Patent Application
20130068949
The invention relates to a charged particle beam device with an aberration corrector mounted therein, and in particular, to a method for adjustment of an aberration corrector in the charged particle beam device.
In a device using a focused charged particle beam (probe beam), such as a scanning electron microscope (SEM), an ion-beam machining device (FIB), and so forth, a probe is caused to scan on a sample to thereby execute processing of an observed image or the sample. In the case of these charged particle beam devices, resolution, and working accuracy are dependent on a size of a probe in section (a probe diameter), and the smaller the probe diameter is, the further the resolution, and the working accuracy can be, in principle, enhanced. Now, development of an aberration corrector for the charged particle beam device has been underway in recent years, and progress has so far been made in commercialization thereof. In the aberration corrector, both an electric field and a magnetic field, not rotationally symmetrical with each other, are applied thereto by use of a multipole lens, thereby giving a reverse aberration to the probe beam. Thereby, it is possible to cancel various types of aberrations including a spherical aberration, and a chromatic aberration, occurring to an objective lens, a deflector lens, and so forth.
In the optical system of a conventional charged particle beam device, lenses symmetric to each other, with respect to axial rotation, have been used, so that if the axes of the respective lenses, and the axes of aperture stops are aligned with each other, and the focus as well as a non-point of the objective lens are adjusted, it has, in principle, been possible to adjust the probe diameter to the minimum value. Further, when focus adjustment and non-point correction are executed, adjustment has been carried out by acquiring an image of the probe under a condition that a focus is changed to thereby select a point where a sharpening degree of the image is at the highest, while comparing the sharpening degrees in at least two directions with each other.
On the other hand, in the charged particle beam device provided with the aberration corrector, both the electric field and the magnetic field, not rotationally symmetrical to each other, are applied by the aberration corrector using the multipole lens. Thereby, in these devices, the effect of high-order aberration becomes pronounced although high-order aberration in the conventional optical system that does not exert the effect. To exploit the maximum performance of the device, aberration types (aberration components) contained in the beam, including these aberrations, and respective amounts of the aberration components must be accurately measured, and multipole electric field and magnetic field must be properly adjusted to thereby remove all the aberration components.
As one of the methods of measuring an aberration, there has been known a method whereby an image is picked up in a state where an electron beam tilted against the optical axis of an optical system for use in measurement is caused to fall on the optical system, an sample image movement amount caused by the tilted electron beam is measured, and magnitude (aberration coefficients) of an aberration is calculated on the basis of a change in the movement amount, accompanying a change in tilt-condition, thereby finding the aberration coefficients (refer to, for example, nonpatent literature 1). Upon the tilted electron beam falling on an electron beam passing over the optical axis of an optical system, a destination position of the electron beam undergoes a change according to the magnitude of an aberration, contained in the optical system, whereupon a visual view offset occurs to a sample image. Magnitude as well as a direction of the visual view offset represents a function of the magnitude as well as the type of the aberration, contained in the optical system, a tilt angle, and a tilt condition. Accordingly, if there is carried out a correlation operation between the image before electron beam tilting and the image after the electron beam tilting, at plurality of tilt angles, and tilt directions, respectively, and the magnitude as well as the direction of the visual view offset, due to the electron beam tilting, is measured on the basis of an amount of a pixel offset in each of the two images, thereby finding a formula for a curve (an aberration figure) expressing a locus of image movements, this will enable the aberration coefficients to be found from the factor of the formula.
Images subjected to beam tilting from a plurality of directions, respectively, are required for accurate measurement of an aberration. In general, an aberration figure reflects all rotationally symmetrical aberrations contained in an optical system, superimposed on irrotationally symmetrical aberrations contained in the optical system, respectively, so that complex curves are depicted in the aberration figure. Accordingly, to accurately trace the curves, it is necessary to pick up images from a multitude of the tilt directions, respectively, and, in principle, the more the number of the images is, the more accurate is the aberration figure that is obtained, so that measurement accuracy of the aberration coefficients is enhanced. In reality, however, there arises a problem that the more the number of the images is increased, the longer time is required for calculation of amounts of displacements, or acquisition itself of images, in scanning, and retention, respectively.
However, in the case of aberration correction using an aberration corrector, to adjust the corrector, measurement as well as a change in the power supply value of the corrector is often repeated a number of times, which is required because a high-order aberration (parasitic aberration) occurs due to ununiformity of the fields, attributable to safety integrity of a power supply, and working accuracy of the aberration corrector when the power supply value of the aberration corrector is varied, and it is necessary to reduce a spherical aberration of a objective lens, while cancelling these aberrations out each other inside the corrector every time those aberrations occur. For this reason, time necessary for one measurement of an aberration is preferably as short as possible.
Accordingly, to solve those problems, it is an object of the invention to realize execution of an aberration measurement with high precision in a charged particle beam device with an aberration corrector mounted therein, while controlling an operation amount during aberration measurement, thereby enabling time necessary for correction as a whole to be shortened.
To solve the problems described as above, the invention provides a charged particle beam device, and representative ones thereof include the following:
1) a charged particle beam device including an electron beam source for radiating an electron beam, an electron optical system for irradiating a sample with the electron beam, an electron beam detector for detecting an electron beam emitted from the sample irradiated with the electron beam, an aberration corrector for removing an aberration component by application of an electric field, and an magnetic field, not rotationally symmetrical to each other, and a deflector disposed on a side of the aberration corrector, adjacent to the electron beam source, for controlling a route of the electron beam passing through the electron optical system. A beam in an optical condition changed by use of the deflector is used for scanning on a sample having a predetermined pattern, thereby acquiring a plurality of images by use of beams differing in path from each other, and an image formed by cutting out one of the plural images in a scope containing the predetermined pattern, narrower than a region of one other image, is used as a reference image, a means for working out an amount of an aberration on the basis of a differential between the reference image and the one other image being provided.
2) Or a charged particle beam device including an electron optical system for irradiating a sample with an electron beam radiated from an electron beam source, an electron beam detector for detecting an electron beam emitted from the sample irradiated with the electron beam, an aberration corrector for removing an aberration component by application of an electric field, and an magnetic field, not rotationally symmetrical to each other, and a deflector disposed on a side of the aberration corrector, adjacent to the electron beam source, for controlling a route of the electron beam passing through the electron optical system. There are further provided with a means for acquiring two-dimensional brightness distribution information on the basis of a secondary charged particle signal detected by the electron beam detector, a means for acquiring two-dimensional information pieces on a plurality of brightness distributions against the sample by changing an optical condition set by the deflector, and a means for working out an amount of a visual view offset between images of the sample, acquired under optical conditions differing from each other, from the two-dimensional information pieces on the plural brightness distributions.
3) Otherwise, a charged particle beam device including an electron optical system for irradiating a sample with an electron beam radiated from an electron beam source, an electron beam detector for detecting an electron beam emitted from the sample irradiated with the electron beam, an aberration corrector for removing an aberration component by application of an electric field, and an magnetic field, not rotationally symmetrical to each other, and a deflector disposed on a side of the aberration corrector, adjacent to the electron beam source, for controlling a route of the electron beam passing through the electron optical system. There are further provided with a means for acquiring one-dimensional brightness distribution information on the basis of a primary charged particle signal detected by the electron beam detector, a means for acquiring one-dimensional information pieces on a plurality of brightness distributions against the sample by changing an optical condition set by the deflector, and a means for working out an amount of a visual view offset between images of the sample, acquired under optical conditions differing from each other, from the one-dimensional information pieces on the plural brightness distributions.
More specifically, in the present invention, (1) the number of pixels or resolution in the first image serving as the reference as well as in the second image are changed to thereby roughly find a movement destination caused by a visual field offset. Thereafter, the number of pixels or the resolution in the first image is rendered identical in condition to that in the second image to thereby precisely measure a visual field offset amount. Or, (2) the sample having lines extending in the horizontal direction and lines extending in the vertical direction is one-dimensionally scanned to thereby measure image displacement amounts from a signal position offset.
In an image-processing operation, a difference operation is proportional to pixel numbers “n”, but a correlation operation is proportional to the square of “n”. Accordingly, if the first image is cut out small in size onto the second image to thereby grasp an approximate position of a visual view offset by use of the difference operation executed in advance, this will render it possible to reduce a calculation amount of the correlation operation taking much time.
According to the present invention, in the charged particle beam device with the aberration corrector mounted therein, execution of an aberration measurement with high precision can be implemented, while controlling an amount of calculation during measuring an aberration, so that time necessary for correction as a whole can be reduced.
There is described in detail hereinafter an embodiment of a scanning electron microscope (SEM) provided with an aberration corrector of a quadrupole-octopole system electric and magnetic fields superimposition type, as a first embodiment of the invention. In
Further, the SEM according to the invention has a function for enabling a beam incident on an object point of the objective lens to be tilted against the optical axis of the objective lens. For example, the SEM according to the present embodiment has the 2-stage deflection coil 3 above the aberration corrector 4, and therefore, it is possible to create an electron beam with the center axis thereof, having a tilt angle (τ), and an azimuth θ, against the objective lens. Data on both beam tilting and the azimuth thereof is stored in the memory inside the PC to be referred to during acquisition of an aberration correction, and so forth.
Now, there is broadly described a method for finding an aberration coefficients 17 from amounts of displacement 16 of an image. If an object point is irradiated with a beam in a state having a given tilt angle, an optical path difference occurs to the electron beam due to the tilting of the beam, whereupon an aberration caused by the tilting of the beam is applied to a sample image. In general, if a function expressing an optical path difference due to an aberration is defined as χ(ω), χ(ω) can be analytically expressed by use of plural-order aberrations. Herein, χ(ω) for aberrations up to a three-order aberration can be represented by formula (1) as follows:
In expression (1), A0, C1, A1, B2, A2, C3, S3, and A3 express an image movement, a defocus, two-fold symmetry astigmatism, an axial coma aberration, three-fold symmetry astigmatism, a tertiary spherical aberration, a star aberration, and four-fold symmetrical astigmatism, respectively. Further, ω expresses complex coordinates on an object surface. Herein, when an incident electron beam is tilted at the tilt angle (τ), χ(ω) can be represented by the following formula (2): it is assumed that the tilt angle (τ) is expressed by a complex number.
In expression (2), A0(τ), and C1(τ) each express the aberration coefficients 17 when the electron beam is tilted by the tilt angle (τ). The respective aberration coefficients at the time when the electron beam is tilted are expressed as the sum of the aberration coefficients when the beam is tilted at the tilt angle (τ), and an aberration coefficients when the electron beam is not tilted. In the case of taking the aberrations up to the three-order aberration into consideration, an image movement emerging due to tilting is represented as follows:
formula 3
A
0(τ)=A0+A1
As is evident from expression (3), all the aberration coefficients up to the third-order aberration coefficients before tilting are included in A0(τ). That is, if the function form of A0(τ), and values of A0(τ) under several tilt-conditions are known, the aberration coefficients can be found by function fitting.
Next, if an irradiation direction of the electron beam that is tilted is represented in complex-number expression, the tilt angle (τ) can be represented by formula (4) on the basis of a tilt angle (t) against the optical axis of a lens, and an azimuth φ of an electron beam incident on a lens plane.
formula (4):
τ=teiφ (4)
If the expression (3) is replaced with expression (4) to be reorganized, the expression (3) can be finally written in the following form.
Herein, mk(t) is a factor expressed by a formula based on linear connection between the respective aberration coefficients before the beam-tilting and “t”. Accordingly, mk(t) can be found by measuring several azimuths φ of A0(τ) at the angle (t) to thereby execute the function fitting, and if equations as obtained are simultaneously solved, this will enable all the aberration coefficients 17 before the beam-tilting to be calculated.
A procedure for use in execution of aberration correction according to the present embodiment is described hereinafter with reference to a flow chart shown in
Further, control of the deflection coil power-supply control unit 20, and input of information necessary for the control are executed by use of a control unit 19, and an input unit 18, respectively.
In
Next, an amount of a visual view offset in the second image is to be found against the first image, and at this point in time, a pixel in the vicinity of the center of the first image is cut out (STEP6), as shown in
A size of the cutout region is preset by a user before a measurement such that divisors are m, n, in the x-direction, and the y-direction, respectively, against the pixel numbers M×N {pix} of a scanning region in whole.
Subsequently, a differential between cells, in the first image, and the second image, respectively, is obtained, and search for a region having a higher differential value is carried out to thereby find a candidate cell at a movement destination. Thereafter, a correlation calculation against the candidate cell is executed to thereby find an amount of movement of the visual view offset, in detail (STEP8).
For example, in the respective cases of
Upon obtainment of the amount of movement, described as above, an tilting angle (τ), and an azimuth θ are newly set (STEP4) to pick up an image of which an amount of a visual view offset is to be measured, thereby repeating this procedure until a target number of data blocks are obtained.
Further, in the case of a visual view offset region in the second image crosses a plurality of regions as shown in
Thus, if the first image is cut out small in size onto the second image to thereby grasp an approximate position of a visual view offset by use of the difference operation executed in advance, this will render it possible to reduce a calculation amount of the correlation operation taking much time. Since the correlation operation is a calculation to roughly search for the visual view offset, there is no need for accurately grasping the amount of the visual view offset on a pixel-by-pixel basis.
With an image-processing operation, the difference operation is proportional to pixel numbers “n”, and the correlation operation is proportional to the square of “n”, so that if a movement-destination candidate region is searched for according to the differential in advance, and the correlation operation is applied only to a limited number of the candidate regions, this will enable an amount of an operation necessary for calculation of a visual field offset amount to be reduced. Further, if only an image in the central cutout portion is used in peripheral regions of the first image, this will enable the number of data blocks, acting as noise, respectively, to be reduced, so that the amount of movement can be measured with high precision.
To accurately carry out the above described embodiment, it is preferable that a sample used for photographing may have many edges in all the directions in order to increases the information amount of an image. Furthermore, using of an isolated pattern where a sample pattern is present only in part of a visual field region can reduce the number of data blocks acting as noise during the difference calculation, and can thus enhance the precision for identifying the movement destination.
Further, in the case of carrying out a measurement using the isolated pattern, even a region without a pattern present may be used as the first image, as shown in
Upon acquisition of data on a desired amount of movement, calculation of the aberration coefficients 17 is executed by aberration coefficients calculation unit 14 on the basis of the data (STEP10) and the result is sent out to a correction power setting unit 15. A table listing the aberration coefficients 17, and respective aberration corrector power supply values corresponding thereto is pre-stored in the aberration correction power setting unit, and a power supply value to be given to the aberration corrector 4 to correct an aberration to a presently desired aberration can be obtained by referring to the table (STEP11). At this point in time, a power supply value to be actually given to the aberration corrector 4 is determined on the basis of the result of an aberration coefficients 17 obtained by the aberration coefficients calculation unit 14, and power supply data to be given to the aberration corrector 4 is sent out to a corrector power-supply controller 21. The power supply value of the aberration corrector 4 is changed via the (aberration) corrector power-supply controller 21, whereupon the aberration correction is executed (STEP12).
The second image as obtained is similarly divided by m, n in the x-direction, and the y-direction, respectively, dividing the image as a whole into m×n of small regions, as is the case with the first embodiment, thereby roughly identifying a movement position by obtaining the differential of the first image on a region-by-region basis. Thereafter, the horizontal scanning frequency of the scan coil is set again to a value identical to, or higher than that during acquiring the first image, and only the vicinity of the destination candidate obtained as a result of differentiation is scanned in high resolution. By execution of the correlation operation as to an image obtained as above, and the first image, movement coordinates in detail can be found.
Since images of visual views in a wide scope can be obtained within the same scanning time by lowering the resolution of the second image, during differential calculation, against the first image, as described above, it is unnecessary to cut out the first image in the case of the present embodiment. Searching in the wide scope is executed in the difference operation, and the correlation operation is executed only in the vicinity of the destination candidate, so that it is possible to reduce an amount of calculation in the correlation operation that takes much time. Furthermore, by scanning only the vicinity of the destination candidate at a high resolution, the correlation operation can be enhanced in precision.
In both the first embodiment and the second embodiment, if a pretreatment using a sharpening filter is applied to the second image, this will reduce effects of defocusing due to tilting to thereby improve the precision of differential searching.
Reduction in the amount of the difference operation can also be achieved when the difference operation is executed by executing calculation every n pieces of pixels instead of taking the differential in all the pixels, as shown in
There is shown another example as still another embodiment of the invention, where an amount of movement is measured by use of a sample 60 having lines 70 arranged at unequal intervals in the horizontal direction of a visual field, and lines 71 arranged at unequal intervals in the vertical direction of the visual field, as shown in
Accordingly, if the sample as described is scanned one line by one horizontal line, and in the vertical direction, respectively, to thereby examine the line profile, the amount of movement can be measured without acquisition of two-dimensional brightness distribution data, and scanning time can be significantly shortened.
In the present embodiment, it need only be sufficient to have a sample provided with lines arranged at unequal intervals in the horizontal direction, and in the vertical direction, respectively. Accordingly, use may be made of a sample 61 having an unequally-spaced grid pattern 72 shown, for example, in
Otherwise, as shown in
There is shown still another example as a further embodiment of the invention, where the present technique is applied to the case of executing an automatic operation using a measuring SEM. The measuring SEM is a device for measuring a distance between two points on data of an image measured by calculation of pixels. In
The load lock chamber 102 is separated from the sample chamber of the main body of the device by a gate valve 31. The gate valve 31 is opened at the time when the sample is to be introduced into the device, whereupon the sample is introduced into the device by a sample transport mechanism 30. Further, adjustment of the device is carried out by use of a standard sample 32 placed on the sample stage 8.
The measuring SEM according to the present embodiment is provided with a boosting electrode 33 disposed above a magnetic-field type objective lens 7. An electrostatic lens is formed by application of an electric field to the boosting electrode, and fine adjustment of a focus can be effected by varying the intensity of the electrostatic lens. A voltage applied to the boosting electrode 33 is varied by controlling a boosting electrode power-supply control unit 34. Further, a voltage (retarding voltage) for forming a retarding field against an incident electron beam is applied to the sample stage 8 by a retarding power-supply control unit 35; however, the focus can also be adjusted by controlling the retarding voltage using the retarding power-supply control unit 35. Usually, response of the magnetic-field type objective lens to an excitation current is delayed due to aftereffects of magnetism, so that a focus can be speedily altered by adjustment of a boosting voltage, or the retarding voltage instead of the excitation current of the magnetic-field type objective lens.
To perform an automatic operation, a user checks a correction state through a GUI screen shown in
| Number | Date | Country | Kind |
|---|---|---|---|
| 2010-124315 | May 2010 | JP | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/JP2011/062190 | 5/27/2011 | WO | 00 | 11/27/2012 |
