JP4313691B2 - Charged particle optics - Google Patents

Description

  The present invention relates to a charged particle optical apparatus that is used in a charged particle beam apparatus such as an electron beam apparatus such as a scanning electron microscope or an ion beam apparatus such as an ion microprobe to irradiate a sample with a charged particle beam in focus.

  In charged particle optical devices such as scanning electron microscopes and transmission electron microscopes, an aberration correction device is incorporated in a charged particle optical system for the purpose of observing a high-resolution image and increasing the probe current density. As this aberration correction apparatus, a method has been proposed in which chromatic aberration is corrected by a combination of an electrostatic quadrupole and a magnetic quadrupole, and spherical aberration is corrected by a four-stage octupole. The principle is introduced in detail in Non-Patent Documents 1 to 3.

  Here, an outline of the principle of the aberration correction apparatus described above will be described with reference to FIG. In FIG. 6, an aberration correction device 10 is disposed in front of the objective lens 7. The aberration correction apparatus 10 has four stages of multipoles 1, 2, 3, and 4. In addition, a stop 8 is provided in the front stage of the aberration correction apparatus 10. Reference sign PP in the drawing indicates the main surface of the objective lens 7.

In such a configuration, the charged particle beam incident from the left side of the figure along the optical axis LO uses the four-stage multipoles 1 to 4 as electrostatic quadrupoles, and these and the objective lens 7 The charged particle beam trajectory is formed, and the charged particle beam is focused on the sample surface 20. In FIG. 6, the X-axis trajectory _Rx and the Y-direction trajectory _{Ry of} the particle beam perpendicular to the Z direction are collected on the same plane, with the optical axis LO direction in which the charged particle beam travels as the Z direction. It is drawn schematically.

  Next, a specific configuration of the multipole elements 1 to 4 will be described.

  FIG. 7 is a diagram illustrating an example of multipoles 1 to 4 using 12-pole elements.

  FIG. 7A is a diagram showing multipoles 1 to 4 configured by electrostatic twelve-pole elements.

In this case, a voltage can be independently supplied to each of the electrodes U _{1 to} U ₁₂ constituting the 12 poles. That is, in the example of this figure, the potentials supplied from the standard and diagonal dipole power supply 101, the standard and diagonal quadrupole power supply 102, the standard and diagonal hexapole power supply 103, and the standard octupole power supply 104 are used. Based on this, the amplifiers A _{1 to} A ₁₂ in the supply unit 105 supply potentials to the corresponding electrodes U _{1 to} U ₁₂ .

  In general, in the case of a dipole, the standard dipole corresponds to the X-axis deflection, and the diagonal dipole corresponds to the Y-axis deflection.

  FIG. 7B is a diagram showing multipoles 1 to 4 configured by electric field / magnetic field superposition type 12-pole elements.

In this case, coils for supplying an excitation current are respectively attached to the ₁₂ electrode / magnetic poles W _{1 to} W ₁₂ made of a magnetic material. The potential and excitation can be controlled for each of the electrode and magnetic poles W _{1 to} W ₁₂ .

The supply unit 105 supplies a potential to each of the electrode / magnetic poles W _{1 to} W ₁₂ . Further, in the example of FIG. Based on the quadrupole power supply 111 standard for, supplies an exciting current to the coil mounted on the electrode-and-pole _W 1 _{to W-12,} which amplifier _B 1 _{.about.B 12} in the supply unit 112 corresponding .

  As described above, Non-Patent Document 4 discloses a method in which electric field / magnetic field superposed 12-poles are used as multipoles and these are used as electric-field and magnetic-field dipoles to 12-poles.

  In the example of FIG. 7B, the voltage of each multipole electrode and the excitation current of the magnetic pole are configured to be controlled independently, but this is a configuration for using a 12-pole as a plurality of multipoles. If the number of multipoles to be used is limited, such as only quadrupoles and octupoles, the number of power supplies can be reduced accordingly.

  FIG. 8 is a diagram showing various multipole elements realized by using 12-pole elements. FIG. 8 shows an example of an electrode arrangement of an electrostatic multipole.

  Usually, a multipole having a function equivalent to a structure having a reference electrode in the X-axis direction is called a normal 2n pole (n = 1, 2,...), And this standard 2n pole is used as an electrode. A multipole having a function equivalent to a structure rotated by ½ of the pitch angle (= 2π / 4n = π / 2n [rad] or 90 / n [deg]) is called a skew 2n pole.

  Similarly, in the case of the magnetic field type, a multipole having a function equivalent to a structure using an electrostatic diagonal 2n pole electrode as a magnetic pole is a standard 2n pole, and an electrostatic standard 2n pole electrode is a magnetic pole. A multipole having a function equivalent to the structure is called an oblique 2n pole.

  The reason why the arrangement of the electrodes and magnetic poles of the standard multipole (or oblique multipole) is different between the electrostatic type and the magnetic type is because the direction in which the charged particles receive the force is selected on the same straight line depending on these fields. . In the following description, when it is not necessary to distinguish between these electrodes and magnetic poles, they may be called poles.

  Next, an actual operation using these multipole elements 1 to 4 will be described with reference to FIG. The standard dipole is a deflecting device in the X direction, and the oblique dipole is a deflecting device in the Y direction. These are held for axial alignment, but the details are omitted.

  First, focus adjustment will be described. This focus adjustment can be grasped as a problem of formation of a reference trajectory.

The reference trajectory is a paraxial trajectory, and the trajectory _Ry in the Y direction passes through the center of the quadrupole by the second multipole 2 by the action of the quadrupole by the first multipole 1 and the second multistage. Due to the action of the quadrupole by the pole 2, the trajectory R _x in the X direction passes through the center of the quadrupole by the third stage multipole 3, and finally by the third stage multipole 3 and the fourth stage multipole 4. A trajectory in which the particle beam is focused on the sample surface 20 by the action of the quadrupole and the objective lens 7. In practice, these mutual adjustments are necessary to achieve complete focus.

  In addition, when the image is not clear only by focus adjustment in the X and Y directions, a quadrupole potential in an oblique direction may be used.

  Next, chromatic aberration adjustment will be described.

In order to first correct chromatic aberration in such a system, the potential φ _q2 [V] of the electrostatic quadrupole of the second stage multipole 2 and the excitation J of the magnetic quadrupole so as not to change the reference trajectory. ₂ [AT] (or magnetic potential) is adjusted, and the chromatic aberration in the X direction is corrected to 0 for the entire lens system. Similarly, the potential φ _q3 [V] of the electrostatic quadrupole of the third-stage multipole 3 and the excitation J ₃ [AT] of the magnetic quadrupole are adjusted so as not to change the reference trajectory. Chromatic aberration is corrected to 0 in the Y direction as a whole of the particle optical system.

  Next, correction of secondary aperture aberration will be described.

Here, secondary aperture correction using a hexapole will be described. Although the secondary aperture aberration should not occur ideally, it actually occurs parasitically in the aberration correction apparatus 10 due to the limit of mechanical accuracy. After correcting the chromatic aberration in the X and Y directions, the secondary aperture aberration in the X direction is corrected to 0 as the entire lens system by the potential ψ _s2 [V] of the electrostatic hexapole of the second-stage multipole element 2. Then, the secondary aperture aberration in the Y direction is corrected to 0 by the potential ψ _s3 [V] of the electrostatic hexapole of the third-stage multipole element 3. Next, the secondary aperture aberration in the direction in which the X and Y directions are combined (for example, the direction of 30 ° or 60 ° with respect to the X axis) is applied to the first-stage multipole 1 and the fourth-stage multipole 4. Each electrostatic hexapole is corrected to zero.

  Next, spherical aberration correction will be described. The spherical aberration correction can be grasped as a problem of third-order aperture aberration correction.

When correcting the spherical aberration, the third-order aperture aberration in the X direction is corrected to 0 as the entire lens system by the potential ψ _O2 [V] of the electrostatic octupole of the second-stage multipole element 2, and the third-stage aberration is corrected. The third-order aperture aberration in the Y direction is corrected to 0 by the potential ψ _O3 [V] of the electrostatic octupole of the multipole element 3. Next, the third-order aperture aberration in the 45 ° direction in which the X and Y directions are combined is corrected to 0 by the electrostatic octupole of each of the first-stage multipole 1 and the fourth-stage multipole 4. In practice, iterative adjustment is required alternately.

  Next, high-order aperture aberration correction will be described. Here, fifth-order aperture aberration correction will be examined.

  In order to minimize the contribution of the fifth-order aberration, a method of correcting the fifth-order aberration with a 12-pole element (see Non-Patent Document 5), adjusting the sign and amount of the third-order aperture aberration, and adjusting the fifth-order aberration And a method for minimizing the contribution of aberration (see Non-Patent Document 3). These methods are not described in detail here.

H. Rose, Optik 33, Heft 1, 1-23 (1971) J. Zach, Optik 83, No. 1, 30-40 (1989) J. Zach and M. Haider, Nucl. Instr. And Meth. In Phys. Res. A 363, 316-325 (1995) M. Haider, W. Bernhardt and H. Rose, Optik 63, No. 1, 9-23 (1982), especially Table 1. H. Rose, Optik 34, Heft 3, 285-311 (1971)

  As a result of the above theory and experiment, for example, the conventional techniques shown in FIGS. 6 to 7 are excellent, but sufficient consideration has not always been made to aim for a microprobe. In the following, problems with the conventional method will be described.

  First, in the case where a hexapole component is generated in the aberration collector due to incomplete mechanical accuracy, not only the secondary aperture aberration correction by the hexapole is required, but also this hexapole component is It also affects fourth-order aperture aberration. Therefore, when aiming for higher resolution, the fourth-order aperture aberration, which has a large contribution, must be corrected before the fifth-order aperture aberration. That is, fourth-order aperture aberration correction using a 10-pole element is necessary.

  Second, when the hexapole component is generated, it is considered that not only the standard hexapole component but also the diagonal hexapole component exists. In the correction of the second-order aperture aberration described in the above-described prior art, the second-order aperture aberration correction in the X direction by the standard hexapole and the second-order aperture aberration correction in the Y direction by the oblique hexapole are performed. . If this oblique component is present, the spherical aberration cannot be completely reduced to zero only by the standard octupole that corrects the spherical aberration (third-order aperture aberration). That is, it is necessary to correct spherical aberration by using an oblique octupole.

  Third, if there is the second oblique hexapole component, the fourth-order aperture aberration cannot be completely reduced to zero with only the standard 10-pole element. That is, it is necessary to correct fourth-order aperture aberration using an oblique 10-pole element.

  Similarly, if there is an oblique hexapole component, the fifth-order aperture aberration cannot be completely reduced to zero only with the standard 12-pole element. That is, it is necessary to correct the aperture aberration of 5 by the oblique 12-pole element.

  It is an object of the present invention to provide a charged particle optical device that solves these problems that have not been considered in the past, realizes optimal aberration correction, and obtains a minimum probe diameter.

  In order to solve the above-mentioned problems, a charged particle optical apparatus according to the present invention focuses a charged particle beam to irradiate a sample, and (1) is arranged along the optical axis of the charged particle beam. (4) A standard octupole potential or magnetic potential is independently given to at least three of the four-stage multipoles, and an oblique octupole is given to at least two multipoles. An independent variable power source capable of supplying five or more independent octupole potentials or magnetic potentials, and (3) at least one of the four stages of multipoles is constant. Or a non-independent variable power supply capable of supplying one or more octupole potentials or magnetic potentials, which gives a standard or oblique octupole potential or magnetic potential dependent on any one of the above values or the independent variable power supply (4) 8 of the five or more independent variable power supplies By adjusting independently of each other child potential or magnetic potential, and a control means for correcting the third-order aperture aberration.

  The control means is preferably a computer that executes a series of correction procedures based on a predetermined program.

  At least two of the three or more independent standard octupole potentials or magnetic potentials are supplied to the second and third multipole elements in the middle, and the two or more independent oblique octupole potentials. Alternatively, at least two of the magnetic potentials are preferably supplied to the first and fourth multipole elements.

  The charged particle optical apparatus according to the present invention focuses the charged particle beam and irradiates the sample, and (1) a four-stage multipole element arranged along the optical axis of the charged particle beam; (2) Among the four-stage multipoles, at least three multipoles are independently given standard ten-pole potentials or magnetic potentials, and at least three multipoles are given oblique ten-pole potentials or magnetic potentials independently. And an independent variable power supply capable of supplying six or more independent 10-pole potentials or magnetic potentials, and (3) at least one of the four-stage multipoles has a constant value or the independent variable power supply A non-independent variable power supply capable of supplying one or more 10-pole potentials or magnetic potentials, which provides a standard or oblique 10-pole potential or magnetic potential dependent on any of the values; (4) 6 Up to 10 or more independent variable power supplies By adjusting independently of each other magnetic potential, it is desirable to have a control means for correcting the fourth-order aperture aberration.

  At least one of the three or more independent standard ten-pole potentials or magnetic potentials is supplied to the second stage multipole element, and at least one of the two or more independent oblique ten-pole potentials or magnetic potentials. One is preferably supplied to the third stage multipole element.

  At least three of the three or more independent standard 10-pole potentials or magnetic potentials are supplied to the first, second, and fourth multipole elements, and the three or more independent oblique 10-pole elements are supplied. Preferably, at least three of the potentials or magnetic potentials are supplied to the first, third, and fourth stage multipole elements.

  Further, the charged particle optical device according to the present invention focuses the charged particle beam and irradiates the sample, and (1) a four-stage multipole element arranged along the optical axis of the charged particle beam; (2) The standard 12-pole potential or magnetic potential is independently given to all of the four-stage multipoles, and the seven independent multi-poles are independently given oblique 12-pole potentials or magnetic potentials. An independent variable power supply capable of supplying a 12-pole potential or magnetic potential; and (3) one multipole of the four-stage multipoles depends on a constant value or any one of the independent variable power supplies. A non-independent variable power source capable of supplying a single 12-pole potential or magnetic potential, which gives a standard or oblique twelve-pole potential or magnetic potential, and (4) a 12-pole potential of the seven independent variable power sources or By adjusting the magnetic potentials independently of each other, the fifth order opening It is desirable to have a control means for correcting the aberration, a.

  The potentials or magnetic potentials of the three independent oblique 12-poles are preferably supplied to the first, second, and fourth multipole elements.

  Alternatively, the charged particle optical apparatus according to the present invention focuses the charged particle beam and irradiates the sample. (1) An electric field / magnetic field superposition type arranged along the optical axis of the charged particle beam. (4) A standard 12-pole potential is applied independently to all of the 4-stage multipoles, and the magnetic potential of the oblique 12-pole is applied to the three multipoles. An independent power supply; and (3) one of the four-stage multipole elements having a fixed value or an oblique 12-pole magnetic field value dependent on one of the independent variable power supplies. (4) adjusting the four independent 12-pole potentials and the 3 independent 12-pole magnetic potentials; And control means for correcting fifth-order aperture aberration.

  Alternatively, the charged particle optical device according to the present invention irradiates the sample with the charged particle beam focused, and (1) an electric field / magnetic field superposition type arranged along the optical axis of the charged particle beam. (4) The magnetic potential of the standard 12-pole is independently given to all of the 4-stage multipoles, and the potential of the oblique 12-poles is given to the 3 multipoles. (3) One of the four-stage multipoles has a constant value or a diagonal twelve-pole with a value dependent on any one of the independent variable power supplies. A non-independent variable power supply capable of supplying one twelve-pole potential, and (4) adjusting the four independent twelve-pole magnetic potentials and the three independent twelve-pole potentials to provide a fifth order And a control means for correcting the aperture aberration.

  In the charged particle optical device according to the present invention, the four-stage multipole element is a twelve-pole element, and at least two central stages of the four-stage twelve pole elements are of an electric field / magnetic field superposition type, And a power source for supplying at least two independent quadrupole magnetic potentials to the electric field / magnetic field superposed multipole.

  It is desirable that the control means corrects aberrations of hexapole components together.

  As described above, the charged particle optical device of the present invention has (1) a four-stage multipole, and (2) an independent variable power source capable of supplying five or more independent octupole potentials (or magnetic potentials). Standard octupole potential (or magnetic potential) is independently applied to at least three multipoles in four stages, and oblique octupole potential (or magnetic potential) is independently applied to at least two multipoles. (3) A non-independent variable power supply capable of supplying one or more constant octupole potentials (or magnetic potentials) or an octupole potential (or magnetic potential) of a value subordinate to any one of the independent variable power supplies. A standard octupole potential (or magnetic potential) or oblique octupole potential (or magnetic potential) is applied to at least one multipole out of four stages, and (4) the five or more octupoles From the change in the image obtained by adjusting the potential (or magnetic potential) independently of each other, spherical aberration (third order By providing an aberration correction device that corrects (mouth aberration), spherical aberration (third-order aperture aberration) can be substantially corrected to zero even when the standard and oblique hexapole components are corrected in the aberration correction device. As a result, when the fourth-order or higher-order aperture aberration is corrected, the remaining spherical aberration component does not deteriorate the probe diameter.

  The charged particle optical device according to the present invention includes (1) a four-stage multipole and (2) an independent variable power source capable of supplying six or more independent ten-pole potentials (or magnetic potentials). At least three multipoles of the stage are independently given standard 10-pole potentials (or magnetic potentials), and at least 3 multipoles are independently given oblique 10-pole potentials (or magnetic potentials), ( 3) It has a non-independent variable power supply capable of supplying one or more constant 10-pole potentials (or magnetic potentials) or a 10-pole potential (or magnetic potential) dependent on either of the independent variable power supplies. Then, a standard 10-pole potential (or magnetic potential) or an oblique 10-pole potential (or magnetic potential) is applied to at least one multipole of 4 stages, and (4) the 6 or more 10-pole potentials ( (Or the magnetic potential) are adjusted independently of each other, and the fourth order opening By providing an aberration correction device for correcting aberrations, the fourth-order aperture aberration can be substantially corrected to zero even when the standard and oblique hexapole components are corrected in the aberration correction device. When the next aperture aberration is corrected, the fourth-order aperture aberration no longer limits the probe diameter.

  Further, the charged particle optical device of the present invention has (1) a four-stage multipole element and (2) an independent variable power source capable of supplying seven independent twelve-pole potentials (or magnetic potentials). The standard 12-pole potential (or magnetic potential) is given to all the multipoles independently of each other, and the oblique 12-pole potential (or magnetic potential) is given independently to the three multipoles. (3) One constant A non-independent variable power source capable of supplying a 12-pole potential (or magnetic potential) of a value of 12 or a 12-pole potential (or magnetic potential) of a value subordinate to any of the independent variable power sources, and one of four stages From the change in the image obtained by applying the oblique 12-pole potential (or magnetic potential) to each multipole, and (4) adjusting the 7 12-pole potentials (or magnetic potentials) independently of each other, 5 Since the aberration correction device for correcting the next aperture aberration is provided, the standard and oblique hexapole components can be used in the aberration correction device. Even if positive is like, substantially able to correct the zero fifth-order aperture aberration, the aperture aberration up to fifth order all be 0, it was significantly improved probe diameter.

Furthermore, when the four-stage multipole is configured by a standard array of 12 poles of electric field superposition type, a standard 12 pole field is created by the potential, and an oblique 12 pole field is created by the magnetic potential. The standard 12-pole potential of the poles is given independently, and the magnetic potential of the oblique 12-poles is given independently to the three multipoles, and these four independent potentials and three or more independent magnetic potentials are adjusted. By providing an aberration correction device that corrects the fifth-order aperture aberration based on the change in the obtained image, correction up to the fifth-order by the 12-pole can be performed. Furthermore, correction with an increased number of poles is no longer necessary, and low-cost and high-performance correction is possible.
Alternatively, when a 4-stage multipole is composed of an oblique arrangement of 12 electric fields and magnetic fields, a standard 12-pole field is created by the magnetic potential, and an oblique 12-pole field is created by the potential. The standard 12-pole magnetic potential is given to the multipole independently, and the oblique 12-pole potential is given independently to the three multipoles, and these four independent 12-pole magnetic potentials and three independent potentials are given. By providing an aberration correction device that corrects fifth-order aperture aberration from changes in the image obtained by adjustment, correction up to the fifth order with a 12-pole element is possible, and correction with an increased number of poles is not required As a result, low-cost and high-performance correction is possible.

  As described above, according to the present invention, it is possible to provide a charged particle optical device that realizes optimal aberration correction and obtains a minimum probe diameter.

  Embodiments of a charged particle optical device according to the present invention will be described below in detail with reference to the drawings.

  FIG. 1 is a diagram showing a first embodiment of a charged particle optical device. Also in the following embodiments, it is assumed that the charged particle optical device has the same structure unless otherwise specified.

  The charged particle optical apparatus according to the first embodiment irradiates a sample by focusing a portion of a charged particle beam on a sample as a probe, and includes a diaphragm 8, a four-stage multipole element 1 along an optical axis LO. , 2, 3, 4 and an objective lens 7 for focusing the charged particle beam on the sample surface 20.

  Further, this charged particle optical device includes power sources 11 and 21 for driving the multipole 1, power sources 12 and 22 for driving the multipole 2, power sources 13 and 23 for driving the multipole 3, and power source 14 for driving the multipole 4. , 24, a power source 17 for driving the objective lens 7, an operation display unit 9 for controlling an acceleration voltage and a working distance by performing an interface with the user, and the power sources 11 to 14, 17 based on the operation of the operation display unit 9 , 21 to 24 are controlled.

  For convenience, the figure shows eight power supplies 11 to 14 and 21 to 24 for supplying potentials or magnetic potentials to the four-stage multipole elements 1 to 4 so that fifth-order aberrations can be corrected. The role of these power supplies will be described in detail later.

  The control unit 19 executes a series of procedures for correcting aberrations using built-in software based on an operation on the operation display unit 9 by an operator who is a user. Based on the result of this procedure, the power supplies 11 to 14, 17, and 21 to 24 are controlled to control the electric and magnetic fields in the multipole elements 1, 2, 3, and 4 and the objective lens 7.

  Such a control part 19 can be comprised by a computer, for example. Further, the operation display unit 9 may be integrated with a console and a display, for example, a personal computer. The control unit 19 or the control unit 19 and the operation display unit 9 correspond to the control means.

  The multipoles 1 to 4 can take any form of an electrostatic type, a magnetic field type, and an electric field / magnetic field superposition type if chromatic aberration correction or fifth-order aperture aberration correction including an oblique multipole component is not taken into consideration. In the case of the magnetic field type, by the current supplied from the power sources 21 to 24, in the case of the electrostatic type, by the potential supplied from the power sources 11 to 14, or in the case of the electric field / magnetic field superposition type, the power sources 11 to 14, 21 to 21 It is controlled by the potential and current supplied from 24.

  In the following, the four-stage multipole elements 1, 2, 3, and 4 and the power supplies 11 to 14 and 21 to 24 included therein will be referred to as an aberration correction device.

  Such an aberration correction apparatus 10 is incorporated in a scanning electron microscope, for example, as shown in FIG. In this case, electrons correspond to charged particles.

  An electron beam is generated in the lens barrel 50 whose inside is in a vacuum atmosphere, the electron gun 51 which gives energy to electrons by an acceleration voltage, the electron beam generated by the electron gun 51 is converged, and an electron beam current is appropriately set. The condenser lens 52 and the objective diaphragm 53 for limiting to a certain value, the aberration correction device 54 (corresponding to the aberration correction device 10 in FIG. 1), and the deflection for scanning the sample 59 by deflecting the electron beam two-dimensionally. 55, an objective lens 57 for focusing the electron beam and irradiating the sample 59, a stage 58 for holding the sample 59 and freely driving the sample 59 in at least the XY plane, and accompanying the irradiation and scanning of the electron beam A detector 60 for detecting a signal such as secondary electrons generated from the sample 59 is provided.

  FIG. 3 is a diagram illustrating a configuration of power supplies 11 to 14 used for driving the multipole element according to the present embodiment.

  The electrostatic multipole will be described below, but the following description also applies to the electric field / magnetic field superposition type.

  Each of the power supplies 11 to 14 includes a power supply unit 31 that generates a potential and a supply unit 32 that supplies a potential to the multipole elements 1 to 4 based on the power supply unit 31.

The power supply unit 31 includes a standard dipole power supply 31 ₁ , an oblique dipole power supply 31 ₂ , a standard quadrupole power supply 31 ₃ , an oblique quadrupole power supply 31 ₄ , a standard hexapole power supply 31 ₅ , an oblique hexapole power supply 31 ₆ , and a standard octupole. power 31 _7, oblique octupole power supply 31 _8, standard 10-pole power supply 31 _9, oblique 10-pole power _{31 10} has a standard 12-pole power _{31 11,} and the oblique 12-pole power _{31 12.}

The supply unit 32 includes twelve amplifiers A _{1 to} A ₁₂ corresponding to the 12-pole elements of the multipole elements 1 to 4. Each of these amplifiers A _{1 to} A ₁₂ supplies potentials to 12 poles based on the potentials of the power supplies 31 _{1 to} 31 ₁₂ of the power supply unit 31.

That is, by adding and supplying by the amplifiers A _{1 to} A ₁₂ based on the potentials of the power supplies 31 _{1 to} 31 ₁₂ of the power supply unit 31, aperture aberration correction up to the fifth order by each multipole potential is independently performed. Yes. Although not shown in the figure, the magnetic field type can be similarly configured.

  Next, the potential (or magnetic potential) and potential distribution (or magnetic potential distribution) of the electrodes (or magnetic poles) of the multipole elements 1 to 4 will be described.

When the radius from the center of the multipoles 1 to 4 is r, the optical axis is the z axis, and the rotation angle from the x axis around the z axis is θ, the potential of the 2n pole at the coordinates (r, θ, z) and The distribution of magnetic potential is
Standard 2n pole potential:
Ψ _N2n (r, θ, z) = V _N2n ψ _2n (z) (r / a) ⁿ cos (nθ)
Diagonal 2n-pole potential:
Ψ _S2n (r, θ, z) = V _S2n ψ _2n (z) (r / a) ⁿ sin (nθ)
Standard 2n pole magnetic potential:
Φ _N2n (r, θ, z) = μ ₀ I _N2n φ _2n (z) (r / a) ⁿ sin (nθ)
Magnetic potential of diagonal 2n poles:
Φ _S2n (r, θ, z) = − μ ₀ I _S2n φ _2n (z) (r / a) ⁿ cos (nθ)
Given in.

Here, V _N2n and V _S2n are potentials applied to the electrodes of the standard 2n multipole and the oblique 2n pole, and I _N2n and I _S2n are coil currents of the magnetic poles of the standard 2n multipole and the oblique 2n pole (more precisely, current and number of turns) ), Ψ _2n (z) and φ _2n (z) are functions representing the potential and magnetic potential distribution in the optical axis LO direction (z direction) of the 2n pole. Further, a in the equation represents the inner diameter of the multipole, and μ ₀ represents the magnetic permeability in vacuum.

In Tables 1 to 6 to be described later, the sign of the magnetic potential of the diagonal 2n pole is inverted in order to reduce the number of tables for easy understanding. In actual aberration calculation, the magnetic potential (magnetic potential) is often calculated as a vector potential. However, in order to make the explanation easy to understand, a scalar potential is used here.
When these expressions are expressed by orthogonal coordinates (x, y, z),
x = r cos θ
y = rsinθ
You can rewrite the above equation using.

Further, the synthesis of the potential or magnetic potential of the standard multipole and the oblique multipole corresponds to the rotation of the multipole. That is, when the standard multipole is considered as a reference, the standard multipole strength A and the oblique multipole strength B are:
A cos α + B sin α = C cos (α−δ)
Here, since A = C cos δ and B = C sin δ,
C = (A ² + B ² ) ^1/2
δ = tan ⁻¹ (B / A)
It is.

  Even in an actual apparatus, instead of specifying A and B, the strength C of the standard multipole after synthesis and the rotation angle δ from the standard position can be specified. In the present embodiment, the description will be made by designating A and B so that the relationship between the potential applied to the actual electrode or the magnetic potential applied to the actual magnetic pole becomes clear.

  Next, as a specific example, when a 12-pole is adopted for the multipoles 1 to 4, the potential (or magnetic pole) to be applied to each electrode (or magnetic pole) of the 12-pole when the amplitude of the potential (or magnetic potential) is 1. (Table 1 to Table 4).

  Table 1 is a table showing the standard multipole potential and the oblique multipole magnetic potential by the standard arrangement 12-pole.

Table 2 is a table showing the oblique multipole potential and the standard multipole magnetic potential by the standard arrangement 12-pole.

Table 3 is a table showing standard multipole potentials and oblique multipole magnetic potentials with obliquely arranged 12-pole elements. Here, θ ′ = θ + 15 [deg] or θ ′ = θ + π / 12 [rad].

Table 4 is a table showing the oblique multipole potential and the standard multipole magnetic potential due to the obliquely arranged 12-pole element.

Next, in Tables 3 and 4 using the obliquely arranged 12-poles, the potentials (or magnetic potentials) of the respective poles normalized based on the potential (or magnetic potential) of the first pole are shown in Table 5 below. Table 6 shows.

In addition, when the multipoles 1 to 4 are configured with 12-poles, the realizable 12-pole is either the standard 12-pole or the diagonal 12-pole depending on whether the electrodes (or magnetic poles) constituting the multipoles are the standard array or the diagonal array. It is limited to either. That is, in the case where it cannot be realized, the potential (or magnetic potential) to be applied to the electrode (or magnetic pole) becomes zero.

  An object of the present application is to remove the influence of the hexapole component generated in the aberration collector due to imperfectness in mechanical accuracy, etc., when correcting higher-order aperture aberrations and aiming for higher resolution. As one of the measures, an aberration collector apparatus composed of a total of eight multipole fields in which four multipole fields by standard multipoles and four multipole fields by oblique multipoles as described above are combined can be considered.

  Therefore, when the present inventor researched such an apparatus, it was found that a new problem occurs when the eight multipole fields are simply controlled independently. For example, there are multiple solutions for correcting a certain aberration, or a multipole with low sensitivity for correcting a certain aberration may result in an excessively high applied voltage or a high voltage applied. As a result, new aberrations occur. As a result of such research, the inventor of the present application reduces the number of multipoles to be independently controlled as will be described in detail below, thereby avoiding the newly generated problems as described above and facilitating the handling of the apparatus. ing.

  Hereinafter, the principle of correction in the aberration correction apparatus 10 will be described.

The inventor of the present application examined the correction potential (or correction magnetic potential or coil current) of the aberration correction apparatus 10 and the aberration coefficient in the X direction and Y direction appearing on the sample surface 20, and as a result, the unit of the correction potential (or correction magnetic potential). It was found that the aberration coefficients in the X direction and the Y direction with respect to the change ΔV are related to each other. That is, if the changes of the aberration coefficients C _nx and C _my including the rotation effect of the nth beam in the X direction and the mth beam in the Y direction are set to ΔC _nx and ΔC _my , Rate of change t _nx = ΔC _nx / (C _nx · ΔV)
t _my = ΔC _my / (C _my · ΔV)
It was found that there is a number n, m that is related to each other (not described in detail, but C _nx and C _my are in a proportional relationship). This will be briefly described below. Here, the opening angles in the X and Y directions of the beam incident on the sample are α _x and α _y . Here, it is assumed that α _x and α _y are sufficiently small values such that sin α _x = α _x and sin α _y = α _y can be regarded approximately.

  First, the concept will be described using specific numerical values shown in Table 7, taking the case of the third-order aperture aberration coefficient as an example. Table 7 gives a standard octupole field and an oblique octupole field to each of the four-stage multipole elements, and when they are changed by a certain amount, the third-order aperture aberration coefficient (there are eight cases in the third-order). The rate of change was investigated. The left four columns in the table represent the X direction, and the right four columns represent the Y direction. Also, the column in which the rate of change is zero within the error range is represented as a blank column for easy viewing. In addition, the value of each change rate due to the oblique octupole field is relatively larger than the change rate of the standard octupole field, which is the value before correction in the denominator (C · ΔV) of the change rate. This is because the aberration C is caused by the oblique hexapole component, and the resulting aberration coefficient is a small value.

Now, by examining the table in detail, the rate of change of the four aberration coefficients of the third to fourth terms α _x ² α _y ^{1 in} the ^{first to} fourth steps that give a diagonal eight-pole field in the X direction is ¹ to 4 in the Y direction. it can be seen match respectively cubic term alpha _x ³ alpha _y change rate and the error range of the four aberration coefficients of the oblique 1-4 stage ⁰ by 4 stages. Similarly, the rate of change of the four aberration coefficients of the standard 1 to 4 stage 3rd order terms α _x ¹ α _y ² giving the standard octupole field in the X direction is the standard 1 to 4 stage 3rd order of the Y direction. The rate of change of the four aberration coefficients of the term α _x ² α _y ¹ is in agreement with each other in the range of error, and is an oblique 1 to 4 step third-order term α _x ⁰ α _y ³ that gives an oblique 8-pole field in the X direction. The change rates of the four aberration coefficients in the Y direction are the same as the change rates of the four aberration coefficients of the third to fourth terms α _x ¹ α _y ² in the diagonal direction in the Y direction within the respective error ranges. I understand. This is the first finding.

Next, the rate of change of each coefficient from the standard first stage to the standard fourth stage will be compared. However, the comparison is performed with absolute values. Then, first, the rate of change of the aberration coefficient of the second-stage term α _x ³ α _y ^{0 and} the rate of change of the aberration coefficient of the third-stage term α _x ⁰ α _y ³ are larger values than the others. I understand. In other words, the second and third stages of the standard have a greater influence on the amount of change when the octupole field is changed than the first and fourth stages of the standard. Next, when the first and fourth stages are examined, it can be seen that both levels are the same. This is the second finding.

  Similarly, the change rate of each coefficient when the potential of the oblique octupole field applied to the first to fourth stage multipoles is changed by a certain amount will be compared. Again, the comparison is absolute. Then, it can be seen that the numerical values in the first and fourth stages are clearly larger than those in the second and third stages. This is the third finding.

  These findings apply in the selection of multipoles that should give an octupole field. The above is the case of the third-order aperture coefficient, but in the case of other higher-order (second-order, fourth-order, and fifth-order), although not described in detail, the same three findings are obtained for each. . Therefore, the same technique can be applied to selection of multipoles that give each multipole field (6-pole field, 10-pole field, and 12-pole field) even in other higher-order cases. Hereinafter, based on the above concept, the case of the second to fifth aperture aberration coefficients will be described.

  The case of the secondary aperture aberration coefficient is as follows. Here, the Y-direction aberration terms that are subordinate to the X-direction aberration terms are listed side by side. This is based on the first finding. The same applies to the case of the third or higher order.

X direction: α _x ² α _y ⁰ , α _x ¹ α _y ¹ , α _x ⁰ α _y ²
Y direction: α _x ² α _y ⁰ , α _x ¹ α _y ¹ , α _x ⁰ α _y ²
The change rate of the coefficient of aberration proportional to the X direction: t _1x , t _2x , t _{3x respectively}
Y direction: t _1y , t _2y , t _3y
From the first finding in the case of the secondary aperture coefficient,
t _2x = t _1y , t _3x = t _2y
Is confirmed. Accordingly, there are four independent parameters, and at least four independent power supplies are required.

  The case of the third order aperture aberration coefficient is as follows.

X direction: α _x ³ α _y ⁰ , α _x ² α _y ¹ , α _x ¹ α _y ² , α _x ⁰ α _y ³
Y direction: α _x ³ α _y ⁰ , α _x ² α _y ¹ , α _x ¹ α _y ² , α _x ⁰ α _y ³
The change rate of the coefficient of aberration proportional to the X direction: t _1x , t _2x , t _3x , t _{4x respectively}
Y direction: t _1y , t _2y , t _3y , t _4y
After all,
t _2x = t _1y , t _3x = t _2y , t _4x = t _3y
It becomes. Accordingly, there are five independent parameters, and at least five independent power supplies are required.

  The case of the fourth-order aperture aberration coefficient is as follows.

X direction: α _x ⁴ α _y ⁰ , α _x ³ α _y ¹ , α _x ² α _y ² , α _x ¹ α _y ³ , α _x ⁰ α _y ⁴
Y direction: α _x ⁴ α _y ⁰ , α _x ³ α _y ¹ , α _x ² α _y ² , α _x ¹ α _y ³ , α _x ⁰ α _y ⁴
The change rate of the coefficient of aberration proportional to the X direction: t _1x , t _2x , t _3x , t _4x , t _{5x, respectively}
Y direction: t _1y , t _2y , t _3y , t _4y , t _5y
After all,
_t2x = _t1y , _t3x = _t2y , _t4x = _t3y , _t5x = _t4y
It becomes. Therefore, the number of independent parameters is 6, and at least 6 independent power sources are required.

  The case of the fifth-order aperture aberration coefficient is as follows.

_{^{X-direction: α x 5 α y 0,}} α x 4 α y 1, α x 3 α y 2, α x 2 α y 3, α x 1 α y 4, α x 0 α y 5
Y _{^{direction: α x 5 α y 0,}} α x 4 α y 1, α x 3 α y 2, α x 2 α y 3, α x 1 α y 4, α x 0 α y 5
The change rate of the coefficient of aberration proportional to the X direction: t _1x , t _2x , t _3x , t _4x , t _5x , t _{6x, respectively}
Y _{direction: t 1y, t 2y, t} 3y, t 4y, t 5y, t 6y
After all,
t _2x = t _1y , t _3x = t _2y , t _4x = t _3y , t _5x = t _4y , t _6x = t _5y
It becomes. Accordingly, there are seven independent parameters, and at least seven independent power supplies are required.

  As mentioned above, the relationship between the aberration coefficients found by the inventors of the present application has been described. Since the appearance of aberration differs between the standard multipole element and the oblique multipole element, which aberration coefficient to focus attention on is corrected as follows. It will be described in the correction procedure. In this case, which multipole is selected is based on the second and third findings.

  Next, an actual correction procedure in the aberration correction apparatus 10 will be described.

In the following description, it is assumed that focus adjustment, chromatic aberration correction, and secondary aperture aberration correction have been completed. Let α _x and α _y be the opening angles in the x and y directions with respect to the optical axis LO of the charged particle beam incident on the sample surface.

  FIG. 4 is a diagram for explaining aberration correction by the multipole elements 1 to 4.

  First, spherical aberration (third-order aperture aberration) will be described with reference to FIG. 4A.

  In the case of correcting spherical aberration (third-order aperture aberration), the number of independent parameters can be reduced to five as described above, so that a total of eight multi-parameters including four standard multipoles and four diagonal multipoles are provided. Five of the poles can be selected. The selection criterion is to select a multipole that has a great influence on the beam position when the potential or magnetic potential of the octupole field applied to the multipole changes. According to the research results of the present inventor, the second and third standard multipoles have a great influence, and it is sufficient to select at least one of the first and fourth standard multipoles. Therefore, here, as shown in FIG. 4A, standard multipoles in the first, second, and third stages are selected. For the slanted multipole, the second and third stages are not preferable for selection because the influence is small. Therefore, the first and fourth diagonal multipoles are selected.

  A case where the objective lens 7 is a magnetic type (or a case where at least one magnetic field type beam focusing lens is included in the middle of the optical system) will be described. In this case, since the beam is rotated by the magnetic field, the effect of correction by the aberration correction device 10 cannot be distinguished from the effect of correction by the standard multipole element and the oblique multipole element unless the amount of rotation is corrected. For this reason, the control of the standard multipole and the control of the oblique multipole are not independent, but are related to each other.

Regarding the corrected potential of the octupole before correction, potentials V ₁ , V ₂ , and V ₃ as standard octupoles are applied to the _first , _second , and _third multipole elements ₁ , ₂ , and ₃ , respectively. Potentials V ₄ and V ₅ as oblique octupoles are applied to the multipole elements 1 and 4 in the fourth stage. With respect to the aberration coefficient before correction on the sample surface 20 with respect to this correction potential, an aberration coefficient proportional to α _x ³ , α _x ² α _y , α _x α _y ² , and α _y ³ that affects the position in the X direction is c _1. , C ₄ , c ₂ , c _5, and an aberration coefficient proportional to α _y ³ that affects the position in the Y direction is c ₃ . Here, among the above coefficients, c ₄ , c ₂ , and c ₅ are proportional to α _x ³ , α _x ² α _y , and α _x α _y ² that affect the position in the Y direction, respectively.

Next, ΔV ₁ , ΔV ₂ , ΔV ₃ are the potential changes of the standard octupoles of the _first , _second , and _third stages, and ΔV ₄ are the potential changes of the first and fourth oblique octupoles. , ΔV ₅ .

Here, the change of the aberration coefficient c _j (j = 1 to 5) due to the change of each ΔV _k (k = 1 to 5) is not independent between the standard multipole element and the oblique multipole element.

Therefore, the change amount ΔV _k for setting the aberration coefficient c _j to 0 is the simultaneous equation [a _jk ] [ΔV _k ] = [− c _j ].
Is required.

Here, [a _jk ] is a matrix of j rows and k columns, [ΔV _k ] is a column vector of k rows, and [−c _j ] is a column vector of j rows.

As a result, the correction potential is V _k + ΔV _k for k = 1 to 5.
It becomes. In this way, the aberration coefficient c _j before correction can be canceled by the aberration coefficient −c _j generated by the aberration correction apparatus 10.

Next, a case where the objective lens 7 is an electrostatic type (or a case where a magnetic field type beam focusing lens is not included in the middle of the optical system) will be described. In this case, the effect of the correction by the aberration correction device affects the standard multipole element and the oblique multipole element independently. Therefore, regarding the correction potential of the octupole at the present time (before correction), the potentials V ₁ , V ₂ , and V as pole quasi-octupoles are applied to the _first , _second , and third stage multipoles 1, 2, and 3. ₃ , potentials V ₄ and V ₅ as oblique octupoles are given to the first and fourth multipole elements 1 and ₄ , and the standard multipole element and the oblique multipole element are controlled independently. Then, with respect to the current aberration coefficient on the sample surface with respect to the correction potential, an aberration coefficient proportional to α _x ³ , α _x ² α _y , α _x α _y ² , and α _y ³ that affects the position in the X direction is c. ₁ , c ₄ , c ₂ , and c _5, and an aberration coefficient proportional to α _y ³ that affects the position in the Y direction is c ₃ . Here, among the above coefficients, c ₄ , c ₂ , and c ₅ are proportional to α _x ³ , α _x ² α _y , and α _x α _y ² that affect the position in the Y direction, respectively.

Next, ΔV ₁ , ΔV ₂ , and ΔV ₃ are the potential changes of the standard octupole of the first stage, the second stage, and the third stage, and the potential change amounts of the first and fourth stages of the diagonal octupole are ΔV. ₄ and ΔV ₅ .

Here, the change of the aberration coefficient c _j (j = 1 to 5) due to the change of each ΔV _k (k = 1 to 5) is independent between the standard multipole element and the oblique multipole element.
Standard multipole (j, k = 1 to 3):

Oblique multipole (j, k = 4-5):

Therefore, the change amount ΔV _k for setting the aberration coefficient c _j to 0 is the simultaneous equation standard multipole (j, k = 1 to 3): [a _jk ] [ΔV _k ] = [− c _j ]
Oblique multipole (j, k = 4-5): [a _jk ] [ΔV _k ] = [− c _j ]
Is required. As a result, the correction potential is V _k + ΔV _k for k = 1 to 5.
It becomes. In this way, the aberration coefficient c _j before correction can be canceled by the aberration coefficient −c _j generated by the aberration correction apparatus 10.

If the second-order aperture aberration is corrected, setting the aberration correction potential as the above result to each multipole will result in five third-order coefficients appearing in the calculation out of a total of eight third-order coefficients. It can be confirmed that not only the aperture aberration coefficient c _j can be canceled out to 0, but also the other three third-order aperture aberration coefficients that do not appear in the calculation become substantially 0 at the same time. That is, it can be seen that the number of independent potentials necessary for correcting the third-order aperture aberration reaches the object with a total of five potentials, three for the standard octupole and two for the oblique octupole.

For the multipole element that changes the aberration coefficient and the potential selected above, a combination is selected in which the change in the aberration coefficient becomes large with respect to a slight change in the potential. In actual correction, when it is difficult to accurately measure various aberration coefficients or the amounts of blur corresponding to these from the image, the blur direction of the image is large with respect to each of the potential changes ΔV _k described above. attention to the direction, may be repeated several times that blur the direction of the image to adjust each V _k so as to minimize. With this operation, all third-order aperture aberrations can be reduced to zero even in a system that requires aberration correction with an oblique hexapole potential. It should be noted that in order to observe or measure the blur, since it is easier to see the blur of the screen when the position of the focus is slightly shifted from the sample surface as is well known, this may be used.

  Next, a supplementary description will be given of a power source (non-independent variable power source) other than a power source (independent variable power source) that can be applied by varying the potential (or magnetic potential) independently. In the above description using FIG. 4A, the potential (magnetic potential) of the remaining standard 4-pole octupole that is not used and the second-stage and third-stage diagonal octupoles are as shown in FIG. 4A. And zero. However, a constant octupole field potential (magnetic potential) may be supplied instead of zero. Alternatively, it can be configured to be dependently linked with other independently variable octupole fields at a constant rate. The merit of doing in this way is that it can be added to correct aberrations from low order to high order so that the potential (magnetic potential) finally applied to the multipole does not become excessive. Because there is. This is because, for example, a potential (or magnetic potential) can be applied in a shared manner with other stages so that one octupole potential (or magnetic potential) does not become too high. In this sense, it is often practically advantageous to supply a constant non-zero octupole field potential (magnetic potential) or to interlock with other independently variable octupole fields. Further, as a modification of this dependent interlocking, there are cases where there is a merit even when the number of independent power supplies is increased from the above five.

  There are also octupole stages that are not suitable for applying a potential (or magnetic potential) (see Table 7). For example, in the second and third oblique multipoles of the above-mentioned oblique octupole, there is a small amount that can correct the third-order aperture aberration caused by the hexapole component even when a large potential (or magnetic potential) is applied. When complete correction is performed at these stages, high-order aberrations are significantly increased. Therefore, particularly when the stability of the power supply and noise are not a problem, a stage having a large change in aberration coefficient (a good sensitivity) with respect to a change in potential (or magnetic potential) is selected. This is the reason why the first and fourth stages are used as the oblique multipoles.

  As a result of further examination in this way, it is known that the rate of change is significant for the second and third stages of the standard octupole that should be supplied with the three potentials independently. (See Table 7). The third potential to be supplied independently may be the first level as described above, but the fourth level or the first and fourth levels may be the same potential for practical use. It is known theoretically and experimentally that there is no problem. The above-mentioned “the first stage and the fourth stage are set to the same potential” can be considered that if the first stage is considered to be independent, the fourth stage is subordinate to the first stage at a ratio of 1: 1.

  The selection of the number of independent power sources and which multipole element to use independently among the first to fourth multipole elements 1 to 4 are as follows. A similar concept can be applied to correction of aberrations.

  Next, correction of fourth-order aperture aberration will be described with reference to FIG. 4B.

  In the case of the correction of the fourth-order aperture aberration, the number of independent parameters can be reduced to six as described above. Therefore, six out of a total of eight multipoles including the four-stage standard multipole and the four-stage oblique multipole. Just select one. Similar to the explanation of the spherical aberration (third-order aperture aberration), the selection criterion is to select a multipole element that has a large influence on the beam position when the potential or magnetic position applied to the multipole element changes. Although details are omitted, according to the research result of the present inventor, as shown in FIG. 4B, the standard multipole has the first stage, the second stage, and the fourth stage, and the oblique multipole has the first stage and the third stage. Let's choose the 4th row.

  As in the case of the spherical aberration (third-order aperture aberration), first, the case where 7 is a magnetic field type objective lens will be described.

Regarding the corrected potential of the 10-pole before correction, potentials V ₁ , V ₂ , and V ₃ as pole quasi-10 poles are given to the first, second, and fourth multipoles 1, 2, and 4, respectively. Potentials V ₄ , V ₅ , and V ₆ as oblique 10-pole elements are applied to the multi-pole elements 1, 3, and 4 in the third, third, and fourth stages. Then, the aberration coefficient before correction on the sample surface with respect to the correction potential is changed to α _x ⁴ , α _x ³ α _y , α _x ² α _y ² , α _x α _y ³ , α _y ⁴ that affects the position in the X direction. The proportional aberration coefficients are c ₁ , c ₄ , c ₂ , c ₅ , and c _3, and the aberration coefficient proportional to α _y ⁴ that affects the position in the Y direction is c ₆ . Here, among the above coefficients, c ₄ , c ₂ , c ₅ , and c ₃ are α _x ⁴ , α _x ³ α _y , α _x ² α _y ² , and α _x α _y ³ that affect the position in the Y direction. Are also proportional to each.

Next, the potential changes of the standard 10 poles of the first stage, the second stage, and the fourth stage are ΔV ₁ , ΔV ₂ , and ΔV _3, and the potentials of the first stage, the third stage, and the fourth stage oblique 10 poles. The amount of change is assumed to be ΔV ₄ , ΔV ₅ , ΔV ₆ .

Here, the change of the aberration coefficient c _j (j = 1 to 6) due to the change of each ΔV _k (k = 1 to 6) is not independent between the standard multipole element and the oblique multipole element, but is related to each other. And

Therefore, the change amount ΔV _k for setting the aberration coefficient c _j to 0 is the simultaneous equation [a _jk ] [ΔV _k ] = [− c _j ].
Is required. As a result, the correction potential is V _k + ΔV _k for k = 1 to 6.
It becomes. In this way, the aberration coefficient c _j before correction can be canceled by the aberration coefficient −c _j generated by the aberration correction apparatus 10.

  Next, similarly to the spherical aberration (third-order aperture aberration), the case where the objective lens 7 is an electrostatic type will be described first.

First, regarding the correction potential of the 10-pole before correction, potentials V ₁ , V ₂ , and V ₃ as standard 10-poles are given to the first, second, and fourth multipoles. Potentials V ₄ , V ₅ , and V ₆ as oblique 10-poles are applied to the multi-poles at the fourth and fourth stages. Then, the aberration coefficient before correction on the sample surface with respect to the correction potential is changed to α _x ⁴ , α _x ³ α _y , α _x ² α _y ² , α _x α _y ³ , α _y ⁴ that affects the position in the X direction. The proportional aberration coefficients are c ₁ , c ₄ , c ₂ , c ₅ , and c _3, and the aberration coefficient proportional to α _y ⁴ that affects the position in the Y direction is c ₆ . Here, among the above coefficients, c ₄ , c ₂ , c ₅ , and c ₃ are α _x ⁴ , α _x ³ α _y , α _x ² α _y ² , and α _x α _y ³ that affect the position in the Y direction. Are also proportional to each.

Next, the potential changes of the standard 10 poles of the first stage, the second stage, and the fourth stage are ΔV ₁ , ΔV ₂ , and ΔV _3, and the potentials of the first stage, the third stage, and the fourth stage oblique 10 poles. The amount of change is assumed to be ΔV ₄ , ΔV ₅ , ΔV ₆ .

Here, the change of the aberration coefficient c _j (j = 1 to 6) due to the change of each ΔV _k (k = 1 to 6) is independent between the standard multipole element and the oblique multipole element.
Standard multipole (j, k = 1 to 3):

Oblique multipole (j, k = 4-6):

Therefore, the change amount ΔV _k for setting the aberration coefficient cj to 0 is the simultaneous equation standard multipole (j, k = 1 to 3): [a _jk ] [ΔV _k ] = [− c _j ]
Oblique multipole (j, k = 4-6): [a _jk ] [ΔV _k ] = [− c _j ]
Is required. As a result, the correction potential is V _k + ΔV _k for k = 1 to 6.
It becomes. In this way, the aberration coefficient c _j before correction can be canceled by the aberration coefficient −c _j generated by the aberration correction apparatus 10.

If the third and lower order aperture aberrations are corrected, setting the aberration correction potential in the above result to each multipole will result in six fourth-order coefficients appearing in the calculation out of a total of ten fourth-order coefficients. It is confirmed that not only the aperture aberration coefficient c _j (j = 1 to 6) can be canceled to 0, but also the four fourth-order aperture aberration coefficients that do not appear in other calculations become substantially 0 at the same time. it can. In other words, it can be seen that the number of independent potentials necessary for correcting the fourth-order aperture aberration reaches the object with a total of 6 independent potentials, 3 for the standard 10-pole element and 3 for the oblique 10-pole element.

For the multipole element that changes the aberration coefficient and the potential selected above, a combination is selected in which the change in the aberration coefficient becomes large with respect to a slight change in the potential. In particular, since the change rate due to the second-stage multipole element is large in the standard 10-pole element, and the change rate due to the third-stage multipole element is large in the oblique 10-pole element, it is preferable to include at least these two elements. In actual correction, when it is difficult to accurately measure various aberration coefficients or the amounts of blur corresponding to these from the image, the change in image blur is large with respect to each of the potential changes ΔV _k described above. attention to the direction, may be repeated several times that blur the direction of the image to adjust each V _k so as to minimize. By this operation, all fourth-order aperture aberrations can be reduced to zero even in a system that requires aberration correction by the oblique hexapole potential.

  Next, correction of fifth-order aperture aberration will be described with reference to FIGS. 4C and 1.

  In the case of correcting the fifth-order aperture aberration, the number of independent parameters can be reduced to 7 as described above, so 7 out of a total of 8 multipoles including a 4-stage standard multipole and a 4-stage oblique multipole. Just select one. Similar to the explanation of the spherical aberration (third-order aperture aberration), the selection criterion is to select a multipole element that has a large influence on the beam position when the potential or magnetic position applied to the multipole element changes. Although details are omitted, according to the research result of the present inventor, as shown in FIG. 4C, all of the standard multipole elements are 1 to 4 stages, and the oblique multipole elements are the first, second, and fourth stages except the third stage. The stage will be selected.

  First, the case where the objective lens 7 is a magnetic field type will be described.

  In the case of using a standard arrangement of 12 poles as a multipole, in order to perform fifth-order aperture correction in a system with an oblique hexapole component, the first and fourth stages are also of the electric field / magnetic field superposition type, The power supplies 21 and 24 for the first and fourth stage multipole coils are required (see FIG. 1).

In order to perform this correction, the potentials V ₁ and V ₂ as the quasi-12 poles are added to the first, second, third, and fourth multipoles of the correction potential of the 12 poles before correction. , V ₃ , and V ₄ are given, and magnetic potentials corresponding to potentials V ₅ , V ₆ , and V ₇ as oblique 12-poles are given to the first, second, and fourth stage multipole elements. Then, with respect to the aberration coefficient before correction on the sample surface with respect to this correction potential, α _x ⁵ , α _x ⁴ α _y , α _x ³ α _y ² , α _x ² α _y ³ , α _x α that affect the position in the X direction. The coefficients of aberration proportional to _y ⁴ and α _y ⁵ are c ₁ , c ₅ , c ₂ , c ₆ , c ₃ , and c _7, and the coefficient of aberration proportional to α _y ⁵ that affects the position in the Y direction is c. ₄ Here, among the coefficients, c ₅ , c ₂ , c ₆ , c ₃ , and c ₇ are α _x ⁵ , α _x ⁴ α _y , α _x ³ α _y ² , α _x that affect the position in the Y direction. It is proportional to ² α _y ³ and α _x α _y ⁴ respectively.

Next, the first 12th stage, the 2nd stage, the 4th stage, and the 4th stage are set to ΔV ₁ , ΔV ₂ , ΔV ₃ , and ΔV ₄ , respectively. The magnetic potentials correspond to the potential changes ΔV ₅ , ΔV ₆ , and ΔV ₇ of the oblique twelve poles of the eye.

Here, the change of the aberration coefficient c _j (j = 1 to 7) due to the change of each ΔV _k (k = 1 to 7) is not independent between the standard multipole element and the oblique multipole element, but is related to each other. And

Therefore, the change amount ΔV _k for setting the aberration coefficient c _j to 0 is the simultaneous equation [a _jk ] [ΔV _k ] = [− c _j ].
Is required. As a result, the correction potential is V _k + ΔV _k for k = 1 to 7.
It becomes. In this way, the aberration coefficient c _j before correction can be canceled by the aberration coefficient −c _j generated by the aberration correction apparatus 10. In addition, when the multipoles 1 to 4 that realize the oblique 12-pole exceed the 12-pole, the oblique 12-pole can be realized only by the potential without using the magnetic field type magnetic potential.

  Next, the case where the objective lens 7 is an electrostatic type will be described.

Potentials V ₁ , V ₂ , V ₃ , and V ₄ as pole quasi-12 poles are applied to the first, second, third, and fourth multipoles, the _first , _second , and fourth stages. Magnetic potentials corresponding to potentials V ₅ , V ₆ , and V ₇ as diagonal twelve poles are given to the multipole elements in the stage. Then, the current aberration coefficients in the specimen surface for the correction potential, alpha _x ⁵ affecting the X-direction ^{_{position, α x 4 α y, α}} x 3 α y 2, α x 2 α y 3, α x α y ⁴ , the coefficient of aberration proportional to α _y ⁵ is c ₁ , c ₅ , c ₂ , c ₆ , c ₃ , c _7, and the coefficient of aberration proportional to α _y ⁵ that affects the position in the Y direction is c _4. And Here, among the coefficients, c ₅ , c ₂ , c ₆ , c ₃ , and c ₇ are α _x ⁵ , α _x ⁴ α _y , α _x ³ α _y ² , α _x that affect the position in the Y direction. It is proportional to ² α _y ³ and α _x α _y ⁴ respectively.

Next, the first 12th stage, the 2nd stage, the 4th stage, and the 4th stage are set to ΔV ₁ , ΔV ₂ , ΔV ₃ , and ΔV ₄ , respectively. The magnetic potentials correspond to the potential changes ΔV ₅ , ΔV ₆ , and ΔV ₇ of the oblique twelve poles of the eye.

Here, the change in the aberration coefficient c _j (j = 1 to 7) due to the change in each ΔV _k (k = 1 to 7) is independent between the standard multipole element and the oblique multipole element.
Standard multipole (j, k = 1 to 4):

      Oblique multipole (j, k = 5-7):

Therefore, the change amount ΔVk for setting the aberration coefficient c _j to 0 is the simultaneous equation standard multipole (j, k = 1 to 3): [a _jk ] [ΔV _k ] = [− c _j ]
Oblique multipole (j, k = 5-7): [a _jk ] [ΔV _k ] = [− c _j ]
Is required. As a result, the correction potential is V _k + ΔV _k for k = 1 to 7.
It becomes. In this way, the aberration coefficient c _j before correction can be canceled by the aberration coefficient −c _j generated by the aberration correction apparatus 10.

If aperture aberrations up to the 4th order are corrected, when the aberration correction potential and magnetic position of the above result are set in each multipole, 7 out of a total of 12 5th order coefficients appear in the calculation. The fifth-order aperture aberration coefficient c _j (j = 1 to 7) can be canceled to zero, and the five fifth-order aperture aberration coefficients that do not appear in other calculations are also substantially zero at the same time. I can confirm that. That is, it can be seen that the number of independent potentials and magnetic potentials necessary for correcting the fifth-order aperture aberration reaches the purpose with a total of 7 for the standard 12-pole element and 4 for the oblique 12-pole element.

For the multipole element that changes the aberration coefficient, potential, and magnetic potential selected above, a combination is selected in which the change in the aberration coefficient becomes large with respect to a slight change in the potential or magnetic potential. In actual correction, when it is difficult to accurately measure various aberration coefficients or the amounts of blur corresponding to these from the image, the image blurs with respect to the above-described potential and magnetic potential change ΔV _k . It is sufficient to pay attention to the direction in which the change of V is large and to adjust each V _k several times so that the blur of the image in this direction is minimized. By this operation, all fifth-order aperture aberrations can be reduced to zero even in a system that requires aberration correction by an oblique hexapole potential.

  FIG. 5 is a diagram for explaining the second embodiment.

In the first embodiment, a 12-pole is selected as the multipole, and the potential (or magnetic potential) applied to this electrode (or magnetic pole) is the standard 2n-pole potential (or magnetic potential) and the diagonal 2n-pole. It has been described that the potential (or magnetic potential) is supplied by adding the amplifiers A _{1 to} A ₁₂ to n = 1 to 6.

However, this is because the concept is easy to understand. As shown in FIG. 5, the unit connected to the input of each amplifier A _i (i = 1 to 12) includes one or more analog-digital converters (DA). A signal corresponding to the sum of the potential (or magnetic potential) of each standard 2n pole and the potential (or magnetic potential) of each diagonal 2n pole may be output from the DA converter. The DA converter may be divided into two or more for coarse adjustment and fine adjustment.

  Next, a third embodiment will be described with reference to Tables 3 to 6.

  In the correction of the fifth-order aperture aberration, the example in which the standard arrangement of 12 poles is used as the 12 pole has been described. However, the same effect can be obtained by using the oblique arrangement of 12 poles.

In this case, magnetic potentials corresponding to V ₁ , V ₂ , V ₃ , V ₄ are given as standard 12-poles to the _first , _second , _third , and _fourth multipoles. Potentials V ₅ , V ₆ , and V ₇ as oblique 12-pole elements are applied to the second and fourth multipole elements, and fifth-order aperture aberrations are corrected in the same manner as the fifth-order aperture aberration correction. I can do it.

  Next, a fourth embodiment will be described.

  In the above example, the description has been made on the assumption that, among the hexapole components, the standard hexapole component and the diagonal hexapole component are mixed at the same time. However, in an actual apparatus, there may be a case where there is no oblique hexapole component among the hexapole components. Accordingly, if there is no oblique hexapole component, an oblique octupole in correcting spherical aberration (third-order aperture aberration), an oblique ten-pole in correcting fourth-order aperture aberration, and an oblique twelve-pole in correcting fifth-order aperture aberration are not required. . For example, when calculating the correction voltage from the aberration coefficient, aberration amount, image blur amount, etc., if it is known that the oblique multipole component does not exist clearly, the equation for obtaining the potential of the oblique multipole from the simultaneous equations There is no problem even if you remove, and the formula becomes simple.

  Next, a fifth embodiment will be described.

  In the example using the magnetic field type 12-pole, an example in which an independent power source is connected to each coil of the 12-pole is shown. By the way, when the standard or oblique component of the second, third, and fourth order aperture aberration correction and the fifth order aperture aberration correction is performed by an electrostatic type 12-pole element, the remaining fifth order aperture aberration is corrected. As for the magnetic field type 12-pole, the 12-pole magnetic field type is used exclusively for the generation of the 12-pole field in the 1st stage and the 4th stage 12 poles, and the 12 coils consider the polarity (adjacent polarity). Can be connected to the series so that one 12-pole power supply can be used for each 12-pole element. In addition, the magnetic field type 12-pole in the second and third stages in the center generates a quadrupole field because it simply adds the standard (or oblique) 12-pole field to the conventional quadrupole field for correcting chromatic aberration. The coil for exclusive use and the coil for exclusive use of the 12-pole are separately provided, and the coil for exclusive use of the 12-pole is connected to the series in consideration of the polarity (so that the adjacent polarities are reversed). On the other hand, a single 12-pole power source can be used. In fact, in this way, not only the number of power supplies is reduced, but according to experiments, there is a merit that generation of useless deflection fields based on subtle deviations in output of each power supply can be reduced.

  Furthermore, if the above idea is extended, correction of aperture aberrations exceeding the fifth order should be possible in the same manner. For example, in the case of a charged particle optical apparatus using a charged particle beam having a larger energy, there is a possibility that an effect by correcting aperture aberration exceeding the fifth order may be expected.

  The present invention is not limited to the form shown in FIG. 1, and an opening angle control lens is provided when an additional lens is provided between the aberration correction apparatus 10 and the objective lens 7 or when the aberration correction apparatus 10 is in front of the aberration correction apparatus 10. The present invention can also be applied to the case where a transfer lens is provided between the aberration correction apparatus 10 and the objective lens 7.

  Although the present invention has been specifically described above, the gist of the present invention is summarized as follows just in case.

  Originally, the secondary aperture aberration occurs depending on mechanical accuracy and assembly accuracy. Therefore, conventionally, the secondary aperture aberration has been made negligible by sufficiently increasing such mechanical accuracy. However, if you try to correct the aberration of the charged particle optical device to the limit, you will have to improve the mechanical accuracy and assembly accuracy to the maximum, but it is technically necessary to increase the mechanical accuracy and assembly accuracy to any extent. It is difficult both economically and economically. Therefore, if the performance of the aberration correction apparatus is to be demonstrated to a high degree, the secondary aperture aberration cannot be ignored.

  If the second-order aperture aberration cannot be ignored, naturally the second-order aperture aberration also affects the higher-order aberration, so that it must be taken into account in the higher-order aberration correction. For this purpose, a total of eight multipoles, that is, four-stage standard multipole elements and oblique multipole elements are independently controlled to find and set an optimal correction condition for aberrations. It will be necessary.

  Certainly, in principle, the large number of factors that must be adjusted independently means that the operation of the device becomes very difficult. Furthermore, there are multiple solutions for correcting a certain aberration, or the voltage actually applied to the multipole becomes an abnormally high value, or a new aberration occurs with the application of the high voltage. Such inconvenience may occur. From such a viewpoint, it is practically desired to appropriately select independent factors and to reduce the number thereof as much as possible.

  The present invention proposes which multipoles out of a total of 8 multipoles should be selected as factors that must be adjusted independently, while minimizing the number of independent factors. That is, the number of independent factors in correcting third-order aperture aberration (spherical aberration) is five, the number of independent factors in correcting fourth-order aperture aberration is six, and correction of fifth-order aperture aberration. The number of independent factors is 7 and a multipole as a specific independent factor is selected. By adjusting using these selected multipole elements, the aberration can be easily corrected, and the operational practicality of the charged particle optical device incorporating the aberration correcting device can be enhanced.

It is a figure which shows the charged particle optical apparatus of 1st Embodiment. It is a figure which shows the scanning electron microscope incorporating the aberration correction apparatus. It is a figure which shows the structure of a power supply. It is a figure explaining the aberration correction by a multipole element. It is a figure explaining 2nd Embodiment. It is a figure which shows the conventional charged particle optical apparatus. It is a figure which shows the 12 type | mold of an electrostatic type | mold with an inner radius a, and an electric field and magnetic field superimposition type | mold. It is a figure which shows the multipole electric potential implement | achieved using 12 poles.

Explanation of symbols

1, 2, 3, 4: Multipole 7: Objective lens 8: Aperture 10: Aberration correction device 20: Sample surface PP: Main surface LO of objective lens 7: Optical axis 9: Operation display unit 19: Control units 11-14 17, 21 to 24: power supply 31: power supply unit 32: supply unit

Claims (15)

In a charged particle optical device that focuses a charged particle beam and irradiates a sample,
(1) four-stage multipole elements arranged along the optical axis of the charged particle beam;
(2) Of the four-stage multipoles, at least three multipoles are independently given standard octupole potentials or magnetic potentials, and at least two multipoles are given oblique octupole potentials or magnetic potentials independently. An independent variable power supply capable of supplying five or more independent octupole potentials or magnetic potentials,
(3) At least one multipole of the four-stage multipoles has a standard value or a potential or magnetic potential of an oblique octupole of a constant value or a value subordinate to any value of the independent variable power supply. A non-independent variable power supply capable of supplying one or more octupole potentials or magnetic potentials,
(4) Control means for correcting third-order aperture aberration by independently adjusting the octupole potential or magnetic potential of the five or more independent variable power supplies;
A charged particle optical device comprising: At least two of the independent standard octupole potentials or magnetic potentials are supplied to the second and third multipoles in the middle, and at least two of the independent oblique octupole potentials or magnetic potentials. The charged particle optical device according to claim 1, wherein the first and fourth multipole elements are supplied to the charged particle optical apparatus. A standard 10-pole potential or magnetic potential is independently given to at least 3 multipoles among the 4-stage multipoles, and an oblique 10-pole potential or magnetic potential is independently given to at least 3 multipoles. By independently adjusting the independent variable power supply capable of supplying 6 or more independent 10-pole potentials or magnetic potentials and the 10-pole potential or magnetic potential of the 6 or more independent variable power supplies, the fourth-order aperture aberration can be reduced. The charged particle optical apparatus according to claim 1, further comprising a control unit for correcting the charged particle optical apparatus. Seven independent 12-pole potentials, each of which has a standard 12-pole potential or magnetic potential independently applied to all of the four-stage multipoles, and each of the three multi-poles independently has an oblique 12-pole potential or magnetic potential. Or an independent variable power source capable of supplying a magnetic potential and a control means for correcting fifth-order aperture aberration by independently adjusting the seven 12-pole potentials or magnetic potentials. Item 4. The charged particle optical device according to any one of Items 1 to 3. The 4-stage multipole is a 12-pole, and at least the central 2-stage of the 4-stage 12-pole is an electric field / magnetic field superposition type,
An independent variable power supply for supplying four independent quadrupole potentials to the four-stage twelve poles and at least two or more independent quadrupole magnetic potentials to the electric field / magnetic field superposed multipoles;
5. The charged particle optical device according to claim 1, wherein In a charged particle optical device that focuses a charged particle beam and irradiates a sample,
(1) four-stage multipole elements arranged along the optical axis of the charged particle beam;
(2) Among the four-stage multipoles, at least three multipoles are independently given standard ten-pole potentials or magnetic potentials, and at least three multipoles are given oblique ten-pole potentials or magnetic potentials independently. An independent variable power supply capable of supplying 6 or more independent 10-pole potentials or magnetic potentials,
(3) At least one multipole of the four-stage multipoles has a standard value or a potential or magnetic potential of a standard value or a slanted 10-pole value depending on any value of the independent variable power source. A non-independent variable power supply capable of supplying one or more 10-pole potentials or magnetic potentials,
(4) control means for correcting fourth-order aperture aberration by independently adjusting the ten-pole potential or magnetic potential of the six or more independent variable power supplies;
A charged particle optical device comprising: At least one of the independent standard 10-pole potential or magnetic potential is supplied to the second-stage multipole, and at least one of the independent oblique 10-pole potential or magnetic potential is supplied to the third-stage multipole. The charged particle optical device according to claim 6, wherein the charged particle optical device is supplied to a multipole element. At least three of the independent standard 10-pole potentials or magnetic potentials are supplied to the first, second, and fourth-stage multipoles, and at least the independent oblique 10-pole potentials or magnetic potentials. 7. The charged particle optical device according to claim 6, wherein three are supplied to the first, third, and fourth multipole elements. Seven independent 12-pole potentials, each of which has a standard 12-pole potential or magnetic potential independently applied to all of the four-stage multipoles, and each of the three multi-poles independently has an oblique 12-pole potential or magnetic potential. Or an independent variable power source capable of supplying a magnetic potential and a control means for correcting fifth-order aperture aberration by independently adjusting the seven 12-pole potentials or magnetic potentials. Item 9. The charged particle optical device according to any one of Items 6 to 8. The 4-stage multipole is a 12-pole, and at least the central 2-stage of the 4-stage 12-pole is an electric field / magnetic field superposition type,
An independent variable power supply for supplying four independent quadrupole potentials to the four-stage twelve poles and at least two or more independent quadrupole magnetic potentials to the electric field / magnetic field superposed multipoles;
The charged particle optical device according to claim 4, wherein In a charged particle optical device that focuses a charged particle beam and irradiates a sample,
(1) four-stage multipole elements arranged along the optical axis of the charged particle beam;
(2) The standard 12-pole potential or magnetic potential is independently given to all of the four-stage multipoles, and the seven independent multi-poles are independently given oblique 12-pole potentials or magnetic potentials. An independent variable power supply capable of supplying a 12-pole potential or magnetic potential;
(3) One of the four-stage multipoles is given a potential or magnetic potential of a slanted 12-pole having a constant value or a value dependent on any one of the independent variable power supplies. A non-independent variable power supply capable of supplying one 12-pole potential or magnetic potential;
(4) Control means for correcting fifth-order aperture aberration by independently adjusting the 12-pole potential or magnetic potential of the seven independent variable power supplies;
A charged particle optical device comprising: 12. The charged particle optical device according to claim 11, wherein the potential or magnetic potential of the independent oblique 12-pole is supplied to the first, second, and fourth multipole elements. The 4-stage multipole is a 12-pole, and at least the central 2-stage of the 4-stage 12-pole is an electric field / magnetic field superposition type,
An independent variable power supply for supplying four independent quadrupole potentials to the four-stage twelve poles and at least two or more independent quadrupole magnetic potentials to the electric field / magnetic field superposed multipoles;
13. The charged particle optical device according to claim 11, wherein the charged particle optical device comprises: In a charged particle optical apparatus that focuses a charged particle beam on a sample and irradiates the sample,
(1) A four-stage multipole using a standard arrangement of 12 poles of a superposed electric / magnetic field type arranged along the optical axis of the charged particle beam;
(2) An independent variable power supply that independently applies a standard 12-pole potential to all of the four-stage multipoles, and independently gives an oblique 12-pole magnetic potential to three multipoles;
(3) One of the four-stage multipoles is given a magnetic value of an oblique 12-pole having a constant value or a value dependent on any of the independent variable power supplies. A non-independent variable power supply capable of supplying a 12-pole magnetic potential of
(4) control means for correcting fifth-order aperture aberration by adjusting the four independent 12-pole potentials and the 3 independent 12-pole magnetic potentials;
A charged particle optical device comprising: In a charged particle optical device that focuses a charged particle beam and irradiates a sample,
(1) A four-stage multipole using an oblique electric field / magnetic field superposition type 12-pole array arranged along the optical axis of the charged particle beam;
(2) An independent variable power supply that independently gives the magnetic potential of a standard 12-pole to all of the four-stage multipoles, and independently gives the potential of the oblique 12-pole to three multipoles;
(3) One multipole element among the four-stage multipole elements is supplied with a potential of an oblique twelve-pole element having a constant value or a value dependent on any one of the independent variable power supplies. A non-independent variable power supply capable of supplying a 12-pole potential;
(4) control means for correcting fifth-order aperture aberration by adjusting the four independent 12-pole magnetic potentials and the 3 independent 12-pole potentials;
A charged particle optical device comprising:

Images