JPH1062149A - Defect inspection equipment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To inspect defect in a pattern formed on a sample at high rate with high accuracy. SOLUTION: An electron beam radiated radially from an electron gun 12 is converted through condenser lenses 14, 16 into a beam parallel with the optical axis and projected to a mask 18 having holes 18A to produce a plurality of multibeams MB. The multibeams MB is deflected through deflectors 20, 22 and reduced by a factor of 2 through electromagnetic lenses 24 26 before impingingon a sample 30. Since the electromagnetic lenses 24, 26 are symmetric magnetic tablet lenses arranged symmetrically to a crossover CO position in the direction of the optical axis, aberration of the lens is canceled and the beam can be converged thin. Defect detection is performed at high rate with high accuracy by scanning the surface of a sample 30 with that beam.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a defect inspection apparatus, and more particularly, to a defect inspection apparatus for inspecting a defect of a sample by irradiating the sample with a charged particle beam and detecting the charged particle beam from the sample. More particularly, the present invention relates to a defect inspection apparatus suitable for inspecting a defect such as a circuit pattern formed on a photomask, a reticle, or a wafer (hereinafter, referred to as a “sample”).

[0002]

2. Description of the Related Art Conventionally, in the manufacturing process of semiconductor devices and the like, a large number of masks or reticles for sequentially transferring circuit patterns and the like onto a wafer have been used. And
For example, when manufacturing the mask itself, a defect inspection apparatus that inspects the pattern for defects to check whether the transfer pattern formed on the mask is accurately formed according to the design data is used. ing.

Further, the defect inspection apparatus can similarly inspect the presence or absence of a defect of a circuit pattern transferred onto a wafer using a mask or a reticle.

For example, a conventional defect inspection apparatus forms a single probe by converging an electron beam or the like, irradiates a pattern on a sample to be inspected with a defect, and generates a secondary electron generated from the sample. Is detected by a detector, and a defect inspection is performed by processing the detection signal. In particular, when inspecting a defect in a certain area (for example, a pattern area) on a sample, the above-described probe is deflected to irradiate while performing raster scanning within the inspection area. Defect inspection has been performed by detecting sequentially generated secondary electrons with a detector.

[0005]

In the above-described conventional defect inspection apparatus, one electron beam obtained by converging an electron beam or the like is used as a probe.
In order to detect a defect having a pattern of about 1 μm, it is necessary to scan the sample with no gap between the sample and the scanning takes a long time, so that there is a disadvantage that the throughput of the defect inspection is reduced.

In order to shorten the scanning time of the probe, it is conceivable to scan the probe at a high speed. However, the S / N ratio of the secondary electron signal from the sample obtained by the detector becomes small, and erroneous detection increases. There was a risk of doing.

[0007] In an electron optical system for guiding an electron beam to a sample as a probe, an electron lens including an electrostatic lens using an electric field and an electromagnetic lens using a magnetic field is used. When these electron lenses are used in the electron optical system of the defect inspection device, they are affected by aberrations generated in the rotation direction around the optical axis of the lens and distortions generated in the radiation direction. The electron beam is blurred,
When the beam position shifts, the resolution of defect inspection deteriorates, and the defect position cannot be detected accurately. In particular, the influence of these aberrations tends to increase as the distance from the optical axis of the lens increases.

The present invention has been made under such circumstances, and an object of the present invention is to provide a defect capable of inspecting a defect of a pattern formed on a sample at high speed and with high accuracy. An object of the present invention is to provide an inspection device.

[0009]

According to the first aspect of the present invention, a sample (30) on which a pattern is formed is irradiated with a charged particle beam (MB), and a charged particle beam (SB) from the sample (30) is irradiated. ) And inspecting the pattern for defects in said pattern, comprising: charged particle beam forming means (12, 14, 16, 18) for forming a plurality of said charged particle beams; and said plurality of charged particle beams ( MB) and the sample (30) are arranged at optical axis direction positions corresponding to the reduction ratio with respect to a crossover (CO) position which is a position obtained by dividing the sample (30) at a predetermined reduction ratio. A symmetric magnetic tablet type charged particle beam optical system (24, 26) comprising two similar shaped electron lenses (24, 26); and a sample (30) from the charged particle beam optical system (24, 26). The plurality of charged particles irradiated to Detecting means for detecting the charged particle beam generated from said sample (30) by line (MB) of (SB) (38A, 38B, 38C, 38D) characterized in that it comprises a.

According to this, a plurality of charged particle beams are formed by the charged particle beam forming means, and the plurality of charged particle beams are irradiated on the sample by the charged particle beam optical system including the symmetric magnetic tablet type lens. The defect inspection is performed by detecting the charged particle beam generated from the irradiation position of the charged particle beam by the detecting means.

As described above, when a defect formed on a pattern formed on a sample is inspected, a plurality of charged particle beams are used to detect a plurality of defects on the sample, as compared with a conventional example in which a single charged particle beam is used for inspection. Since the defect inspection can be performed at the same irradiation point at the same time, the inspection time is greatly reduced, and the throughput can be improved. In addition, the charged particle beam optical system that irradiates a sample with a plurality of charged particle beams has a crossover position centered at a predetermined reduction ratio between a position where a plurality of charged particle beams are formed and a sample. In addition, a symmetric magnetic tablet type lens in which two similar electronic lenses are arranged at symmetric positions in the optical axis direction corresponding to the reduction ratio is adopted. For this reason, even if the aberration occurs when the charged particle beam passes through one of the electron lenses, the opposite aberration appears when the charged particle beam passes through the other electron lens and acts to cancel the aberration. As a result, the influence of aberration is almost eliminated, and the charged particle beam can be converged finely, so that the resolution can be improved and a highly accurate defect inspection can be performed.

Here, in order to scan the charged particle beam on the sample surface, the sample stage on which the sample is placed may be configured to be movable. Means (12, 14, 16, 18) and charged particle beam optical system (2
Charged particle beam scanning means (2) for scanning a plurality of charged particle beams (MB) irradiated to the sample (30) between the charged particle beam scanning device (4, 26).
0, 22). The charged particle beam scanning means scans a plurality of charged particle beams on a sample to perform defect inspection, thereby performing a wide range of defect inspection in a short time. Further, since the charged particle beam scanning means is arranged between the charged particle beam forming means and the charged particle beam optical system, the space between the charged particle beam optical system and the sample can be effectively used. For example, a space can be secured between the charged particle beam optical system and the sample, in which a detection unit for detecting a charged particle beam generated from the sample, a multi-lens for guiding the charged particle beam to the detection unit, and the like can be provided.

The number of the charged particle beam scanning means is not particularly limited, but as in the third aspect of the present invention,
It is more preferable that the charged particle beam scanning means (20, 22) of the second aspect is constituted by an even number of deflectors (20, 22) along the optical axis direction. The reason why the deflector, which is the charged particle beam scanning means, is an even-numbered stage is that when one of the deflectors deflects the charged particle beam, the angle of incidence of the charged particle beam incident on the charged particle beam optical system changes. This is because the aberration generated in the line optical system increases, but if an even-numbered deflector is used, the other deflector can deflect the light so that the incident angle does not change. In particular, when the deflector has an even number of stages, the angle of incidence on the charged particle beam optical system does not change simply by making the absolute value of the deflection amount equal and reversing the deflection direction, so that it can be easily controlled. There is. As a result, even if the charged particle beam is scanned by the charged particle beam scanning means, the influence of aberration and the like does not increase, so that the charged particle beam can be narrowed and converged to perform a highly accurate defect inspection.

Further, according to a fourth aspect of the present invention, in the defect inspection apparatus according to any one of the first to third aspects, the charged particle beam forming means (12, 14, 16, 1) is provided.
8) A charged particle beam source (12) for generating a charged particle beam
An electron lens (14, 16) for forming a charged particle beam generated by the charged particle beam source into a substantially parallel charged particle beam having a predetermined line width; A mask (18) for forming a charged particle beam (MB).

According to this, a plurality of charged particle beams can be formed with a simple configuration of a charged particle beam source, an electron lens, and a mask, and the number, size and arrangement of the plurality of charged particle beams are as follows. For example, it can be easily realized by changing the number, size, arrangement and the like of the holes formed in the mask.

The invention described in claim 5 is the first invention.
The defect inspection apparatus according to any one of claims 1 to 4, wherein the charged particle beam forming means (12, 14, 16, 1)
8) The electronic lens (24) arranged on the side is shifted toward the crossover (CO) position by a predetermined distance K (K is a positive or negative real number), and the electronic lens (24) on the sample (30) side is shifted. 26) is shifted toward the crossover (CO) position by a distance corresponding to a product of a predetermined distance K and a reduction ratio 1 / M.

According to this, in the charged particle beam optical system constituted by a symmetric magnetic tablet type lens, two electron lenses are arranged so as to maintain symmetry about a crossover position. Therefore, to shift the position of the electron lens while maintaining its symmetry, one of the electron lenses (here, the electron lens arranged on the charged particle beam forming means side)
Is shifted to the crossover side by a predetermined distance K (K is a positive or negative real number), the other electron lens (here, the electron lens arranged on the sample side) is charged to the crossover side by the distance K and the charge. It is shifted by a distance multiplied by 1 / M, which is the reduction ratio of the particle beam optical system. As a result, in the charged particle beam optical system constituted by the symmetric magnetic tablet type lens, two similar electron lenses are arranged at the optical axis direction position corresponding to the reduction ratio with the crossover position as the center, and the symmetry is maintained. Therefore, the space between the charged particle beam optical system and the mask, or the space between the charged particle beam optical system and the sample can be increased while reducing aberrations. Can be installed without difficulty.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIG.
A description will be given based on FIG.

FIG. 1 shows a schematic configuration of a defect inspection apparatus 10 according to the present embodiment. This defect inspection device 10
Is a defect inspection apparatus that inspects a fine circuit pattern of a semiconductor element formed on a wafer (hereinafter, referred to as a “sample”) for defects of a fine circuit pattern using an electron beam.

In this defect inspection apparatus 10, an electron gun 12 as a charged particle beam source, condenser lenses 14, 16 and a mask 18 (charged particle beam forming means is constituted by the electron gun 12, the condenser lenses 14, 16 and the mask 18). Has been),
Deflectors 20 and 22 as charged particle beam scanning means, electromagnetic lenses 24 and 26 (symmetric magnetic tablet type lens) as charged particle beam optical system, aperture 28, sample 30, multi-lens 34, multi-aperture plate 36, detector 38A, 38
B, 38C, 38D, signal processing circuit 40, comparison circuit 4
2, a storage unit 44 and the like.

The electron gun 12 is a charged particle beam source that emits an electron beam forming a probe for defect inspection, and a thermoelectron is extracted from an incandescent filament or the like by an electric field. Here, the electron gun 12 is used as a charged particle beam source. However, when a defect inspection is performed using a charged particle beam such as a proton or an ion in addition to the electron beam, a charged particle beam source corresponding to the charged particle beam is used. Is used. For example, when performing a defect inspection using ions, a charged particle beam source is an ion source that ionizes atoms or molecules by discharging gas and extracts the ions or atoms by an appropriate electromagnetic field.

The condenser lenses 14 and 16 are connected to the electron gun 1
The electron beam emitted radially from 2 is converged by a convex lens action or an electron beam PB substantially parallel to the optical axis AX.
It is a lens that changes to Also, this condenser lens 1
4 and 16, the plurality of electron beams M
The crossover CO position where B intersects can be set to a desired position.

The mask 18 has a plurality of holes 18A formed in a doped conductive silicon plate. When the electron beam PB made substantially parallel by the condenser lens 16 is irradiated, each of the holes 18A is formed. The electron beam passes through the hole 18A of the
To form For example, in this embodiment, holes 18A of 0.2 μm square are formed in the mask 18 at a pitch of 400 μm in a direction perpendicular to the paper surface (this is defined as Y-axis direction) in 10 rows, and in a paper surface direction (this is defined as X-axis direction). 10 rows are provided, a total of 10
Zero holes 18A are formed in a matrix. The charged particle beam forming means for forming a plurality of charged particle beams includes the electron gun 12 and the condenser lenses 14 and 16 described above.
And a mask 18.

The deflectors 20 and 22 are disposed in two stages in the direction of the optical axis AX between the mask 18 and an electromagnetic lens 24 to be described later so as to surround the optical path of the electron beam MB. . 10 formed by the above-described mask 18
The zero electron beam MB is simultaneously deflected by the deflectors 20 and 22, and the inspection area on the sample 30 is raster-scanned. Here, the reason why the deflector is configured in two stages is that
For example, by making the absolute value of the amount of deflection of the deflector 20 equal to the absolute value of the amount of deflection of the deflector 22 and reversing the direction of deflection, even if the electron beam MB is deflected, the angle of incidence on the electromagnetic lens 24 changes. (See the deflection light PL shown by the broken line in FIG. 1), and it is possible to prevent the occurrence of aberrations due to the change in the incident angle of the electron beam MB.

The electromagnetic lenses 24 and 26 convert a plurality of electron beams MB formed on the mask 18 into a certain reduction ratio 1 / M.
After reducing by (here, １ ／), an image is formed on the sample 30 to be inspected for defects. These two electromagnetic lenses 24,
Reference numeral 26 is designed to have a similar shape around a crossover CO position, which is a point on the optical axis where a plurality of electron beams MB intersect. That is, the electromagnetic lens 24 and the electromagnetic lens 26 are arranged so that the ampere turns (AT: current × the number of coil turns) are equal and the directions of the generated magnetic fields are opposite to each other. , A lens group called a symmetric magnetic tablet type. More specifically, the electromagnetic lens 2
Numeral 4 is provided on the mask 18 side of the crossover CO position, and a coil for generating a magnetic field is arranged so as to surround the optical path. The electromagnetic lens 26 is provided closer to the sample 30 than the crossover CO position, and a coil similar in shape to the electromagnetic lens 24 is arranged to surround the optical path. The charged particle beam optical system that guides a plurality of charged particle beams onto the sample includes the electromagnetic lenses 24 and 26.

The aperture 28 is provided in the vicinity of a crossover CO position where a plurality of electron beams intersect, and is formed of a non-magnetic metal plate such as aluminum or copper having an opening in the center thereof through which the electron beam passes. Have been. This aperture 28 allows the reduction ratio of the electromagnetic lenses 24 and 26 between the mask 18 and the sample 30 to be 1 / M.
In this case, the condenser lens 14 is provided at a position internally divided into M: 1 (here, 2: 1 because the reduction ratio is 1/2). It is adjusted at 16.

Here, the sample 30 is a wafer on which a circuit pattern for performing a defect inspection is formed, but when inspecting a defect such as a photomask or a reticle, these are the samples.

Here, the multi-lens 34 is provided in a space between the electromagnetic lens 26 and the sample 30, and a non-magnetic metal plate is disposed so as to surround the periphery of the sample 30. A plurality of lens apertures 34A, 34B, 34C, 34D having a predetermined aperture are formed corresponding to the positions where secondary electrons are emitted from the sample 30. FIG.
Is a sectional view, only four lens apertures are shown. However, actually, a number of lens apertures corresponding to the number of electron beams MB (here, 100) are formed at corresponding positions. I have. In the description of the present embodiment,
The four lens apertures will be described as an example. Further, a predetermined convergence potential is applied to the multi-lens 34 from a potential control unit (not shown), and equipotential surfaces (not shown) are formed in respective portions of the lens apertures 34A, 34B, 34C, and 34D. Act as For this reason, the plurality of electron beams MB
When each of the above irradiation points is irradiated, secondary electrons are emitted from each irradiation point, and when the secondary electrons pass through the corresponding lens apertures 34A, 34B, 34C, 34D, they are individually converged by the convex lens action. And a detector 38 described later.
A, 38B, 38C, 38D.

The multi-aperture plate 36 is a non-magnetic metal plate disposed between the multi-lens 34 and detectors 38A, 38B, 38C, 38D described later, and the non-magnetic metal plate has a predetermined aperture. A plurality of openings 36A, 36B, 36
C and 36D are formed. The openings of the multi-aperture plate 36 are formed at respective passage positions connecting the lens aperture of the multi-lens 34 and the detector, corresponding to the emission positions of the respective secondary electrons on the sample 30. A constant convergence potential is applied to the multi-aperture plate 36 from a potential control unit (not shown), so that the openings 36A, 36B, 36
An equipotential surface (not shown) is formed on each of the portions C and 36D, and functions as a convex lens here. For this reason,
For example, the secondary electrons converged by passing through the lens aperture 34A of the multi-lens 34 are converged again at the opening 36A, and can be reliably incident on the detector 38A. The same applies to the other openings 36B and 36C. Since FIG. 1 is a cross-sectional view, similarly to the above-described lens aperture, the openings 36A,
Although only four of 36B, 36C and 36D are shown,
Actually, a number of openings corresponding to the number (100) of electron beams MB are formed at corresponding positions. In the description of the present embodiment, the above-described four openings will be described as an example.

Detectors 38A, 38B, 38C, 38D
For example, when a plurality of electron beams MB are irradiated on a circuit pattern formed on the sample 30 and secondary electrons generated from each irradiation point of the sample 30 are incident on the circuit pattern, a detection signal corresponding to the incident electron dose is generated. And an electron multiplier that generates the The detectors 38A, 38B, 38C,
1 is a cross-sectional view of FIG. 1, only four are shown in the same manner as the lens aperture of the multi-lens 34 and the opening of the multi-aperture plate 36 described above. (100 detectors) are arranged at corresponding positions. In the description of the present embodiment, the above-described four detectors will be described as an example.

The signal processing circuit 40 includes detectors 38A, 38
This is a circuit for converting a detection signal of secondary electrons output from B, 38C, 34D and the like into a digital signal for comparison with reference data of a pattern according to a design value.

The comparison circuit 42 detects the presence / absence of a defect and the position of the defect by comparing the detection signal of the secondary electron converted into a digital signal with reference data of a pattern as designed.

The storage unit 44 stores in advance reference data obtained when a pattern formed according to a design value having no defect is inspected.

Next, the operation of the defect inspection apparatus 10 configured as described above will be described.

First, the operator places a sample (wafer on which a circuit pattern is formed) 30 to be inspected for defects in the X and Y directions.
After mounting on a sample table (not shown) movable in the two-dimensional plane direction, the operation of the defect inspection is started.

As shown in FIG. 1, the electron beam radially emitted from the electron gun 12 is once converged and crossed by the condenser lens 14,
As a result, an electron beam PB substantially parallel to the optical axis is irradiated onto the mask 18 in which the plurality of holes 18A are formed. The parallel electron beam PB irradiated on the mask 18
By passing through A (100 holes), 100 electron beams MB (hereinafter, referred to as “multi-beam MB”) are formed.

The multi-beams MB formed by the mask 18 have a beam width of 0.2 μm square and are arranged in a matrix of (10 rows × 10 columns) at a pitch of 400 μm. When the multi-beam MB passes through the symmetric magnetic tablet type electromagnetic lenses 24 and 26, the multi-beam MB 1 /
The image is reduced to 2 and is imaged on a circuit pattern on the sample 30 to be irradiated.

When the circuit pattern on the sample 30 is irradiated with the multi-beam MB, secondary electrons are emitted from each irradiation point of the sample 30. The amount of secondary electrons emitted from each irradiation point varies depending on the state of the pattern formed on the sample (for example, the presence or absence of the pattern, the thickness of the pattern, and the like). For this reason, as shown in FIG.
The secondary electrons emitted from the upper irradiation point are converged by, for example, the lens action of the lens aperture 34A of the multi-lens 34, separated through the opening 36A of the multi-aperture plate 36, and detected by the detector 38A. . Similarly, secondary electrons emitted from other irradiation points are detected through a path of lens aperture 34B → opening 36B → detector 38B, and secondary electrons emitted from other irradiation points are: Secondary electrons detected via the path of the lens aperture 34C → the opening 36C → the detector 38C and emitted from the other irradiation points are further converted to the lens aperture 34D.
Detection is performed individually by passing through a path of → opening 36D → detector 38D.

As described above, the secondary electrons emitted from each irradiation point on the sample 30 are detected by the detectors 38A and 38A corresponding to the irradiation points.
The signals are detected by 38B, 38C, 38D and the like, and the signal processing circuit 40 performs signal processing for converting the detected signals into digital signals that can be compared with reference data. The digital signal that has been subjected to this signal processing is stored in a storage unit 44 by a comparison circuit 42.
By comparing with the reference data of the pattern according to the design value stored in the sample pattern, it is possible to detect the presence / absence and the position of the defect of the circuit pattern on the sample 30 to be inspected.

The above-described defect inspection apparatus 1 of the present embodiment
According to 0, a multi-beam MB is used as a probe for performing a defect inspection. FIG. 2 shows the electromagnetic lenses 24, 2
FIG. 3 is a view of the sample 30 from the 6 side, and FIG.
100 multi-beams MB10 imaged on plane 0
The irradiation pattern of MB109 is shown. Here, since the reduction ratio of the electromagnetic lenses 24 and 26 is ２, the multiple beams MB10 to MB109 formed on the sample 30 are formed.
Is 0.1 μm square, and the pitch between beams is 2 μm.
00 μm. If the deflection amounts of the deflectors 20 and 22 are respectively set to 200 μm in the X and Y directions on the sample 30, 2000.1 μm × 2000.
A region R of 1 μm square is raster-scanned. When the raster scanning in the region R is completed, the sample stage is moved by a predetermined distance in the X or Y direction by the sample stage drive mechanism (not shown) (the sample 30 placed on the sample stage is also integrally driven), and again. The same raster scanning as described above is performed. As described above, by repeatedly performing the raster scanning in the region R while sequentially moving the sample 30, the defect inspection of the predetermined region on the sample 30 can be completed.

As described above, in the defect inspection apparatus 10 of the present embodiment, the multibeams MB10 to MB
Since B109 is used and the defect inspection is performed while simultaneously deflecting and raster-scanning, the scanning range can be reduced as compared with the conventional inspection by scanning with one electron beam. The time is shortened, and the throughput of the defect inspection can be greatly improved.

In the present embodiment, the charged particle beam optical system is a symmetric magnetic tablet type in which two similar electromagnetic lenses 24 and 26 are symmetrically arranged about the crossover CO position. The influence of the aberration can be reduced by mutually canceling out the aberration in the rotation direction around the center and the distortion aberration in the radial direction. As described above, by reducing the various low yields generated in the electromagnetic lens, the deviation of the irradiation position of the beam is reduced, so that the defect inspection can be performed accurately and the beam can be converged finely with high resolution. Defect inspection can be performed. In addition, when the lens has a low aberration, the influence of the aberration that increases as the position of the lens through which the electron beam passes away from the optical axis AX can be reduced, so that the lens diameter can be increased. Therefore, defect inspection can be performed at high throughput by increasing the number of multi-beam MBs, and by increasing the beam interval between the multi-beam MBs, crosstalk generated between adjacent beams can be reduced, and the By increasing the S / N ratio of the detection signal, the resolution of the defect inspection can be improved.

That is, this will be described with reference to FIG. 3 showing the multi-beam MB incident on the electromagnetic lenses 24 and 26. For example, in the comparative example of FIG. 3A, only the lens aperture of radius R1 can be used in order to avoid the influence of aberration at the peripheral portion of the lens.
If it is set to 1, the number of beams becomes nine (in FIG. 3, only the multi-beam MB passing on the paper surface is shown. Actually, beams of the same density are applied to the near side and the far side of the paper surface). There). On the other hand, in the present embodiment, since the symmetric magnetic tablet type lens is employed, the radius R2 (R1 <R2) due to the reduced aberration at the lens peripheral portion as shown in FIG. 3B. If the beam aperture is kept at P1, the number of beams can be increased (here, increased from 9 to 13), so that the scanning time is further reduced and the throughput of defect inspection is increased. Can be improved.

FIG. 3 shows another example of the present embodiment.
In FIG. 3C, a lens aperture having a radius R2 can be used since a symmetric magnetic tablet type lens is employed as in FIG. 3B. Here, in FIG. 3 (C), the usable lens diameter is increased. For example, by setting the same number of beams (9) as in FIG. 3 (A), the beam interval becomes P2 (P1 <P2). And the crosstalk between adjacent beams can be reduced, and the S / N ratio of the detection signal in the detector can be increased, thereby improving the resolution of defect inspection. be able to.

Further, in this embodiment, the inspection area on the sample 30 for performing the defect inspection is irradiated with the multi-beam MB, and the secondary electrons emitted from each irradiation point are converted into the lens of the multi-lens 34 corresponding to each irradiation point. Apertures 34A, 34B, 34
C, 34D and openings 36A, 36 of the multi-aperture plate 36.
B, 36C and 36D, and are surely detected by the detectors 38A, 38B, 38C, 38D and the like corresponding to the respective irradiation points.
Even in the case where the defect inspection is performed by raster-scanning the sample 30 while simultaneously deflecting it at 0 and 22, secondary electrons continuously generated from each irradiation point can be reliably detected in real time.

Further, the electromagnetic lenses 24, 2 of the present embodiment
Numeral 6 denotes similar electromagnetic lenses arranged at symmetrical positions in the optical axis direction with the center of the crossover CO as the center. The ampere turns (AT: current × the number of coil turns) of both lenses are equal to each other, and the magnetic field generated Are excited so that their directions are opposite to each other. In general, the symmetric magnetic tablet type electromagnetic lenses 24 and 26 are provided with the center of the electromagnetic lens 24 at the midpoint between the mask 18 and the crossover CO position shown in FIG. It is said that the center of the electromagnetic lens 26 should be provided at the midpoint between the positions. However, in the present embodiment, similar electromagnetic lenses 24 and 26 are arranged at positions in the optical axis direction corresponding to the lens reduction ratio 1 / M (here, 1/2) with the crossover CO position as the center. Thus, the electromagnetic lenses 24 and 26 can be moved in a predetermined optical axis direction under conditions that always satisfy the symmetry.

For example, as shown in FIG. 1, it is assumed that the electromagnetic lens 24 is shifted toward the crossover CO position by a predetermined distance K (here, K = 40 mm, but may be any positive or negative real number). Then (in the direction of arrow M1 in FIG. 1), the electromagnetic lens 26 is correspondingly moved to the crossover CO position side by a distance (K × 1 / M = K) corresponding to the product of the predetermined distance K and the reduction ratio 1 / M. / 2 = 20 mm) (in the direction of arrow M2 in FIG. 1). FIG. 1 shows an arrangement after the electromagnetic lenses 24 and 26 are shifted from the midpoint by the above-described distance. Thus, the electromagnetic lens 2
4 and 26 are moved while maintaining the symmetry of the arrangement around the crossover CO position, thereby reducing the aberration and reducing the aberration between the mask 18 and the electromagnetic lens 24 on the mask 18 side and the sample 30 and the sample side. Of the electromagnetic lens 26 can be widened. For this reason, it is possible to arrange the deflectors 20 and 22 and the multi-lens 34 and the like having a two-stage configuration in their respective spaces.

In the above embodiment, the electromagnetic lens 2
4 and 26 are moved in a direction approaching the crossover CO position (when the predetermined distance K is positive), and conversely, both lenses are moved in a direction away from the crossover CO position (the predetermined distance K is negative). Case) is also possible.

As described above, since the electromagnetic lenses 24 and 26 maintain their symmetry even when moved in the direction of the optical axis AX, the effect that the aberration generated by one of the electromagnetic lenses is canceled by the other electromagnetic lens. , And the effect of reducing aberrations (for example, rotational aberration and radial distortion) can be maintained. Therefore, the electron beam as a probe can be converged finely, and defect inspection can be performed with high accuracy (high resolution).

In this embodiment, even-numbered (here, two-stage) deflectors 20 and 22 are used because when one charged deflector deflects a charged particle beam, it enters the electromagnetic lens 24. The angle of incidence of the charged particle beam at the time of the change changes the aberration generated in the electromagnetic lens 24. However, in the case of a two-stage deflector, the absolute value of the deflection amount is made equal and the deflection direction is reversed. This is because there is an advantage that a simple control method can be used so that the angle of incidence on the charged particle beam optical system does not change (see the deflection light PL in FIG. 1). Therefore, the deflector 2
When the multi-beam MB is scanned by using 0 and 22, the influence of aberrations and the like does not increase, and the resolution can be improved and highly accurate defect inspection can be performed by converging the charged particle beam finely. In addition, the multi-beam MB can be scanned to perform a defect inspection over a wide area at high speed.

As described above, the defect inspection apparatus according to the present embodiment irradiates a circuit pattern on a sample with a plurality of electron beams (multi-beam MB) and performs a raster scan on the circuit pattern.
By detecting secondary electrons generated from the sample, a defect inspection with high throughput can be performed.

Further, the defect inspection apparatus of the present embodiment employs a symmetric magnetic tablet type electromagnetic lens and arranges the electromagnetic lens at a symmetrical position centered on the crossover CO position to generate the lens inside the lens. Rotational aberrations, radial distortions, and the like that cancel each other out become low aberrations, and the beam can be narrowly converged. As a result, the resolution of defect inspection is improved, and highly accurate defect inspection can be performed.

In the above embodiment, an electron beam is used as a charged particle beam. However, the present invention is not limited to this, and a defect inspection apparatus is constructed using a charged particle beam such as ions or protons. It is also possible.

In the description of the above-described embodiment, the mask 1 in which 100 holes 18A of (10 rows × 10 columns) are formed.
8 is irradiated with an electron beam and 100 beams passing through the hole 18A are used as a multi-beam MB. However, the present invention is not limited to this, and the number of beams, the size of the beams, or the arrangement thereof is arbitrarily determined. You can choose. Also, the method of forming the multi-beam MB is not limited to the method of branching the electron beam with the mask, and the beam may be formed individually using a plurality of electron beam sources.

[0055]

According to the present invention, as described above,
According to the invention described in Item 5, there is an unprecedented excellent effect that a defect of a pattern formed on a sample can be inspected at high speed and with high accuracy.

[Brief description of the drawings]

FIG. 1 is a diagram schematically showing a configuration of a defect inspection apparatus according to one embodiment.

FIG. 2 is a diagram showing an example of an irradiation pattern of a multi-beam irradiated on a sample.

3A and 3B are diagrams illustrating a multi-beam MB incident on an electromagnetic lens. FIG. 3A illustrates a multi-beam MB as a comparative example.
(B) is a multi-beam MB of the present embodiment with the same beam spacing as the comparative example, and (C) is a multi-beam MB of the present embodiment with the same number of beams as the comparative example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Defect inspection apparatus 12 Electron gun 14 Condenser lens 16 Condenser lens 18 Mask 20,22 Deflector 24,26 Electromagnetic lens 28 Aperture 30 Sample 34 Multi lens 34A, 34B, 34C, 34D Lens aperture 36 Multi aperture plate 36A, 36B, 36C , 36D Aperture 38A, 38B, 38C, 38D Detector CO Crossover MB Multi-beam (Electron beam) AX Optical axis

Claims (5)

[Claims] 1. A defect inspection apparatus that irradiates a charged particle beam to a sample on which a pattern is formed, detects charged particle beams from the sample, and inspects the pattern for defects. Means for forming a charged particle beam;
The positions at which the plurality of charged particle beams are formed and the sample are arranged at optical axis direction positions corresponding to the reduction ratio with a center at a crossover position which is a position divided at a predetermined reduction ratio. A charged particle beam optical system of a symmetric magnetic tablet type comprising two electron lenses having similar shapes; and a charged particle beam generated from the sample by the plurality of charged particle beams irradiated on the sample from the charged particle beam optical system. A defect inspection apparatus, comprising: a detection unit for detecting. 2. The apparatus according to claim 1, further comprising a charged particle beam scanning unit for scanning a plurality of charged particle beams irradiated on the sample between the charged particle beam forming unit and the charged particle beam optical system. The defect inspection apparatus according to claim 1. 3. The defect inspection apparatus according to claim 2, wherein the charged particle beam scanning unit is an even-numbered deflector arranged along the optical axis direction. 4. The charged particle beam forming means includes: a charged particle beam source for generating a charged particle beam; and a charged particle beam generated by the charged particle beam source. The defect inspection according to any one of claims 1 to 3, further comprising: an electron lens that forms a beam, and a mask that branches the charged particle beam to form a plurality of charged particle beams. apparatus. 5. The electronic lens disposed on the charged particle beam forming means side is shifted toward the crossover position by a predetermined distance K (K is a positive or negative real number), and the electron lens on the sample side is displaced. 5. The defect inspection apparatus according to claim 1, wherein the defect inspection apparatus is shifted toward the crossover position by a distance corresponding to a product of a predetermined distance K and a reduction ratio 1 / M. 6.