by fitting (S202). Next, the proximity effect correction coefficient η _c at the reference dose D _base whose pattern dimension is equal to the dimension of the resist image is obtained (S203). Then, (σ _B , D _th ) is independently set for each area density U, η _c is taken in common, and the obtained data is fitted again with the threshold model equation (S204). Here, the eta _c, using a value obtained in S203. Next, after _obtaining the proximity effect correction coefficient η _{c ′} at the reference dose D _base whose pattern dimension is equal to the dimension of the resist image (S205), η _c and η _{c ′} are compared (S206). When the difference between η _c and η _{c ′} is equal to or smaller than the predetermined value, the fitting is finished. When the difference between η _c and η _{c ′} is larger than the predetermined value, the process returns to S204 and the subsequent processing is repeated. In addition, the value _calculated | required by S205 is used for (eta) _c in a repetition process.

  FIG. 5 shows an example in which the relationship between the proximity effect correction coefficient η and the dimension CD is obtained for each of a plurality of patterns having different area densities U from continuous values obtained by fitting the discrete values of the dimension CD and the dose D. It is. In FIG. 5, the dotted line indicates the difference in dimension CD between patterns. The value of η when the difference in dimension CD is the minimum is the optimum proximity effect correction coefficient that satisfies the proximity effect correction condition.

FIG. 6 is an example of a graph showing the correlation between the proximity effect correction coefficient η and the reference dose D _base with respect to the pattern dimension CD. That is, the change amount of the dimension CD when the proximity effect correction coefficient and the reference irradiation amount are changed is shown along the continuous line obtained by the above fitting. By using this relationship, it is possible to change the dimensions while satisfying the proximity effect correction condition. CD on the horizontal axis is a dimensional bias from the reference irradiation condition, and represents a dimensional variation range given for the purpose of loading correction. The reference irradiation condition is that a combination of a proximity effect correction coefficient and a reference irradiation amount at which the irradiation amount in a line pattern of a line width of 1 to 1 with an area density of 50% coincides with an iso-focal dose, and a standard proximity effect correction coefficient η Data points calculated as shown in FIG. 6 or interpolation thereof are defined as ₀ and the standard reference dose D _{base, 0} .

Here, for comparison, a method for obtaining the correlation CD (η, D _base ) between the proximity effect correction coefficient η and the reference irradiation amount D _base with respect to the conventional pattern dimension CD will be described.

In the conventional method, the optimum η under the experimental conditions (D _base , σ _B ) is obtained by a drawing experiment in which D _base is fixed and η is changed. For example, the pattern set shown in FIG. 2 is arranged, drawing is performed on the mask while changing each value of the proximity effect correction coefficient, the reference irradiation amount, and the influence range, and the line width dimension CD of each pattern after drawing is measured.

FIG. 7 is an example in which the relationship between the dose and the size in the influence range (σ _B ) of a certain proximity effect is graphed. In this example, the four reference doses D _base are drawn using the relationship of D = D _base D _p (η, U) by changing the proximity correction coefficient in five ways for each reference dose. When U is close to 1, since the amount of change D _p when varying η decreases, data is discretely as shown in FIG.

FIG. 8 shows an example in which data of a certain reference dose D _base is extracted from the data shown in FIG. 7 and the relationship between the proximity effect correction coefficient and the dimension is obtained by linear interpolation for each pattern having a different area density U. It is. In FIG. 8, the horizontal axis represents the proximity correction coefficient used for drawing, and the vertical axis represents the dimensions. In FIG. 8, a dotted line indicates a difference in dimension CD between patterns. The value of η when the difference in dimension CD is the minimum is the optimum proximity effect correction coefficient that satisfies the proximity effect correction condition. When the data of FIG. 8 is obtained using four types of D _base as shown in FIG. 2, four optimal combinations of proximity effect correction coefficients and reference irradiation amounts are calculated for each range of influence of the proximity effect. Then, the optimum influence range σ is defined as the influence range in which the difference between the line width dimensions of the pattern area densities of about 0%, 50%, and 100% in the combination of the optimum proximity effect correction coefficient and the reference irradiation amount becomes the smallest. Decide _B.

Next, based on the combination of the optimum proximity effect correction coefficient, the reference irradiation amount, the influence range σ _B , and the four dimensions at this time, the optimum proximity effect correction coefficient, the reference irradiation amount, and the dimension are successively obtained by interpolation. To have a good correlation. Of the proximity effect correction coefficient on the correlation continuous line with respect to the dimension and the reference irradiation amount, the proximity effect correction coefficient and the reference irradiation amount that coincide with the iso-focal dose in the line pattern having a line width of 1 to 1 with an area density of 50%. Are defined as a standard proximity effect correction coefficient η ₀ and a standard reference dose D _{base, 0} , and set as a reference irradiation condition.

FIG. 9 shows an interpolation based on four combinations of the optimum proximity effect correction coefficient and the reference dose, centering on the standard proximity effect correction coefficient η ₀ and the standard reference dose D _{base, 0} and the dimensions at this time. It is the figure which showed the variation | change_quantity of dimension CD, ie, correlation CD ((eta), _Dbase ) of a correction line width dimension when a proximity effect correction coefficient and a reference irradiation amount are changed along the correlation continuous line which performed.

FIG. 10 is an example in which the deviation between the pattern dimension and the design dimension when drawing is compared between the present embodiment (FIG. 10A) and the conventional method (FIG. 10B). The graph of FIG. 10A calculates the irradiation dose D for each area density in each dimension CD from the relationship between D _base and η for each dimension CD shown in FIG. 7 is obtained using the irradiation amount D used when creating the composite data 7 and the true relationship of the dimensions. 10B is also obtained in the same manner as described above after calculating the dose D for each area density in each dimension CD from the relationship between D _base and η for each dimension CD shown in FIG. . In these examples, 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 as a sample. Moreover, the resist FEP-171 (brand name) by Fuji Film Co., Ltd. is assumed as a resist film. 10 (a) and 10 (b), the horizontal axis is the shift width from the dimension under the reference irradiation condition, and is the dimension fluctuation width given as loading correction. The vertical axis represents a value obtained by subtracting the actual dimension from the target dimension after correcting the loading of the chromium film. However, in FIGS. 10A and 10B, noise added as a measurement error to the data of FIGS. 3 and 7 is not added. That is, dimensional variation according to the type of resist is not taken into consideration.

As can be seen from FIG. 10, by using the parameters obtained by the present embodiment, a pattern with a substantially uniform dimension can be obtained regardless of the deviation from the design dimension of the chromium film. When the parameters obtained by the conventional method are used, the deviation of the obtained pattern from the design dimension becomes large. In this embodiment, the relationship between the dose and the dimension is continuously obtained by using an approximate expression corresponding to the physical properties of the resist, and D _base and η for each CD are calculated without a substantial interpolation operation in fine steps. On the other hand, in the conventional method, (1) D _base and η for each CD are calculated by interpolating the relationship between the proximity effect correction coefficient and the dimension with a linear expression, and (2) using this result, a rough result is obtained. By linear interpolation for the step CD. In the conventional method, (1) is a dimensional variation in parameters (CD = −20, −7, 7, 20 nm) calculated by D _base used in the experiment, and (2) is a dimensional variation between these. , Each has become a cause of spread. In the conventional method, it is necessary to predict the optimum value of η for each D _base during a drawing experiment and determine the range of η used for drawing. On the other hand, in the present embodiment, it is only necessary to change the value of the electron beam irradiation amount D _base so as to vary appropriately, which is simpler than the conventional method.

After the proximity effect correction coefficient η with respect to the pattern dimension CD and the correlation CD (η, D _base ) between the reference dose D _base are obtained as described above, the process of S102 in FIG. 1, ie, the loading effect correction is performed. A step of obtaining the coefficient γ and the influence range σ _{L of the} loading effect is performed.

  In the present embodiment, a small area (first area) for proximity effect correction divided in a mesh shape with a dimension of μm order, for example, 1 μm or less, is defined as a local area in the drawing area on the mask. . In addition, as a global region, a small region (second region) for correcting a loading effect, which is divided in a mesh shape with a size of μm to mm, for example, 0.5 mm to 1 mm, is defined.

  Next, in S103, a correction dimension value CD (x, y) for correcting the shift of the pattern line width dimension due to the loading effect in the second region is calculated. Here, the loading effect correction dimension value CD (x, y) depends on the pattern density dependency value L (x, y) depending on the pattern area density in the second region and the mask in-plane position of the second region. The position-dependent value P (x, y) is affected. Accordingly, first, a loading effect correction dimension value (pattern density dependent value) L (x, y) depending on the pattern density is calculated. L (x, y) is obtained by 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 with a loading effect influence range σ _L. The loading effect radius is in the order of cm, and L (x, y) can be obtained by the following equation by setting the second region to be one tenth or less of the loading effect radius. Note that ρ (x, y) is the pattern area density of the second region, and S _mesh is the area of the second region.

  Next, a position-dependent loading effect correction dimension value (position-dependent value) P (x, y) is obtained by experiment. For example, as a sample, a mask is prepared 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. Then, with respect to the drawing area of the mask, 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 set as a mask for each second area. EB is drawn on the entire surface. Thereafter, the resist film is developed and the line width dimension CD is measured (Measurement 1). Next, the chromium film is etched using this resist pattern as a mask, and the line width dimension CD of the obtained chromium film pattern is measured (measurement 2). Then, a difference value obtained by subtracting the line width dimension CD for each pattern set position in measurement 1 from the line width dimension CD for each pattern set position in measurement 2 is used as a position-dependent loading effect correction dimension value (position dependent value) P ( x, y). This loading effect correction dimension value is a value for each pattern set position, but if it is interpolated to be a value for each second area, a more accurate loading effect correction dimension value P (x, y ) Is obtained.

  Then, a loading effect correction dimension value (pattern density dependency value) L (x, y) depending on the pattern area density and a loading effect correction dimension value (position dependency value) P (x, y) depending on the position in the mask surface. Is the loading effect correction dimension value CD (x, y) in each second region. By considering not only the pattern density dependent value but also the position dependent value, a more accurate loading effect correction dimension value can be obtained. Note that the loading effect correction dimension value in the second region may include not only the dimension caused by the loading effect but also a dimension for correcting non-uniformity in the mask surface other than the etching origin. For example, dimensional non-uniformity due to uneven development in the developing device can be mentioned.

Next, in S104, based on the loading effect correction dimension value CD (x, y) in the second region, an electron beam reference dose map in each second region is created. 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. Is done.

In S105, a proximity effect correction coefficient map in each second region is created based on the loading effect correction dimension value CD (x, y) in each second region. This map is created by obtaining the proximity effect correction coefficient 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. 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 ) shown in FIG. Correction of the same dimension value can be realized.

Next, in S106, the proximity effect correction dose D of the electron beam that corrects the proximity effect in the first region, which is a unit small region for proximity effect correction, using the reference dose map and the proximity effect correction coefficient map. _p is calculated. Below is an equation for calculating the proximity effect correction dose D _p in the present embodiment. Note that Base Dose (x, y) is D _base (x, y), and (x ′, y ′) indicates the position in the mask plane.

As shown in the above equation, the proximity effect correction dose _Dp is the reference dose D _base (x, y), the proximity effect correction coefficient η (x, y), the proximity effect influence range σ _{B, and the} zeroth order proximity effect correction. Data U (x, y), (when i> 0) i-th proximity effect correction data V _i (x, y) can be used. Here, the first and subsequent values of d _i defined by the proximity effect correction coefficient η (x, y), the 0th order proximity effect correction data U (x, y), and the i th order proximity effect correction data V _i (x, y) are used. By adding the term, the dose can be calculated with high accuracy. By taking the first-order sections following as d _i, can be calculated accurately proximity effect correction dose D _p. It should be noted that the number of calculations to be performed can be arbitrarily set within the limits of calculation capacity and calculation time. Effectively, by calculating up to the third order, it is possible to keep the calculation error within about 0.5%.

After calculating the proximity effect correction dose D _p in each first region, to calculate a proximity effect correction dose D _p in the actual irradiation position (x, y). The proximity effect correction dose D _p at the actual irradiation position (x, y) is interpolated by using the proximity effect correction dose D _p in the four first regions surrounding the actual irradiation position (x, y). Calculate to find.

The following formula is a diagram showing an example of the distribution function formula g (x, y) in the present embodiment. Using this equation, 0th-order proximity effect correction data U (x, y) and i-th order proximity effect correction data V _i (x, y) are calculated.

Next, in S107, the dose D at the actual irradiation position is calculated. Specifically, based on the proximity effect correction dose D _p as described above, the dose D of the electron beam is calculated.

  Next, in S108, the irradiation time T of the electron beam at each position in the drawing area is calculated. Since the irradiation amount D can be defined by the product of the irradiation time T and the current density J, the irradiation time T can be obtained by dividing the irradiation amount D by the current density J.

  Next, in S109, drawing is performed on the resist with an electron beam based on the obtained dose. That is, in an electron beam drawing apparatus described later, the control computer outputs a signal to the deflection control unit so that beam irradiation to the sample is turned off at the obtained irradiation time T. The deflection control unit controls the blanking circuit so as to deflect the electron beam in accordance with the obtained irradiation time T along the signal. Then, after irradiating the sample with a desired dose D, the electron beam deflected by the blanking circuit is shielded by the aperture so as not to reach the sample.

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

  As shown in FIG. 11, 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. 12 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. 11, 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. The input unit 20 also receives a result of drawing a plurality of patterns with different area densities by changing the electron beam irradiation amount, and data on the loading effect correction coefficient γ and the loading effect influence range σ _L.

Based on the result read from the input unit 20, the first calculation unit of the control computer 19 fits the discrete values of the dimensions of each pattern and the dose to obtain continuous values. Next, in the second calculation unit of the control computer 19, the proximity effect correction coefficient for each of a plurality of patterns having different area densities, for each D _{base of} fine steps, from the continuous value obtained by the first calculation unit. Then, the relationship between the pattern and the size is obtained, and then, from this relationship, the proximity effect correction coefficient that minimizes the dimensional difference between the patterns and the average dimension of the pattern when using the proximity effect correction coefficient are obtained.

  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, and an irradiation time calculation unit 31d.

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, based on the proximity effect correction dose D _p, the dose D of the electron beam is determined.

  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.

  The deflection control unit 30 in FIG. 11 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, and 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.

  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 proximity effect correction and the loading effect correction have been described in the above embodiment, the fogging effect correction can be performed together with these corrections. In this case, in addition to the small area (second area) for correcting the loading effect as the global area described above, the fogging effect correction divided in a mesh shape with a dimension of μm to mm, for example, 0.5 mm to 1 mm, is performed. A small area (third area) is defined. The second region and the third region can be regions separated by the same dimension, or can be regions separated by different dimensions. As a method of irradiating an electron beam with an irradiation amount that simultaneously corrects a dimensional variation due to a fogging effect, a proximity effect, and a loading effect, for example, the method described in Patent Document 1 can be referred to.

  When performing the fogging effect correction, the charged particle beam writing method according to the present invention applies a plurality of first patterns having different area densities to a first region that is larger than the influence range of the proximity effect and smaller than the influence range of the fogging effect. Drawing by changing the irradiation amount of the beam, measuring the dimension of the first pattern after drawing to obtain the relationship with the irradiation amount, and fitting the discrete values of the first pattern size and the irradiation amount From the continuous values obtained, the relationship between the proximity effect correction coefficient and the dimension is obtained for each of the plurality of first patterns having different area densities, and from this relationship, the proximity effect correction that minimizes the dimensional difference between the first patterns. The step of obtaining the coefficient, and the second region provided at a distance greater than the range of influence of the fogging effect from the first region, the same pattern as the first region, and the influence of the fogging effect around this pattern A second pattern composed of a uniform background pattern arranged with a width larger than the surrounding is drawn by changing the irradiation amount of the charged particle beam, and the dimension of the second pattern after drawing is measured to determine the irradiation amount. The first pattern under irradiation conditions in which proximity effect correction is good with the first pattern is obtained from the step of obtaining the relationship between the first pattern and the continuous value obtained by fitting the discrete value of the dimension and dose of the second pattern. And obtaining a fogging effect correction coefficient that minimizes the dimensional difference between the second patterns.

  In addition, when performing the fogging effect correction, the charged particle beam drawing apparatus according to the present invention includes an input unit for inputting a result of drawing a plurality of first patterns having different area densities by changing the irradiation amount of the charged particle beam, A first calculation unit that obtains a continuous value by fitting a discrete value of the dimension and dose of the first pattern, and a proximity effect for each of a plurality of first patterns having different area densities from the continuous value A relationship between the correction coefficient and the dimension is obtained, and from this relationship, a second calculation unit that obtains a proximity effect correction coefficient that minimizes the dimensional difference between the first patterns, and the charged particle beam using the proximity effect correction coefficient. An irradiation amount calculation unit for obtaining an irradiation amount and a drawing unit for drawing with a charged particle beam based on the irradiation amount. The input unit receives a result of drawing a second pattern in which a region having a uniform area density for providing a fogging effect around the first pattern is added while changing the irradiation amount of the charged particle beam. In addition, a third calculation unit that obtains a continuous value by fitting a discrete value of the dimension of the second pattern and the irradiation amount, and irradiation with good proximity effect correction in the first pattern from the continuous value And a fourth calculation unit that obtains a fogging effect correction coefficient that minimizes a dimensional difference between the first pattern and the second pattern. Similar to the first calculation unit and the second calculation unit, the third calculation unit and the fourth calculation unit can be provided in the control computer. Then, the irradiation amount calculation unit calculates the irradiation amount of the charged particle beam using the proximity effect correction coefficient and the fogging effect correction coefficient.

Furthermore, the charged particle beam drawing apparatus of the present invention may have an evaluation unit that evaluates the accuracy of fitting. For example, if the difference between the fitting value and the experimental value in FIG. 3 is greater than or equal to the threshold value, the drawing process is stopped to prevent drawing with an incorrect parameter. Further, the area density for each loading correction amount is 0% from the correlation between the proximity effect correction coefficient η and the reference irradiation amount D _base with respect to the pattern dimension CD calculated in S101 and the relationship between the irradiation amount and the dimension obtained by the fitting in S204. , 50% and 100% pattern dimensions are calculated, and predicted values of proximity error and loading correction error are obtained. With this process, the drawing accuracy of the electron beam can be determined at 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 the case where another charged particle beam such as an ion beam is 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 51 Pattern to be drawn 52 Frame Area 53 Sub-deflection area 4 electron beam

Claims (5)

A first step of drawing a plurality of patterns having different area densities by changing the irradiation amount of the charged particle beam, measuring a dimension of the pattern after drawing, and obtaining a relationship with the irradiation amount;
From a continuous value obtained by fitting a discrete value of a predetermined condition of the dimension and the irradiation amount, a relationship between the proximity effect correction coefficient and the dimension is determined for each of the plurality of patterns having different area densities, the difference of the dimension between the pattern from this relationship have a second step of obtaining a proximity effect correction coefficient that minimizes,
The charged particle beam drawing method characterized in that the second step obtains the minimum proximity effect correction coefficient for each continuous dimension and an irradiation dose correlated with the proximity effect correction coefficient . A third step of obtaining an irradiation amount of the charged particle beam using the proximity effect correction coefficient;
And a fourth step of drawing with a charged particle beam on a predetermined resist based on the dose determined in the third step,
The charged particle beam drawing method according to claim 1, wherein an approximate expression corresponding to a type of the resist is used for the fitting. The charged particle beam drawing method according to claim 2 , wherein the approximate expression is a threshold model expression or a cubic polynomial. An input unit for inputting a result of drawing a plurality of patterns with different area densities by changing the irradiation amount of the charged particle beam; and
A first calculation unit that obtains a continuous value by fitting a discrete value of a predetermined condition of the dimension of the pattern and the irradiation amount;
From the continuous value, a relationship between the proximity effect correction coefficient and the dimension is obtained for each of the plurality of patterns having different area densities, and from this relationship, the proximity effect correction coefficient that minimizes the size difference between the patterns is obtained. A second calculating unit to be obtained;
An irradiation amount calculating unit for determining an irradiation amount of the charged particle beam using the proximity effect correction coefficient;
Possess a drawing unit for the drawing with a charged particle beam based on the irradiation dose,
The charged particle beam drawing apparatus, wherein the second calculation unit obtains the minimum proximity effect correction coefficient for each continuous size and an irradiation amount correlated with the proximity effect correction coefficient .   The charged particle beam drawing apparatus according to claim 4, wherein a threshold model equation or a cubic polynomial is used for the fitting.