JP2001244195A - Deflection aberration correction in charged particle beam projection system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem, where there are misalignment of a main field position and an aberration dependent on a main field position other than an axial aberration when an off-axis pattern is projected in a charged particle beam projection exposure system, and a correction system composed of a deflector and a focus corrector is considered to be used for correcting the projection exposure system on these aberrations, but the correction system degrades the projection system in optimized resolution and sub-field shape reproducibility and cannot correct it sufficiently on a deflection aberration, moreover sever limitations are imposed on the position of the focusing corrector, so that it is difficult that a deflection aberration correction system is constituted. SOLUTION: Three astigmatism correctors and three focus correctors are provided and driven corresponding to the position of a main field, by which deflection aberrations can be eliminated from a charged particle beam system while secusong it high in resolution and sub-field reproducibility, and furthermore the changed particle beam system, in which strict limitations are not imposed on the position of the correctors and a controlling method for the same can be obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention generally relates to a charged particle beam projection system, and more particularly to an electron projection lithography tool having an extremely high resolution and manufacturing a semiconductor integrated circuit. It concerns a suitable device.

[0002]

2. Description of the Related Art As is generally known, as the integration density increases, the performance and functions increase, and the economics of manufacturing also improves. At present, the minimum line width of electronic components and their connections in an integrated circuit is limited by the wavelength of the electromagnetic radiation exposing a wafer coated with a lithographic resist in its manufacture.
Accordingly, attempts have been made to further reduce the minimum line width in order to increase the degree of integration using charged particle beams including electron beams.

However, in the case of the electron beam system, it is necessary to sequentially expose a region much smaller than the exposure by the electromagnetic wave. That is, one chip area is exposed by performing many exposures sequentially. The number of exposures is, of course, determined by the number of pixels in the chip and the number of pixels that can be exposed simultaneously.
For example, current integrated circuit designs have millions of pixels, but even if they are exposed by a so-called probe forming system having a shaped beam, the number of pixels that can be exposed at one time is less than one hundred. is there. Therefore, the throughput of such an apparatus is generally low even if exposure is performed sequentially at high speed.

In order to obtain commercially acceptable throughput, a wide range of electron beams is shaped by a portion of a mask or reticle (eg, a subfield) to increase the number of simultaneously exposed pixels and reduce the number of sequential exposures. Attempts have been made to form complex patterns. Such attempts include large area projection optics (Large Area Projecti
on Optics, LAPRO) is used, but accurate alignment of the subfield image is required on the resist simultaneously with exposure fidelity of the subfield pattern. And the accuracy of the alignment is less than one-tenth of the minimum line width to be reproduced, so that the electrical connections at the joints that make up the entire pattern, such as subfield boundaries, can get.

For example, it is easily understood that if an alignment error at an edge of a subfield occurs by a fraction of the conductor width, the cross-sectional area of the conductor decreases at the same rate, affecting the signal propagation time. Or diffusion of metal occurs (nucleat
ion of metal migration) leads to unexpected breakage of the chip. The causes of such an alignment error are distortion of the subfield image, alignment error of the subfield image, and a combination thereof. Limiting the deflection area of the beam and moving both the reticle and the wafer complicates maintaining image fidelity to the required level. Hold the wafer and reticle on a movable stage to achieve both high throughput and minimum deflection,
The wafer is moved at high speed synchronously in opposite directions to bring the wafer position corresponding to the reticle subfield into a predetermined position within the deflection region of the system. The electron optical system (electron optical system for which aberration correction has been performed) has been subjected to focus calibration and calibration for sequentially aligning subfields to a set position (main field position) with respect to the reticle stage and the wafer stage. I have.

An example of an electron beam projection apparatus optimized for resolution and subfield shape at a predetermined position for sequential exposure is described in US patent application Ser. No. 09 / 131,113 (corresponding Japanese application: Japanese Patent Application Laid-Open No. 2000-58450). Yes, deflection yoke, 5 dynamic correctors: 3 dynamic focus coils (focus corrector), 2 dynamic stigmators, to optimize resolution and subfield shape Are disclosed. This disclosure is not a known technique for the present invention, but is an important "related technique".

However, the beam must be appropriately deflected, and more desirably, the aberration must be corrected not only in the set main field position but also in the vicinity of the set exposure position. That is, although the moving stage has high positional accuracy, slight errors such as 20 μm on the reticle stage and 10 μm on the wafer stage are inevitable for the speed to satisfy the acceptable throughput. Because I want to. These stage errors must be corrected by the electron optics,
The subfields must be correctly aligned on the wafer with an accuracy of a few nm. It is desired that a wide range of correction by the electron optical system be performed, and thereby, a higher stage speed and a higher system throughput can be accommodated.

[0008] A wide range of correction by the electron optical system has a resolution,
It is difficult to maintain the shape and direction of the subfield. The source of this difficulty is that optimal optical performance is obtained by changing the excitation conditions of the deflection yoke and the correction system as a function of the position error of the stage to be corrected. All correction systems and deflectors affect the subfield position, which are interacting with each other. In addition, its effects and interactions are dependent on the main field deflection position. Despite these difficulties, the subfield position must be corrected, with an accuracy of 1/10000. The correction must then be calculated in real time as the stage position error is measured and the information is transmitted to the optical system.

[0009] Yet another complicating factor in achieving a wide electro-optic correction range is main field deflection distortion. LAPRO optimized for resolution and subfield shape accuracy (see JP-A-2000-58450 on the previous day)
Suffers relatively large deflection distortion. The relationship between the deflector excitation conditions and the position, M, of the deflected beam will not deviate significantly from linearity for very small deflections. However, for a large deflection, the relationship becomes non-linear due to the deflection aberration. A general way to describe the deflection position, M, is to convert the two components, Mx and My, to a complex number Mx + i
It is represented by My and is M (italic). Called the deviation of from linearity and deflection distortion, in principle, the sum of the odd function term of M, for example, a third order distortion, the fifth-order distortion is proportional to M ⁵ which is proportional to M ^3.

By describing this related art and the present invention, it is believed that the terms used in describing LAPRO aberrations can be explained. In particular, regarding aberrations, they fall into four groups. This classification is determined from the geometric relevance to the three parameters. The parameters are a: beam convergence angle, B: position with respect to the center of the subfield, and M described above. FIG. 1 shows the target plane of the system, with definitions of parameters M and B. Aberrations are classified as follows. : 1) axial aberration dependent only on a; 2) numerical aperture a
And 3) a main field aberration that depends on M and does not depend on B; 3) a subfield aberration that depends on a and B but does not depend on M; 4) a hybrid aberration that depends on a, M and B. Since there is no general term for hybrid aberrations, it is useful to partition them by dependence on M, B, a. Like M, B and a are also represented as complex numbers. That is, each, Bx + iby, noted as a _x + ia _y. In general, the aberration coefficient is a complex number, and an actual aberration vector can be obtained by multiplying the aberration coefficient by appropriate coordinates. (Or its complex conjugate, eg, the product with Mc = Mx-iMy with subscript c).

This term is used to describe the earlier US patent application Ser. No. 09 / 131,11.
3 discloses that main field astigmatism (MMa _c ), which excites two astigmatism correctors, causes one hybrid distortion aberration (MMBc) that distorts the image blur and subfield shape. Can be erased. In addition, exciting the dynamic focus coil can eliminate field curvature (aMMc) of the main field and other hybrid distortions that affect the shape and direction of the subfield. The above four aberrations depend on MM or MMc. In this description, the term "quadratic" is used to represent the product of both. These aberrations can be corrected by performing M quadratic function corrected excitation. Therefore, these four aberrations are described as “second-order main field aberrations”.

[0012] However, US Patent 09 / 131,113 does not mention main field deflection distortion MMMC,
This is a typical main field location (e.g., Mx = 2.
375mm, My = 0.125mm) may be 1-3 microns in size. This distortion results in additional position errors, which must be added to the stage position error, which causes subfield alignment difficulties. This is a phenomenon that occurs even when the amount of deflection required by the movement of the movable stage is reduced. (The above-mentioned amount of distortion represents the magnitude of the aberration vector. For simplicity, hereinafter, the magnitude will be used to characterize the aberration. Is 2
It is well known that it is a dimensional vector, generally broken down into x and y components, or into radial and azimuthal components.

Theoretically, a direct correction of the deflection distortion correction is possible, and the excitation of the deflection is performed by several well-calibrated deflectors. However, there are higher-order (third and higher) non-linear characteristics in the distortion, which must be corrected. Otherwise, the range in which the main field is arranged with sufficient beam position accuracy is greatly restricted.

Furthermore, in practice, to maintain the desired beam geometry and trajectory, a deflection correction system must be provided with a main field deflection system (these components are arranged upstream of the main part of the charged particle beam column or the whole system). (Often). As described above, the interaction between the correction deflection system for correcting the deflection aberration, the main field deflection system, and other correction systems is very large, and the calibration is very complicated.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problem, and an object of the present invention is to correct a beam position-dependent order lower than a beam position-dependent order of distortion to be corrected. The purpose of the present invention is to correct the deflection distortion of the charged particle beam device by exciting the correction system by a signal.

Another object of the present invention is to provide a simplified calibration method for correcting deflection aberration in a charged particle beam apparatus optimized for minimum resolution and faithful reproduction of subfield shapes. It is. Yet another object of the invention is to provide
An object of the present invention is to provide a charged particle beam apparatus free from main field distortion, hybrid distortion, main field astigmatism, and field curvature.

Still another object of the present invention is to provide a charged particle beam apparatus capable of easily correcting a position error of a movable stage. It is still another object of the present invention to provide a charged particle beam apparatus which can sufficiently correct the main field and hybrid distortion and field curvature in practice, although the position of the correction system is less limited. It is.

[0018]

SUMMARY OF THE INVENTION In order to achieve the above object, there is provided a method of operating a charged particle beam apparatus in which resolution and subfield shape are optimized. The method includes
Step of deflecting a charged particle beam to one of a plurality of predetermined positions, optimized resolution using three dynamic focus systems and three astigmatism correction systems for correcting third-order and higher-order aberrations And a step of correcting deflection distortion without impairing subfield fidelity reproducibility. The three astigmatism correction systems and the three dynamic focus systems are independently excited at predetermined positions to which the beam is directed, preferably along a curve-variable trajectory.

The charged particle beam apparatus according to the present invention has a deflection system for deflecting a charged particle beam to one of a plurality of predetermined positions, and an aberration correction system for simultaneously correcting deflection distortion, field curvature and astigmatism. Have. The aberration correction system has 3
There are three astigmatism correction systems and three focus systems, which are arranged along the trajectory of the charged particle beam, and a device that excites each of the astigmatism correction systems and the focus coil according to a predetermined deflection position is arranged. I have.

The means for solving the problem will be described in more detail. A first means for solving the problem is a method of operating a charged particle beam projection apparatus in which resolution and faithful reproducibility of a subfield shape are adjusted, and deflects a charged particle beam to predetermined positions, and A dynamic focus corrector and three astigmatism correctors were used to correct the deflection distortion while maintaining the optimized resolution and subfield shape fidelity. Thereby, deflection distortion and aberration are reduced without requiring strictness of the installation position of the correction member, as compared with the correction by the conventional deflector. In addition, the dependence of the correction drive amount on the position of the main field becomes lower, and the driving of the members becomes easier.

A second means for solving the problem is that, when implementing the first means, the astigmatism corrector is excited by the selected main field position. This allows
Deflection aberrations and aberrations are reduced for desired positions.

A third means for solving the problem is to correct third-order aberrations when the first or second means is implemented. As a result, the connection of the subfields is improved, and a pattern with a fine line width can be realized.

In a fourth means for solving the problem, when the third means is implemented, the fifth order aberration is further corrected. Thereby, a finer pattern can be realized by joining.

In a fifth means for solving the problem, when the fourth means is implemented, the seventh order aberration is further corrected. Thereby, a finer pattern can be realized by joining.

A sixth means for solving the problem is a method of operating a charged particle beam apparatus, in which charged particles are deflected to predetermined positions, and the resolution and subfield shape fidelity reproducibility are determined for each deflection position. The first, second and third astigmatism correctors and the first, second and third dynamic focus coils are used to correct deflection aberration, field curvature and astigmatism while maintaining It was decided to excite. This makes it possible to reduce the aberration while maintaining the resolution and the subfield shape faithful reproducibility without strictly setting the arrangement positions of the members used in the correction system. Further, the dependence of the correction drive amount on the main field becomes lower, and the correction drive becomes easier.

A seventh means for solving the problem is a sixth means.
When implementing the means, the third order aberration is corrected. As a result, the connection of the subfields is improved, and a pattern with a fine line width can be realized.

According to an eighth means for solving the problem, when the seventh means is implemented, the fifth-order aberration is further corrected. Thereby, a finer pattern can be realized by joining.

A ninth means for solving the problem is to correct even the seventh-order aberrations when the eighth means is implemented. Thereby, a finer pattern can be realized by joining.

A tenth means for solving the problem is a charged particle beam apparatus, which comprises a deflection system for deflecting the charged particle beam to a predetermined position, and a three-dimensional arrangement arranged along the path of the charged particle beam.
An aberration correction system having two astigmatism correctors and three focus coils for simultaneously correcting deflection aberration, field curvature, and astigmatism; and an astigmatism corrector and a focus coil depending on the predetermined position. And means for exciting As a result, it is possible to obtain a device capable of eliminating the deflection aberration and the main field aberration even if the position of the correction system member is not strict. Further, the order of the dependence of the drive amount on the main field position is reduced, and the correction drive is facilitated.

An eleventh means for solving the problem is that, when the tenth means is implemented, the three astigmatism correctors arranged along the path of the charged particle beam do not overlap with each other, and the three fields do not overlap. Three astigmatism correctors and three focusing coils are arranged so that the fields of the focusing coils do not overlap. Thereby, the initial performance of the members of each correction system is sufficiently exhibited, and the deflection aberration and the optical aberration are reduced.

A twelfth means for solving the problem is that, when the tenth or eleventh means is implemented, the deflection system includes a charged particle beam at a position on the reticle in order to form a charged particle beam pattern. Is deflected. By using this, a pattern to be exposed can be selectively illuminated, and a necessary pattern can be sequentially transferred to form a pattern of the entire chip.

A thirteenth means for solving the problem is that, when the twelfth means is implemented, the deflection system has a deflector for deflecting the charged particle beam to each position of the target. By using this, a pattern to be selectively exposed can be formed on the wafer, and a necessary pattern can be sequentially transferred to form a pattern of the entire chip.

A fourteenth means for solving the problem is that, when any one of the tenth to thirteenth means is implemented, the deflection system has a means for guiding a charged particle beam along a curved variable optical axis. . As a result, the pattern of the subfield located at a position distant from the mechanical center is also projected with low aberration, and an increase in the deflection width leads to an increase in throughput.

A fifteenth means for solving the problem is that, when any one of the tenth to fourteenth means is implemented, the charged particles are electrons. Thereby, accurate control of the charged particle beam by the low intensity magnetic field and the low intensity electric field becomes possible.

A sixteenth means for solving the problem is a method of manufacturing a semiconductor device using a charged particle beam projection apparatus whose resolution and subfield shape fidelity are adjusted, wherein the charged particle beam is positioned at a predetermined position. Deflection and correction of deflection aberration using three dynamic focus correctors and three astigmatism correctors while maintaining the adjusted resolution and subfield shape fidelity, thereby forming the element members of the semiconductor device. Therefore, the resist was exposed to charged particle beams to complete the semiconductor device. According to this method, aberrations are reduced, and the connection between subfields is improved, so that a chip having a fine pattern can be obtained with high throughput.

A seventeenth means for solving the problem is a method of manufacturing a semiconductor device using a charged particle beam device, comprising: a deflector for deflecting a charged particle beam to a predetermined position as a charged particle beam device; An aberration correction system having three astigmatism correctors and three focusing coils arranged along the path of the particle beam to simultaneously correct deflection aberration, curvature of field, and astigmatism; Using a charged particle beam apparatus having an astigmatism corrector and a means for exciting a focusing coil, exposing the resist with a charged particle beam to form an element member of the semiconductor element, and completing the semiconductor element. did. Thereby, deflection aberration and optical aberration are reduced, and
A semiconductor chip with good connection of subfield patterns can be obtained with high throughput.

[0037]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The objects, embodiments and effects of the present invention will be better understood by the following examples. FIG. 1 shows the target surface of the system, the main field coordinates on the target surface, M = Mx + iMy, the subfield coordinates, and B = Bx + iBy.

FIG. 2 shows a sectional view of a charged particle beam column optimized for the fidelity reproducibility and resolution of the subfield shape. FIG. 3 shows a sectional view of the charged particle beam column of the present invention. In the following description, an electron beam will be described as an example of a charged particle beam. FIG. 2 shows a reticle imaging optical system having a large area reduction projection optical system (LAPRO) and key components of an electron beam projection system in which the resolution and the fidelity reproducibility of the subfield shape are optimized. FIG. 3 is a schematic sectional view of FIG. As you can see,
FIG. 2 shows a charged particle beam column to which the present invention is to be improved, but is not prior art to the present invention. Therefore, FIG. 2 is “related art”.

Elements that are not important to the understanding of the present invention are:
The description, including the rest of the electron beam column, is omitted and simplified. Although the electron beam column portion between the reticle and the target is described as being sufficient for explanation, the portion between the electron source and the reticle is omitted. Shown in dashed lines is a schematic diagram of a coil 12 forming the deflector of the electron beam projection system, which coil is also used in the present invention, and which directs the beam on the reticle and on the target, respectively. It is excited to be directed to the desired position (although only the deflector between the reticle and the target is shown) and is preferably directed to the desired position along a curved variable optical axis.
The lenses of the system are shown schematically at 14 and 16. These coils and lenses do not take into account their size proportions and do not represent any particular electron beam projection system.

The reticle imaging optical system shown in FIG.
The collimator section C (collimator se
ction C) and a projection section P (projection section P), and have lenses 14 and 16, respectively. These two parts have lenses, such as a symmetric doublet.
oublet) or sometimes an asymmetric doublet (when the image is reversed by crossover). What this term indicates is that the magnification of the image is m, the on-axis ratio of the magnetic field distribution area on the axes of the collimator unit and the projection unit is m, the intensity ratio of the field is 1 / m, and the ratio of the resulting focal length. Is m. Reduction ratio about 4 through the axis of the system
A system with 1, 1/4 is shown.

As shown, on-axis position indices C1 to C9 and P1 to P4 are assigned to the collimator unit and the projection unit. This index is determined arbitrarily and is only convenient for describing the position on the electron beam axis z in the explanation. As a special index, the index C5 is approximately at the center of the lens 14, and the index P3 is approximately at the center of the lens 13. The indices P1 and P2 are located on the opposite side of the crossover point 22.

The spacing of the indices is also essentially arbitrary, but generally reflects the relative strength of each lens at the corresponding on-axis position. (The smaller the distance, the greater the strength). Therefore, P2 and P3
The distance between P3 and P4 or about half between C4 and C5 is due to the reduction. Therefore, these index positions are different when the arrangement of the magnetic member having the lens and the deflector is different. However,
Here, what should be kept in mind is related technology (Japanese Unexamined Patent Application Publication No.
000-58450) described the installation of the dynamic correction system component, but the position index of the component and the dynamic correction position index therein were very strict. However, there is an advantage that the strictness of the arrangement position of the dynamic correction system in the present invention is relaxed, and that little care is required.

Another shown in FIG. 2 is two two
There is a dual-axis astigmatism corrector 30, 40 and three focusing members 50, 60, 70. The illustrated positions of these element members are approximately the most preferred positions in a practical embodiment of an electron beam projection column optimized for resolution and faithful reproduction of the subfield shape.

The two lenses 14 and 16 form an image of the pattern of the reticle, typically at position 18, on a target surface 20, for example, the surface of a wafer. A magnetic deflector (eg, 24) creates a magnetic field that is superimposed on the magnetic field of the lens and, when properly excited, reduces aberrations in the system (for aberration reduction with a curvilinear optical axis, for example,
USP 5,635,719). The actual trajectory depends on the excitation currents applied to the deflector, and the ratio of those currents causes the beam trajectory to pivot about a crossover point 22, where the beam intersects the geometric axis of the system.
Therefore, any subfield of the reticle can be selectively imaged on the target. Some of the aberrations caused by this deflection and the imaging system are due to the dynamic correction members 30, 4
0, 50, 60, 70. There is a strong selection system for the relative positional relationship of these members, which is roughly shown in the figure.

More specifically, when the two astigmatism correctors are properly excited, the main field astigmatism (MMa _c ) causing image blur and the one hybrid distortion aberration (MMBc) causing subfield shape distortion are reduced. Can be countered. Further, by properly exciting the dynamic focus coil, the main field curvature (aMMc) and other hybrid distortion aberrations (MBMC) that affect the subfield shape and direction can be canceled. These four aberrations are the second-order main field aberrations described above.

As noted above, it is known and accepted by those skilled in the art that at least one correction member is required for each type of image distortion or aberration to be corrected, and Each such correction member introduces higher order distortions and aberrations into the image (although the higher order distortions and aberrations are less). 2 for the five correction members in FIG.
There are two astigmatism correctors 30, 40, which are biaxial members and each have two degrees of freedom in terms of excitation. The degree of freedom of the dynamic focusing coil is one.
Therefore, the correction system has seven degrees of freedom as a whole. The four quadratic main field aberrations are complex numbers, with the exception of the main field curvature. Therefore, there are seven degrees of freedom (3 * 2 + 1).

Although such a device configuration is effective for correcting secondary main field aberrations and canceling main field curvature of field while maintaining faithful reproducibility and resolution of the subfield shape to the maximum. As previously noted, the non-linearities associated with deflection remained uncorrected. In addition, the maximum effect cannot be obtained unless the arrangement of the five correction members is strictly aligned.

The inventor has found that the addition of a third astigmatism corrector, rather than an additional deflection correction member, cancels main field deflection distortion, thereby simplifying stage error handling. On the other hand, it is possible to prevent the subfield distortion and achieve high resolution while maintaining the correction of the secondary main field aberration. Further, it has been found that the strictness of the arrangement position of the correction member can be relaxed by the third astigmatism corrector.

FIG. 3 shows a cross section of the charged particle beam column of the present invention more clearly. This lens barrel corresponds to the lens barrel in FIG. FIG. 2 shows the main field deflector 24, the dynamic focusing coils 50, 60, 70 and the dynamic astigmatism corrector 40,
Although 50 is entered, it is important to understand that
That is, the relative position of the correction member is not so strict as compared with the case (1), and its preferable relative position will be described below. Thus, it can be seen that, at the most basic level, the present invention has an additionally driven astigmatism corrector 80, which performs deflection distortion correction while fully fulfilling the pre-addition function. It is that you are. The astigmatism corrector to be added has two degrees of freedom for correction, and enables correction of complex main field distortion. Incidentally, the main field distortion has two degrees of freedom expressed by complex quantities.

[0050]

DESCRIPTION OF THE PREFERRED EMBODIMENTS To illustrate the advantages of the third astigmatism corrector, a system optimized by five correctors having two astigmatism correctors and one astigmatism corrector are added. Comparison of the aberrations of the different systems. The calculation of aberrations for a given imager configuration is performed using commercially available software and the results are tabulated.

The following example corresponds to the preferred embodiment of the present invention, and the parameters of the electron beam column are as follows. : Object plane position (mm) -600.000 Crossover position (mm) -120.000 Image plane position (mm) 2.000 Aperture angle (mrad) 6.000 Beam acceleration voltage (Volts) 110000.000 Energy spread width (between peaks, eV) 2.000 Main field position Mx, My (mm) 2.375, 0.375 Shaped beam size (mm) 0.250, 0.250 Subfield position, Bx, By (mm) 0.125, 0.125 Two astigmatism correctors introduced as related technologies, three dynamic focus coils Has a second-order main field aberration,
In particular, main field curvature of field, main field astigmatism, two correctable hybrid distortion terms (MM
B, MBM), the following third-order aberrations were obtained by simulation of an optical system including an adjuster of a corrector for canceling out the following aberrations. (The values in the table represent the magnitudes of various aberration vectors at given main field and subfield positions, and the units are microns.) Table 1 Spherical aberration Coma Field curvature Astigmatism Deflection distortion Chromatic aberration On-axis 0.018----0.009 MF-0.007 0.000 0.000 1.029 0.005 SB-0.018 0.018 0.005 0.000 0.000 HB aMB--0.002 0.013--HB MMB-----0.000-HB MBM----0.000-HB MBB---- 0.002-where MF means main-field, SB means shaped beam, and HB means hybrid. Two astigmatism correctors, three dynamic focus coils introduced as related technologies, main field curvature of field and main field astigmatism are eliminated, similar to two correctable hybrid distortion terms (MMB, MBM) The following third-order aberrations were obtained by a simulation of an optical system including an adjuster of a corrector for minimizing main field distortion. Table 2 Spherical aberration Coma Field curvature Astigmatism Deflection distortion Chromatic aberration On-axis 0.018----0.009 MF-0.007 0.000 0.000 0.032 0.005 SB-0.018 0.018 0.005 0.000 0.000 HB aMB--0.002 0.013--HB MMB---- 0.043-HB MBM----0.032-HB MBB----0.002-Here, MF means main-field, SB means shaped beam, and HB means hybrid. As can be seen, when two astigmatism correctors and three dynamic focus coils are used, the hybrid aberrations MMB and MBM can be eliminated, but the main field distortion is caused (Table 1) and the main field distortion is reduced. As a result, the hybrid aberration increases. (Table 2).

However, when a third astigmatism device and three dynamic focusing coils are attached to this optical system and a simulation for adjusting so as to cancel the second-order main field aberration is performed, the following third-order aberration is obtained. . Table 3 Spherical aberration Coma Field curvature Astigmatism Deflection distortion Chromatic aberration On-axis 0.018----0.009 MF-0.007 0.000 0.000 0.000 0.005 SB-0.018 0.018 0.005 0.000 0.000 HB aMB--0.002 0.013--HB MMB---- 0.000-HB MBM----0.000-HB MBB----0.002-Here, MF means main-field, SB means shaped beam, and HB means hybrid. As can be seen from Table 3, three astigmatism correctors and three
The two dynamic focusing coils can be adjusted to eliminate main field distortions as well as main field aberrations. The mechanical peculiarities and relative arrangement of the compensator are such that the astigmatism compensator is placed far away from the compensator so that their fields do not overlap and the dynamic focus coil is sufficiently far from the compensator If they do not overlap with those places, the conditions are not severe. The overlap of the astigmatism corrector field and the dynamic focus coil field is allowed. When this simple condition is met,
Optimization has been made, allowing optimization over a wide range of positions and geometries of astigmatism correctors and dynamic focusing coils. However, if the position and geometrical configuration are different,
Although the excitation conditions will be different.

In order to obtain a perfect astigmatism corrector, the astigmatism corrector had a toroidal shape in the previous simulation and had the following shape. Table 4 Astigmatism corrector 1 Astigmatism corrector 2 Astigmatism corrector 3 Length 92 92 32 Inner radius 40 40 15 Outer radius 75 75 35.5 Turn angle ± 6 ° ± 6 ° ± 19 ° Number of turns 12 12 16 The on-axis position of the correction coil and the excitation current are as follows when the main field position is Mx = 2.375 mm and My = 0.375 mm. Table 5 On-axis position I0 ° I45 ° Astigmatism corrector 1 -340.5 -0.0667 -0.0745 Amps Astigmatism corrector 2 -236.5 -0.0622 -0.0695 Amps Astigmatism corrector 3 -120.0 -0.0189 -0.0211 Amps Dynamic focus 1 -418.0 46.25 Ampere turn Dynamic focus 2 -161 -1.12 Ampere turn Dynamic focus 3 -98.5 1.239 Ampere turn The dynamic focus current changes by the product of M and M, and the astigmatism corrector current is M It changes in proportion to the product of Mc. As can be seen, the corrector current is relatively low, and the demands on the current driver to produce a corrected current value corresponding to a particular main field position are not severe. The manner in which such a current is applied to the correction coil (or voltage for an electrostatic corrector) is not strict in practicing the present invention, and many suitable techniques will occur to those skilled in the art. Are known. And, as will be readily appreciated, the correction current can be easily and quickly calculated for aberrations when deflecting to a predetermined deflection position.

Further, deflection correction near a predetermined exposure position
Is corrected using one or more deflection yokes.
It's easier than As mentioned above, the deflection yoke is M^Three
Must be excited with a signal proportional to
M with correction^TwoIs excited by a signal dependent on Low order
Proportional means that the required signal change is
The compensator correction is slower. This advantage of the present invention
The point is that the two correction signals are close to the main field position.
How it changes in response to a stage error
Correction system with a deflector and a correction system with a corrector
Thinking in detail for both
You. Both systems have a defined main field position, M
n, and the necessary correction signal is
Sdnom and Scnom are used for the corrector correction. Predetermined
Slightly different from the main field position Mn +
Consider the correction signal at dM. Express necessary correction signal
To expand the series based on a predetermined value. Pair in two cases
Then, by developing the correction signal and using dM / Mn << 1, the position Mn
 And a relational expression of the correction value at Mn + dM is obtained. The numbers are
It is as follows. Correction at the deflector Sdnom (Mn) = Kd × Mn^Three  Sd (Mn + dM) = Kd × (Mn + dM)^Three ~ Sdnom × (1 + 3 × dM / M
n) Correction by astigmatism corrector Scnom = Kc × Mn^Two Sc (Mn + dM) = Kc × (Mn + dM)^Two ~ Scnom × (1 + 2 × dM / M
n) From the above equation (the symbol x represents multiplication),
Astigmatism corrector for given positional deviation from a fixed position
The correction signal at the position is more shifted than the correction signal at the deflector.
Only 2/3 of the ratio.

Although it has been described above that the correction of the second-order main field aberration and the third-order deflection distortion is performed by secondary excitation, as will be readily understood by those skilled in the art, the present invention includes distortion. It can also be applied to correction of higher-order aberrations. For example, looking at the fifth order aberration, corresponding to the third order of the main field aberrations including four M ^2, there are four fourth-order main field aberration varying M ^4. These aberrations are corrected by three astigmatism correctors and 3
It can be corrected by a system of two dynamic focus coils, which are also excited to correct the fifth order main field deflection distortion (M ⁵ ). Even though these fifth-order correctors have the same physical element as the third-order corrector (the correction current is further superimposed on the correction current described above), they are different physical elements. May be. In either case, the correction signal proportional to M ⁴ is required for fifth-order aberration correction. As with the third-order aberration, the advantage of the astigmatism corrector when comparing the fifth-order distortion correction with the astigmatism corrector and the deflector is that there is little change in the correction signal near the predetermined main field position. It is. Similarly to the third-order aberration described above, the change in the correction signal decreases at a rate of 4/5 in the case of the fifth order and at a rate of 6/7 in the case of the seventh order. It can be seen that the rate of change of the correction signal when using the present invention is 1/3 of the change of the correction signal by the deflector, and when applied to higher order aberrations, the correction system by the deflector Retains higher improvement as compared to The application of the present invention to a semiconductor manufacturing method is described in
This can be performed in the same manner as the embodiment disclosed in 58450.

[0056]

As described above, according to the present invention, one third astigmatism corrector is provided, and the deflection distortion and the second-order main field aberration are eliminated by exciting the corrector system. The calibration of the charged particle beam apparatus is simplified by reducing the change in the excitation current of the optical component for correcting the aberration near the calibration position. The correction system reduces aberrations to much smaller values, which allows for resolutions of 10 to 30 times smaller than one micron, while at the same time optimizing the subfield shape and not causing higher order aberrations.
Calibration of the position error of the moving stage is also simplified. This is because there is essentially no deflection nonlinearity. Furthermore, when obtaining the optimal solution of the resolution and the fidelity of the subfield shape, if correctors of the same type are arranged sufficiently far along the charged particle beam column and their fields do not substantially overlap each other. Therefore, the strictness of the arrangement position of the corrector is not strict.
Also, when correcting the position error of the stage, the correction can be easily performed using the correction system of the electron optical system. By applying the present invention to a semiconductor manufacturing method, a fine pattern can be manufactured with high yield, and high throughput can be realized.

While one embodiment has been used to describe the invention, it will be readily apparent to those skilled in the art that modifications may be made within the scope of the appended claims.

[Brief description of the drawings]

FIG. 1: Target plane of the system, main field coordinates on the target plane, M = Mx + iMy and subfield coordinates, B
It is a figure which shows = Bx + iBy.

FIG. 2 is a cross-sectional view of a charged particle beam column optimized for faithful reproducibility and resolution of a subfield shape.

FIG. 3 is a sectional view of the charged particle beam column of the present invention.

Claims (17)

[Claims] 1. A method of operating a charged particle beam projection apparatus in which resolution and fidelity reproducibility of a subfield shape are adjusted, wherein the charged particle beam is deflected to predetermined positions, and three dynamic focus correctors are provided. And correcting the deflection distortion while maintaining the optimized resolution and the fidelity of the subfield shape using three astigmatism correctors. 2. The operating method according to claim 1, wherein the astigmatism corrector is excited based on the selected main field position. 3. The operating method according to claim 1, wherein a third-order aberration is corrected. 4. The operating method according to claim 3, wherein the fifth order aberration is further corrected. 5. The operating method according to claim 4, wherein the seventh order aberration is further corrected. 6. A method for operating a charged particle beam apparatus, comprising: deflecting charged particles to predetermined positions; deflecting aberration and image plane while maintaining optimum resolution and subfield shape faithful reproducibility for each deflection position. To correct curvature and astigmatism, first, second and third astigmatism correctors and first, second,
An operating method characterized by exciting a third dynamic focusing coil. 7. The operating method according to claim 6, wherein the aberration includes a third-order aberration. 8. The operation method according to claim 7, wherein the aberration further includes a fifth-order aberration. 9. The operation method according to claim 8, wherein the aberration further includes a seventh-order aberration. 10. A charged particle beam apparatus, comprising: a deflection system for deflecting a charged particle beam to a predetermined position; three astigmatism correctors and three astigmatism correctors disposed along a path of the charged particle beam. Deflection aberration, field curvature, with focusing coil,
A charged particle beam apparatus comprising: an aberration correction system for simultaneously correcting astigmatism; and a unit for exciting an astigmatism corrector and a focusing coil at the predetermined position. 11. The charged particle beam apparatus according to claim 10, wherein the fields of the three astigmatism correctors arranged along the path of the charged particle beam do not overlap, and the fields of the three focusing coils do not overlap. Characterized in that three astigmatism correctors and three focusing coils are arranged so that they do not overlap. 12. The charged particle beam apparatus according to claim 10, wherein the deflection system deflects the charged particle beam to a position on a reticle in order to form a pattern of the charged particle beam. A charged particle beam device comprising a vessel. 13. The charged particle beam apparatus according to claim 12, wherein the deflection system has a deflector for deflecting the charged particle beam to each position of a target. 14. The charged particle beam apparatus according to claim 10, wherein the deflection system has a unit for guiding the charged particle beam along a curve-variable optical axis. Particle beam device. 15. The charged particle beam device according to claim 10, wherein the charged particles are electrons. 16. A method of manufacturing a semiconductor device using a charged particle beam projection apparatus having resolution and subfield shape fidelity adjusted, the method comprising: deflecting a charged particle beam to a predetermined position; Using a corrector and three astigmatism correctors, correct the deflection aberration while maintaining the adjusted resolution and subfield shape fidelity, and apply resist to the resist with charged particle beams to form the element members of the semiconductor device. A method of manufacturing a semiconductor device, comprising exposing the semiconductor device to light. 17. A method for manufacturing a semiconductor device using a charged particle beam device, comprising: a deflector for deflecting a charged particle beam to a predetermined position as a charged particle beam device; and arranging the charged particle beam along a path of the charged particle beam. Correction system that has three corrected astigmatism correctors and three focusing coils to simultaneously correct deflection aberration, curvature of field, and astigmatism, and astigmatism corrector and focus based on the respective deflection positions. Using a charged particle beam apparatus having means for exciting a coil, exposing a resist with a charged particle beam to form an element member of the semiconductor element, and completing the semiconductor element. .