JP2002015976A - Proximity effect correcting method and device- manufacturing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a proximity effect correcting method or the like capable of rightly correcting a proximity effect, even when a beam blur is changed locally. SOLUTION: When two triangles are viewed, along the line of a doze distribution shown in (B), since both the triangles are similar, from the ratio of respective sides, the relation Δ resize = 2 DB'min-DB(x, y)}B(x, y)/0.76 is established. In this case, Δresize is a dimensional correction quantity, DB is distribution of a backscattered electron quantity and B is distribution of the beam blur which occurs in a transfer optical system. As can be seen from this equation, it is sufficient to perform resizing of a quantity which is proportional to DB'min-DB(x, y)}B(x, y).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a minimum line width of 0.1 μm.
The present invention relates to a proximity effect correction method in electron beam transfer exposure that aims to form a high-precision and high-density pattern of less than a high throughput. In particular, the present invention relates to a proximity effect correction method capable of correctly correcting a proximity effect even when a beam blur in a transfer optical system is locally changed. Also,
The present invention relates to a device manufacturing method for performing an electron beam exposure process using such a proximity effect correction method.

[0002]

2. Description of the Related Art At present, exposure of each wafer (sensitive substrate) in lithography of a semiconductor integrated circuit is mainly performed by a so-called stepper using ultraviolet rays. Electron beam exposure has been used for drawing a mask to be mounted as a pattern master on a stepper, but has not yet been used for a mass production lithography process of a wafer. However, recently, in order to expose a highly integrated and ultra-fine pattern, it has been proposed to use electron beam transfer exposure for each wafer.

[0005] Incidentally, electron beam exposure is considered to have a drawback of low throughput, and various techniques have been developed to solve the drawback. At present, a figure portion batch exposure method called cell projection, character projection or block exposure has been put to practical use. In the figure partial batch exposure method, a repetitive circuit small pattern (about 5 μm square on a wafer) is repeatedly transferred using a mask in which a plurality of similar small patterns are formed, with one small pattern as one unit. Perform exposure. However, even in this method, the variable shaping method is used for pattern portions having no repeatability. For this reason, it is impossible to obtain a desired throughput in the mass production lithography process of the wafer.

As an electron beam transfer exposure system aiming at a much higher throughput than the figure partial batch exposure system, a mask having a circuit pattern of an entire semiconductor chip is prepared, and an electron beam is irradiated to a certain area of the mask. In addition, an electron beam reduction transfer device that reduces and transfers an image of a pattern in the irradiation range by a projection lens has been proposed.

In this type of apparatus, when the entire area of the mask is irradiated with an electron beam to transfer the pattern at once, the pattern cannot be transferred with high accuracy. In addition, it is difficult to manufacture a mask serving as an original. Therefore, a method which has been studied vigorously recently does not expose one die (chip on a wafer) or a plurality of dies at once, but has a large optical field as an optical system, but a pattern has a small area (sub-area). This is a method in which transfer exposure is performed by dividing into (field) (herein, referred to as a split transfer method). At this time, for each of the small areas, exposure is performed while correcting aberrations such as the focal point of the image of the small area formed on the surface to be exposed and the field distortion. This makes it possible to perform exposure with high resolution and accuracy over an optically wide area as compared with batch transfer of the entire die.

When exposing a sensitive substrate such as a wafer by irradiating it with an electron beam, there is a proximity effect in which the actual exposure amount of each part changes according to the pattern distribution in the vicinity due to the scattered electrons from the substrate. I do. In the proximity effect, electrons incident on the sensitive substrate surface are scattered and spread to reduce energy stored in a predetermined position, or electrons incident on an exposed portion are widely scattered to give energy to surrounding non-exposed portions. It happens by things.

Incidentally, the electron beam transfer optical system naturally has blur. Bokeh has geometric aberrations such as spherical aberration which exist as in the optical system using ordinary light.
In addition, as a phenomenon peculiar to an electron beam, blurring caused by repulsion of electrons in a beam due to Coulomb interaction is also added. In general, blur caused by geometric aberration increases as the distance from the optical axis to the beam increases. In an exposure apparatus of the division transfer system for projecting and exposing a part of a mask all at once, the blur increases as the center of a subfield (collective exposure area), which is a unit of one transfer exposure, is farther from the optical axis. In addition, even within a subfield, there is a general tendency that blur is greater at the end than at the center. Also, blur due to Coulomb interaction is
As the pattern area in the subfield increases and the amount of current increases, the current tends to increase. Such a phenomenon (proximity effect and blur) causes an unintended increase or decrease in the pattern line width, which is a problem.

As a method of correcting the above-described proximity effect, an improved GHOST method and a method of locally resizing a mask pattern have been proposed.

The improved GHOST method refers to a GHOST method for correcting and inverting an inverted pattern of a pattern to be transferred to make the background dose uniform, so that the correction exposure dose becomes zero at the maximum pattern density. It is a way to add. Here, the GHOST method can correct only the influence of backscattering. The acceleration voltage is 100k
In the case of V or more, blurring of the pattern due to the influence of forward scattering can be ignored. However, in this method, it is necessary to perform the correction exposure for all the wafers, so that there is a problem that one additional process is required.

Next, a description will be given of a resizing method for locally correcting a pattern dimension. The conventional resizing method is based on the premise that beam blur is uniform. However, in practice, when the beam blur fluctuates in the subfield or in the field of view of the entire optical system, or
When blurring fluctuates due to increase or decrease of the beam current, the line width of the pattern may not be accurately corrected.

SUMMARY OF THE INVENTION The present invention has been made in view of such a problem, and has as its object to provide a proximity effect correction method and the like that can correctly correct the proximity effect even when the beam blur is locally changed. I do.

[0012]

In order to solve the above problems, a first proximity effect correction method according to the present invention comprises: forming a pattern to be transferred onto a sensitive substrate on a mask; When projecting an electron beam patterned through the mask onto the sensitive substrate and transferring and exposing the pattern onto the sensitive substrate, the dimensions of the elements of the pattern on the mask are changed beforehand. a method of correcting the proximity effect keep giving resizing); the backscattered electron amount in the sensitive during substrate distribution D _B (x, y) is calculated, however, x, y is on the sensitive substrate surface Calculate coordinates, distribution B (x, y) of beam blur generated in the transfer optical system,
It calculates the minimum value D _B · _min of the _{D B (x, y),} [D B ·
_{min -D B (x, y)} ] × B (x, y) = calculates R,
A dimensional change proportional to the R is given to each element of the pattern. Since the beam blur of each part is calculated in advance and reflected in the resize amount, the line width accuracy of the pattern can be improved.

The second proximity effect correction method according to the present invention includes:
Form the pattern to be transferred on the substrate on the mask,
Illuminate the mask with an electron beam, and
Of the converted electron beam onto the sensitive substrate and
Transfer exposure on the sensitive substrate,
Dimensional change (resizing) for each element of the pattern on the mask
Correcting the proximity effect while giving;
Distribution D of backscattered electrons in sensitive substrate_B(X,
y), where x and y are the seats on the sensitive substrate surface
Target, beam blur distribution B (x,
y), and -D _B(X, y) · B (x, y) = R
Is calculated, and a dimensional change proportional to the R is calculated for the pattern.
It is given to each element. In this second way
Is D in the first method._B・_minIs approximating to 0
Simplifies calculations.

According to a third proximity effect correction method of the present invention, a pattern to be transferred onto a sensitive substrate is formed on a mask,
Illuminating the mask with an electron beam, projecting an electron beam patterned through the mask onto the sensitive substrate, and transferring and exposing the pattern onto the sensitive substrate; On the basis of the electron amount distribution D _B (x, y), x and y define the coordinates on the sensitive substrate surface, the dimensional change (resizing) amount of each element of the pattern on the mask, and the GHOST correction exposure amount. a method of correcting the proximity effect _{by; D B (x, y)} · maximum value D of the _B
max estimates, if the value of the estimate is less than a predetermined value, the D _B (x, y) in the calculation, based on the pattern area ratio in the small radius of the circle than the spread radius of the backscattered electrons Approximate calculation is performed.

In general, in the method of assuming the backscattered electron amount distribution with a 2 Gaussian distribution and calculating the amount of energy stored in the resist by the backscattered electrons, the calculation time increases as the pattern density increases. However, if the pattern area ratio inside a certain circle in the pattern to be transferred is obtained, the calculation time can be reduced. For example, D
_B (x, y) maximum value D _B · _max and the minimum value D _B · _min difference and the ratio between the primary electron dose D _P (dose of electron beam incident on the sensitive substrate surface) of the (D _B · _max -D of _{If B} · _min ) / D _P is 10% or less, the calculation of the energy storage amount may be approximated from the pattern area ratio in a circle having a radius smaller than the spread radius of the backscattered electrons.

According to a fourth proximity effect correction method of the present invention, a pattern to be transferred onto a sensitive substrate is formed on a mask,
Irradiating the mask with an electron beam, projecting an electron beam patterned through the mask onto the sensitive substrate, and transferring and exposing the pattern onto the sensitive substrate. Calculating the distribution D _B (x, y) of the backscattered electron amount in the sensitive substrate;
x and y are coordinates on the sensitive substrate surface, and a beam blur distribution B (x, y) generated in the transfer optical system is calculated, and D _B (x, y
and the difference between the maximum value D _B · _max and the minimum value D _B · _min of y), the ratio of the primary electron dose D _P (dose of electron beam incident on the sensitive substrate _{surface) (D B · max -D B} · min ) / D _P is calculated, and when this ratio is larger than a predetermined value, the correction exposure is performed by the GHOST method, and in other cases, the proximity effect correction is performed only by the resize method without performing the correction exposure. I do.

According to a fifth proximity effect correction method of the present invention, a pattern to be transferred onto a sensitive substrate is formed on a mask,
Irradiating the mask with an electron beam, projecting an electron beam patterned through the mask onto the sensitive substrate, and transferring and exposing the pattern onto the sensitive substrate. Calculating the distribution D _B (x, y) of the backscattered electron amount in the sensitive substrate;
x and y are coordinates on the sensitive substrate surface, and a beam blur distribution B (x, y) generated in the transfer optical system is calculated, and D _B (x, y
calculates the maximum value D _B · _max and the ratio D _B · _max / D _P of the primary electron dose D _P (dose of electron beam incident on the sensitive substrate surface) of the y), if the ratio is greater than a predetermined value Is characterized in that correction exposure is performed by the GHOST method, and in other cases, proximity effect correction is performed only by the resize method without performing the correction exposure. In the fifth method, the D _B · _min in the fourth method is calculated so approximates 0 is simplified.

In the above proximity effect correction method, mask pattern resizing for proximity effect correction is performed, and mask bias dimension correction (for example, Jpn.
J. Appl. Phys. Vol. 37 (1998) pp. 6767-6773).

A device manufacturing method according to the present invention includes a lithography step using the proximity effect correction method according to any one of the above aspects.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A proximity effect correction method according to an embodiment of the present invention will be described below with reference to the drawings.

First, an outline of a division transfer type electron beam projection exposure technique, which is one of the background arts of the present invention, will be described with reference to the drawings. FIG. 4 is a diagram showing an outline of an imaging relationship and a control system in the entire optical system of the charged particle beam projection exposure apparatus of the division transfer system. An electron gun 101 arranged at the uppermost stream of the optical system emits an electron beam downward. Electron gun 1
01, two-stage condenser lenses 102 and 103
The electron beam is converged by these condenser lenses 102 and 103 and the blanking aperture 1
The crossover CO is imaged at 07.

A rectangular opening 104 is provided below the second stage condenser lens 103. This rectangular opening (illumination beam shaping opening) 104 is a mask (reticle).
Only an illumination beam that illuminates one subfield of 110 (a pattern small area that is one unit of exposure) is passed. The image of the opening 104 is formed on the mask 110 by the lens 109.

Below the beam forming aperture 104, a blanking deflector 105 is arranged. Deflector 105
Deflects the illumination beam as needed to hit the non-opening of the blanking aperture 107 so that the beam does not hit the mask 110. Below the blanking opening 107,
An illumination beam deflector 108 is provided. The deflector 108 sequentially scans the illumination beam mainly in the horizontal direction (X direction) in FIG.
Illumination of each subfield. An illumination lens 109 is arranged below the deflector 108. The illumination lens 109 forms an image of the beam shaping aperture 104 on the mask 110.

The mask 110 actually extends in a plane perpendicular to the optical axis (XY plane), and has many subfields. On the mask 110, a pattern (chip pattern) forming one semiconductor device chip as a whole is formed. The mask 110 is mounted on a movable mask stage 111. By moving the mask 110 in the direction perpendicular to the optical axis (YX direction), each subfield on the mask spreads over a wider field of view than the illumination optical system. Can be illuminated. In the mask stage 111,
A position detector 112 using a laser interferometer is provided, so that the position of the mask stage 111 can be accurately grasped in real time.

A projection lens 115 is provided below the mask 110.
119 and a deflector 116 are provided. The electron beam passing through one subfield of the mask 110 is
An image is formed on a predetermined position on the wafer 123 by the projection lenses 115 and 119 and the deflector 116. An appropriate resist is applied on the wafer 123, the resist is given a dose of an electron beam, and the pattern on the mask is reduced and transferred onto the wafer 123.

A crossover CO is formed at a point that internally divides the mask 110 and the wafer 123 at a reduction ratio, and a contrast opening 118 is provided at the crossover position. The opening 118 blocks the electron beam scattered by the non-pattern portion of the mask 110 from reaching the wafer 123.

Above the wafer 123, the backscattered electron detector 1
22 are arranged. This backscattered electron detector 122
The amount of electrons reflected by the exposed surface of the wafer 123 and the mark on the stage is detected. For example, by scanning a mark on the wafer 123 with a beam that has passed through the mark pattern on the mask 110 and detecting reflected electrons from the mark at that time, the relative positional relationship between the masks 110 and 123 can be known.

The wafer 123 can be moved in the X and Y directions via an electrostatic chuck (not shown).
24. By synchronously scanning the mask stage 111 and the wafer stage 124 in directions opposite to each other, it is possible to sequentially expose each part in the chip pattern extending beyond the field of view of the projection optical system. Note that the wafer stage 124 is also provided with a position detector 125 similar to the mask stage 111 described above.

Each of the lenses 102, 103, 109, 1
15, 119 and each deflector 105, 108, 116
Each coil power control unit 102a, 103a, 109
a, 115a, 119a and 105a, 108a, 11
It is controlled by the controller 131 via 6a. Further, the mask stage 111 and the wafer stage 124 are also controlled by the controller 131 via the stage controllers 111a and 124a. The stage position detectors 112 and 125 send signals to the controller 131 via interfaces 112a and 125a including an amplifier and an A / D converter. Also, the backscattered electron detector 1
22 also sends a signal to the controller 131 via the same interface 122a.

The controller 131 grasps the control error of the stage position and corrects the error by the image position adjusting deflector 116. Thus, the reduced image of the subfield on the mask 110 is accurately transferred to the target position on the wafer 123. Then, the respective sub-field images are joined on the wafer 123, and the entire chip pattern on the mask is transferred onto the wafer.

Next, a proximity effect correction method according to the first embodiment of the present invention will be described. FIG. 1 is a diagram schematically showing the dose distribution energy accumulation amount on the wafer (sensitive substrate) surface in the proximity effect correction method according to the first embodiment of the present invention. (A) is a figure which shows the dose distribution in the place where the backscattering of an electron is the least, and (B) is a figure which shows the dose distribution in a general place. The vertical axis in FIG. 1 is the energy storage amount, and the energy storage amount D due to the backscattered electrons.
_The amount obtained by adding _B (x, y) and the amount of primary electron energy accumulated by electrons incident on the sensitive substrate surface is shown.

It is assumed that the dose distributions of (A) and (B) have the same shape in each chevron. However, in the dose distribution of (B), a mountain-shaped portion is placed on a certain background dose 20. (A), (B)
When the minimum value of the energy storage amount is set to zero and the maximum value of the energy storage amount is set to 1 for both the chevron portions of FIG.
Dashed lines are drawn on the lines 2, 0.5, 0.88 and 1. The part where the energy storage amount is 0 to 0.12 has a gentle slope. It is assumed that the portion of the energy storage amount of 0.12 to 0.88 can be approximated by a straight line. Energy storage amount
The portion from 88 to 1 has a gentle slope again, and the top has a gentle convex shape. (A) and (B) are each symmetrical.

(B) has a uniform dose ΔD compared to (A).
There are many only ose. Here, the minimum value of the energy storage amount of (A) and D _B · _min, when the minimum value of the energy storage amount of the _{(B) D B (x,} y), ΔDose = D B · min -D
_B (x, y).

Here, it is assumed that the pattern (A) is formed on the wafer without correction. At that time, negative resizing may be performed on the pattern in the general place of (B). For example, assuming that a line 11 having an energy storage amount of 0.5 in (A) is a threshold value during development, in FIG.
Is formed. This is resized so that a pattern having a width of 15 in the figure where the dose distribution intersects with the line having the energy storage amount of 0.5 in FIG.

Here, two triangles along the dose distribution of (B) were drawn. One is a triangle formed by the slope on the left side of the dose distribution of (B) and sides 17 and 19, and the other is the slope on the left side of the dose distribution and the sides 21 and 19 of the dose distribution of (B).
23 is a triangle formed. Width 15 and width 13 in the figure
Is the dimension correction amount Δresize, the length of the side 17 is Δressize / 2. The height of the side 19 is an interval ΔDose between the threshold line 11 and the line having the energy storage amount of 0.5 in FIG. The length of the side 21 is equal to the beam blur distribution B (x, y). The height of the side 23 is an interval between the energy storage amounts of 0.12 to 0.88, and the difference is 0.76.

Since these two triangles are similar, from the ratio of each side, 0.76: ΔDose = B (x, y): Δresize
/ 2. Therefore, by modifying the above _{equation, Δresize = 2 {D B ·} min -D B (x, y)} B (x, y) /0.76 ··· (1) However, ΔDose = D _B · _min -D _B (x, y) become. That is, resizing _{{D B · min -D B (} x,
y) An amount proportional to｝ B (x, y) may be performed.

It should be noted that, in most devices, there is a region with almost no pattern in the device pattern.
Since D _B · _min is negligibly _{small, {D B · min -D B} (x, y)} of the equation (1) B (x, y) portions of, -
It can be approximated by D _B (x, y) · B (x, y).

[0038] In this first embodiment, the distribution B (x, y) of beam blur is at 100nm, (D _B · _max -D
Assuming that _B · _min ) is about 0.1, it is substituted into the above equation (1), and Δresize = 2 × 0.1 × 100 / 0.76 = 26
nm. Under this condition, the distribution of backscattered electron amount D _B (x,
Assuming that the calculation error of y) is 10%, the proximity effect correction error is 2.6 nm, which is an allowable range. Therefore, in this case, the proximity effect correction can be performed only by the resize method without performing the correction exposure by the improved GHOST method (for example, the method disclosed in Japanese Patent Application Laid-Open No. 6-208944).

Here, in FIG. 1, when the interval of ΔDose exceeds 0.38, the threshold at the position (B) becomes lower than the line of the energy storage amount of 0.12.
In this region, the slope of the dose distribution becomes extremely small.
For this reason, even if the developing conditions fluctuate even a little, the pattern line width formed greatly fluctuates. Therefore, D _B (x,
y), the ratio D _B · _max / D _P between the maximum value D _B · _max and the primary electron dose D _P (dose of the electron beam incident on the sensitive substrate surface) is
If it exceeds 0.38, the line width error becomes large even when resizing. That is, in such a case, the proximity effect correction cannot be effectively performed only by the proximity effect correction by resizing. In such a case, it is preferable to perform correction exposure by the improved GHOST method.

[0040] as large as the value of D _B · _min can not be ignored,
When the _{(D B · max -D B ·} min) / D P exceeds 0.38, it is preferable to correct exposure with improved GHOST method. The interval of ΔDose 0.38 is changed by setting a threshold value during development and a range (blur amount) that can be approximated by the straight line. Further, regardless of the value of D _P, D _B ·
It is also possible to determine whether or not to perform the correction exposure by the improved GHOST method by considering only the value of _max .

Next, a description will be given of a proximity effect correction method according to a second embodiment of the present invention. FIG. 2 is a view showing a semiconductor device formed on a wafer (sensitive substrate) in the proximity effect correction method according to the second embodiment of the present invention. FIG. 2 shows one chip (die) 3 of the microprocessor.
1 is shown. A RAM section 33 in which patterns are densely arranged is formed in an area of the chip 31 which is separated by a thick broken line in the upper part of the drawing. Chip 3
1, a random logic section 35 in which patterns are arranged at a relatively low density is formed in a region separated by a thick dashed line below the diagram. In a region other than these on the chip 31, a peripheral circuit portion 37 in which patterns are arranged at a low density is formed. The periphery of the chip 31 is a gap portion 39 between the chip 31 and an adjacent chip. In the center of the figure, the field of view (stripe)
Is indicated by a dashed line.

At four places on the chip 31, circles 41, 4
2, 43 and 44 are shown. Circle 41 is chip 3
1 is centered in the upper left corner. The circle 42 is the RAM unit 3
The center is in the upper left corner of the figure in the dashed area separating 3. The circle 43 is located at a position near the center of the RAM 33 where patterns are arranged at high density. The circle 44 is located at a position near the center of the random logic section 35 where patterns are arranged at a relatively low density. Circles 41, 42, 43, 44
Are circles each having a radius of 25 μm. The area ratio of the pattern inside the circle is calculated, and the amount of energy stored in the resist due to backscattering near the center of the circle is estimated.

Since the center of the circle 41 is at the corner of the chip 31, 3/4 of the circle is the gap portion 39 in which no pattern is formed. Further, a region with a circle 41 is a peripheral circuit portion 37 in which a pattern is formed at a low density. Therefore, the amount of stored energy here is four circles 41 to 4
Since most reduced in 4, the minimum value D _B · _min energy storage amount. The circle 43 is located in the RAM unit 33 where patterns are arranged at high density, and the amount of energy stored here is the largest among the four circles 41 to 44.
The maximum value D _B · _max energy storage amount.

In FIG. 2, lines 51 and 53 are blurred contour lines, respectively. The outside of the broken line 53 is a region where the blur is 80 nm or more, and the inside of the solid line 51 is a region where the blur is 70 nm or less. That is, the larger the area of the solid line 51, the smaller the blur, and the higher the accuracy of pattern formation. The center of the portion surrounded by the solid line 51 is the center of each subfield.

Since the beam current is small in the peripheral circuit portion 37 where the pattern is formed at a low density, the blur is small, and the region 51 where the blur is 70 nm or less is relatively wide. However, since the beam current is large in the RAM section 33 in which the patterns are densely arranged, the blur is large, and the region 51 where the blur is 70 nm or less is relatively narrow.

Further, the center line 55 of the field of view (stripe)
In the vicinity, since the blur is small, the area of the contour line 51 is relatively large, but at the end of the stripe far from the center line 55,
Since the blur is large, the area of the contour line 51 is relatively narrow. In other words, the contour lines 51 and 53 are calculated by calculating the beam current from the pattern density, taking into account the Coulomb effect at that value, and comprehensively taking into account the geometric optical aberration and chromatic aberration of the electron optical system. It is.

Here, the entire pattern area of FIG.
Is divided at the corner, the energy accumulation amount D _B (x, y) of the resist by backscattered electrons in each region and to calculate the blur B (x, y) of the beam at that location, {D _B · _min - D _B (x,
y) The dimensional change proportional to the value of｝ B (x, y) was defined as the dimensional change of the pattern in that region.

At this time, the minimum value D _B ·
In the circular area 41 which gives _min, smaller as D _B · _min is negligible, above _{{D B · min -D B (} x, y)} B (x,
The portion of y), can be approximated by _{-D B (x, y) ·} B (x, y).

In a place where patterns are regularly arranged as in the RAM section 33 of FIG. 2, the amount of backscattered electrons accumulated in the resist is constant everywhere except for the peripheral portion in the RAM section 33. Therefore, for example, if the area ratio of the pattern inside the circle 43 is calculated and one representative value is calculated, the remaining calculation may be omitted.

Further, before calculating the pattern area ratio,
First, the entire pattern can be resized by a dimension called a mask bias determined by experience with the line. This means resizing the pattern size uniformly, for example, every 10 nm. After that, the correction amount is calculated and the correction is performed as described above.

Next, an embodiment of a device manufacturing method using the above-described proximity effect correction method will be described. FIG. 3 shows a flow of manufacturing a microdevice (a semiconductor chip such as an IC or LSI, a liquid crystal panel, a CCD, a thin-film magnetic head, a micromachine, etc.).

In step 1 (circuit design), a circuit of a semiconductor device is designed. Step 2 (mask fabrication) forms a mask on which the designed circuit pattern is formed. At this time, beam blur due to the proximity effect or the space charge effect may be corrected by locally resizing the pattern. Step 3 (wafer manufacturing)
Then, a wafer is manufactured using a material such as silicon.

Step 4 (oxidation) oxidizes the surface of the wafer. Step 5 (CVD) forms an insulating film on the wafer surface. Step 6 (electrode formation) forms electrodes on the wafer by vapor deposition. Step 7 (ion implantation) implants ions into the wafer. In step 8 (resist processing), a photosensitive agent is applied to the wafer. In step 9 (electron beam exposure), the circuit pattern of the mask is printed and exposed on the wafer by the electron beam transfer device using the mask created in step 2. In step 10 (light exposure), the circuit pattern of the mask is printed and exposed on the wafer by an optical stepper using the light exposure mask similarly formed in step 2. In step 9, the above-described proximity effect correction method is used.

In step 11 (development), the exposed wafer is developed. In step 12 (etching), portions other than the resist image are selectively removed. Step 13
In (resist removal), the unnecessary resist after etching is removed. By repeating steps 4 to 13, multiple circuit patterns are formed on the wafer.

Step 14 (assembly) is called a post-process, and is a process of forming a semiconductor chip using the wafer produced by the above process, and includes an assembly process (dicing and bonding) and a packaging process (chip encapsulation). And the like. In step 15 (inspection), inspections such as an operation confirmation test and a durability test of the semiconductor device manufactured in step 14 are performed. Through these steps, a semiconductor device is completed and shipped (step 16).

Although the proximity effect correction method and the like according to the embodiment of the present invention have been described with reference to FIGS. 1 to 4, the present invention is not limited to this, and various changes can be made. it can.

[0057]

As is clear from the above description, according to the present invention, the proximity effect can be correctly corrected even if the beam blur is locally changed.

[Brief description of the drawings]

FIG. 1 is a diagram schematically showing a dose distribution on a wafer (sensitive substrate) in a proximity effect correction method according to a first embodiment of the present invention. (A) is a figure which shows the dose distribution in the place where the backscattering of an electron is the least, and (B) is a figure which shows the dose distribution in a general place.

FIG. 2 is a diagram showing a semiconductor device formed on a wafer (sensitive substrate) in a proximity effect correction method according to a second embodiment of the present invention.

FIG. 3 shows a flow of manufacturing a micro device (a semiconductor chip such as an IC or an LSI, a liquid crystal panel, a CCD, a thin-film magnetic head, a micromachine, or the like).

FIG. 4 is a diagram showing an outline of an image forming relationship and a control system in the entire optical system of the electron beam projection exposure apparatus of the division transfer system.

[Explanation of symbols]

11 Threshold 13, 15 Pattern Width 17, 19, 21, 23 Triangle Side 31 Chip 33 RAM 35 Random Logic 37 Peripheral Circuit 39 Gap 41, 42, 4
3,44 yen 51,53 Blurred contour line 55 Center line of stripe 101 Electron gun 102,103
Condenser lens 104 Illumination beam shaping aperture 105 Blanking deflector 107 Blanking aperture 108 Illumination beam deflector 109 Condenser lens 110 Mask 111 Mask stage 112 Mask stage position detector 115 First projection lens 116 Image position adjustment deflector 118 Contrast aperture 119 Second projection lens 122 Backscattered electron detector 123 Wafer 124 Wafer stage 125 Wafer stage position detector 131 Controller

Claims (7)

[Claims] Forming a pattern to be transferred on a sensitive substrate on a mask, irradiating the mask with an electron beam, projecting a patterned electron beam through the mask onto the sensitive substrate, A method of correcting a proximity effect by giving a dimensional change (resizing) to each element of the pattern on the mask in advance when transferring and exposing a pattern on the sensitive substrate; Distribution of electron amount D _B
(X, y) is calculated, however, x, y is calculated coordinates on the sensitive substrate surface, the distribution B of the beam blur occurring in the transfer optical system (x, y), the minimum of the D _B (x, y) calculating a value _{D B · min, [D B} · min -D B (x, y)] · B (x, y) = calculates R, dimensional changes in proportion to the R to each element of the pattern Proximity effect correction method characterized by giving. Forming a pattern to be transferred onto a sensitive substrate on a mask, irradiating the mask with an electron beam, projecting a patterned electron beam through the mask onto the sensitive substrate, A method of correcting a proximity effect by giving a dimensional change (resizing) to each element of the pattern on the mask in advance when transferring and exposing a pattern on the sensitive substrate; Distribution of electron amount D _B
(X, y) is calculated, however, x, y is calculated coordinates on the sensitive substrate surface, the distribution B of the beam blur occurring in the transfer optical system (x, y), -D _B (x, y) · B A proximity effect correction method, wherein (x, y) = R is calculated, and a dimensional change proportional to the R is given to each element of the pattern. 3. A pattern to be transferred onto a sensitive substrate is formed on a mask, the mask is illuminated with an electron beam, and a patterned electron beam passing through the mask is projected onto the sensitive substrate to project the electron beam onto the sensitive substrate. When transferring and exposing the pattern onto the sensitive substrate, the distribution D of the backscattered electron amount in the sensitive substrate
_B (x, y), where x and y are the coordinates on the sensitive substrate surface, the dimensional change (resizing) amount of each element of the pattern on the mask, and the GHOST correction exposure amount to determine the proximity effect. Correction method; D _B (x, y)
Estimate the maximum value D _B · _max of, when the value of the estimate is less than a predetermined value, the D
A proximity effect correction method, wherein, when calculating _B (x, y), an approximate calculation is performed based on a pattern area ratio within a circle having a radius smaller than the spread radius of the backscattered electrons. 4. A pattern to be transferred onto a sensitive substrate is formed on a mask, the mask is illuminated with an electron beam, and a patterned electron beam passing through the mask is projected onto the sensitive substrate to form a pattern. the pattern at the time of transferring exposing the sensitive substrate, a method of correcting the proximity effect; backscattered electron amount during the sensitive substrate distribution D _B (x,
y) is calculated, however, x, y coordinates on the sensitive substrate surface, the distribution B (x of beam blur occurring in the transfer optical system, y) is calculated, and the maximum value D _B · a D _B (x, y) calculating the difference between the _max and the minimum value D _B · _min, the primary electron dose D _P ratio (dose of electron beam incident on the sensitive substrate _{surface) (D B · max -D B} · min) / D P, A proximity effect correction method comprising: performing a correction exposure by the GHOST method when the ratio is larger than a predetermined value; and performing a proximity effect correction only by the resize method without performing the correction exposure in other cases. 5. A pattern to be transferred onto a sensitive substrate is formed on a mask, the mask is illuminated with an electron beam, and a patterned electron beam passing through the mask is projected onto the sensitive substrate. the pattern at the time of transferring exposing the sensitive substrate, a method of correcting the proximity effect; backscattered electron amount during the sensitive substrate distribution D _B (x,
y) is calculated, however, x, y coordinates on the sensitive substrate surface, the distribution B (x of beam blur occurring in the transfer optical system, y) is calculated, and the maximum value D _B · a D _B (x, y) _max and primary electron dose D
_P ratio D _B · with (dose of electron beam incident on the sensitive substrate surface)
_max / _Dp is calculated, and when this ratio is larger than a predetermined value, the correction exposure is performed by the GHOST method. In other cases, the proximity effect correction is performed only by the resize method without performing the correction exposure. Proximity effect correction method. 6. The proximity effect correction method according to claim 1, wherein mask pattern resizing for the proximity effect correction is performed, and mask bias dimension correction is performed on the entire surface of the mask. 7. A device manufacturing method, wherein a proximity effect correction method is performed by a method according to claim 1 in a lithography step using an electron beam.