JP2007250223A - Scanning electron microscope - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To materialize high resolving power (low aberration) in an objective in which a sample is disposed on the further electron gun side than the inside magnetic pole and the outside magnetic pole of the objective. <P>SOLUTION: A keynote for designing the magnetic field type lens is how to strengthen B in a magnetic flux distribution B(z) on the optical axis of the lens, and how to thin the thickness of the lens (that is, the z width of the B distribution). For this purpose, a magnetic field is required to be concentrated on the vicinity of an electron beam irradiating portion on a sample by reviewing the shape and disposition of the inside magnetic pole and the outside magnetic pole of the objective 13. For this purpose, first the inside magnetic pole 13a is formed into a truncated cone as shown in Fig. 4, and the sample 12 is disposed near the base side of the truncated cone. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a scanning electron microscope (hereinafter abbreviated as SEM), and more particularly to an SEM technique capable of high-resolution observation at low acceleration.

  A conventional SEM employs a magnetic field type objective lens, and is devised to shorten the focal length of the lens from the viewpoint of high resolution.

  In Patent Document 1, the sample is arranged between the magnetic poles of the objective lens to shorten the focus. In the case of a thin film sample, it is disclosed that not only an SEM image but also a STEM (Scanning Transmission Electron Microscope) image is useful. In Patent Document 2, both the inner magnetic pole and the outer magnetic pole of the objective lens are arranged on the electron gun side with respect to the sample, and the sample is arranged in a leakage magnetic field (or magnetic flux) formed by both magnetic poles on the sample side. . In Patent Document 3, as shown in FIG. 2, both the inner magnetic pole 13a and the outer magnetic pole 13b of the objective lens 13 are arranged on the opposite side of the sample 12 from the electron gun. Thereby, the magnetic path gap is arranged so that the magnetic field intensity generated by the objective lens 13 is sufficiently strong on the sample. It is also disclosed that a semiconductor wafer sample or the like can be inclined by changing the sample mounting side portion of the objective lens from a planar shape to a convex shape. In any shape of the objective lens, there is no structure that blocks the flight of secondary electrons between the sample 12 and the secondary electron detector 10, so that the collection rate of secondary electrons is almost 100%. . Patent Document 4 discloses an objective lens similar to that shown in FIG. 2, that is, an objective lens that is installed on the opposite side of the sample from the electron gun and whose inner magnetic pole protrudes toward the sample. However, the sample-side metal plate connected to the outer magnetic pole is a non-magnetic material such as stainless steel, and does not have a structure that restricts the magnetic flux to the vicinity of the sample.

  In Non-Patent Document 1, a lens described as a snorkel lens as shown in FIG. 3 is a conventional objective lens related to the present invention, in order to increase the magnetic flux density near the end of the inner magnetic pole 13a. A frustoconical shape is disclosed.

Japanese Patent Application Laid-Open No. 2004-214065 JP-A-8-227678 JP-A-6-181041 USP 3,870,891 E. Munro and O.C. Wells (1976), Scanning Electron Microscopy / 1976 (Part 1) p.27

  An object of the present invention is to increase the resolution of an objective lens disclosed in Patent Document 3 and Non-Patent Document 1, that is, an objective lens in which a sample is arranged closer to the electron gun than the inner magnetic pole and outer magnetic pole of the objective lens ( (To reduce aberrations).

  The basic design policy of the magnetic field type lens for solving the above-mentioned problem is how to increase B in the magnetic flux density distribution B (z) on the optical axis of the lens and to increase the thickness of the lens (that is, the z width of the B distribution). It's all about thinning. For this purpose, it is necessary to review the shape and arrangement of the inner and outer magnetic poles of the objective lens 13 and concentrate the magnetic field only in the vicinity of the electron beam irradiation part on the sample. For this purpose, first, as shown in FIG. 4, the inner magnetic pole 13a is formed into a truncated cone shape, and the sample 12 is arranged near the truncated cone side of the truncated cone. As a result, the magnetic lines of force emerging from the frustoconical magnetic pole can be concentrated in the vicinity of the electron beam irradiation part on the sample. Next, in order to narrow the B (z) distribution width, the outer magnetic pole 13b has a tapered cross section that is thinned toward the inner magnetic pole, and is brought closer to the inner frustoconical magnetic pole so as to come out of the frustoconical magnetic pole. Thus, the magnetic field lines 20 that have passed through the sample are taken in so as not to protrude greatly toward the electron gun side. However, it is necessary to secure the thickness of the tapered portion to such an extent that magnetic saturation does not occur.

The combination of the inner frustoconical magnetic pole and the outer magnetic pole having a tapered cross-sectional shape can increase the lens magnetic flux density distribution B (z) and narrow the distribution z width, thereby realizing an unprecedented low aberration objective lens. This enables high-resolution SEM observation.

  An example of the present invention will be described below with reference to the drawings.

  The apparatus configuration of the scanning electron microscope according to the first embodiment of the present invention will be described (see FIG. 1). A voltage of 3 to 6 kV is applied to the field emission electron gun 2 to apply the extraction electrode 3, the electron beam 1 is extracted, and the acceleration electrode 4 accelerates (or decelerates) to 0.5 to 30 kV. The electron beam 1 is accelerated along the ideal optical axis (the trajectory of the electron beam when not deflected).

  The accelerated electron beam 1 is focused by the first focusing lens 5, and unnecessary portions of the beam are removed by the objective lens aperture 6. The electron beam 1 that has passed through the objective lens stop 6 is obtained by the first combination of the second focusing lens 7 and the first objective lens 11 or the second combination of the second focusing lens 7 and the second objective lens 13. Squeezed onto the sample 12. The narrowed electron beam 1 is scanned two-dimensionally on the sample by a two-stage deflection coil 8. Secondary electrons (or backscattered electrons) 14 emitted from the sample are accelerated toward the electron gun 2 by the secondary electron accelerating electrode 17 disposed in the first objective lens 11 without being greatly spread around the optical axis. After being pulled up, it is removed from the optical axis by the ExB filter 9 and detected by the secondary electron detector (or backscattered electron detector) 10. However, in the case of reflected electrons, to be precise, once the reflected electrons are lifted, they are collided with a conversion plate (not shown) to generate secondary electrons for conversion, and the converted secondary electrons are converted into the ExB filter 9. Therefore, the secondary electron detector (or backscattered electron detector) 10 detects the light beam off the optical axis.

  A detection signal detected by the secondary electron detector (or backscattered electron detector) 10 is amplified by an amplifier and used as a luminance signal of a scanning image synchronized with the deflection signal of the deflection coil 8, and is used as an SEM image. For example, a CRT or liquid crystal display screen) 22 is formed. The second combination of the second focusing lens 7 and the second objective lens 13 is used particularly in the case of the high-resolution observation mode at the time of low acceleration with an acceleration voltage of 0.5 to 5 kV. On the other hand, the first combination of the second focusing lens 7 and the first objective lens 11 is used in the normal observation mode at an acceleration voltage of 0.5 to 30 kV. In the normal observation mode at a low acceleration of 0.5 to 5 kV, the SEM observation at a low magnification can be performed. The first objective lens 11 and the second objective lens 13 are not operated simultaneously, and are switched by the mode switching means 20 for each mode.

  In the case where the thickness of the sample 12 is about 0.2 μm or less, an electron beam having a high acceleration voltage of about 15 to 30 kV or more passes through the sample 12. The transmitted electrons 15 are divided into scattered electrons having a large scattering angle and small non-scattered electrons with respect to the incident direction of the electrons, and are detected by a dark field detector and a bright field detector, respectively. The transmission electron detector 16 is a combination of a dark field detector and a bright field detector. If the detection signals from these are used as luminance signals, a STEM image similar to the SEM image can be obtained. The first inner magnetic pole 13a of the objective lens 13 is provided with a hole for passing the transmitted electron 15, the hole shape in the vicinity of the sample 12 is a cone, and the cone apex angle is in the range of 60 to 90 °. The first inner magnetic pole 13 has a magnetic pole end face formed toward the field emission electron gun 2 and has a magnetic pole end face perpendicular to the ideal optical axis.

  In a dark field image using a scattered electron beam, it has been empirically known that detection of scattered electrons having a wider angle is preferable as a scattering angle, but the top of the hole cone is restricted due to spatial interference with the outer magnetic pole 11b having a tapered cross section. The angle is optimally 60 ° to 90 °. In a SEM apparatus that does not incorporate STEM observation, of course, this conical hole is essentially unnecessary. However, in SEM observation of a thin film sample, noise generation due to backscattered electrons from the truncated cone pole of the transmitted electron 15 occurs. From the standpoint of inhibition, it is preferable to provide a cylindrical hole with a small diameter.

Next, axial magnetic flux density distribution B (z) Tai径D [mm] of the inner magnetic pole 13a in the distribution, the inner diameter d of the outer magnetic pole 13b [mm], the projection amount z _L relative to the inner magnetic pole 13a of the outer magnetic pole 13b [mm] Describe the impact of. The aforementioned parameters (D, d, and z _L ) and other parameters (outer diameter D _out [mm] and conical half apex angle α [°] of the inner magnetic pole 13a, outer diameter d _out [mm] of the outer magnetic pole 13b)) Is shown in FIG. FIG. 6 and FIG. 7 show typical B (z) distribution and d dependence of the spherical aberration coefficient Cs [mm] and the chromatic aberration coefficient Cc [mm], respectively. However, (α,
D, z _L ) = (45, 2, 0) and (D _out , d _out ) = (20, 60). The sample is placed at a position 0.5 mm (that is, z = −0.5) in front of the electron gun side with the base surface of the inner magnetic pole 13a as the z origin. d = 5, 6, 8, 12 corresponds to d / D = 2.5, 3, 4, 6 D _out and d _out are several times larger than D and d, and do not affect the characteristic curves of FIGS. Therefore, a scaling rule is established between Cs, Cc, d, and D, where, for example, if d / D is fixed to a specific value and D is doubled, Cs and Cc are also doubled. Will do. An important point in FIG. 6 is that the B (z) distribution curve is increased / decreased by d / D (2.5⇔6) at the position z = −2 mm, that is, with a point P away from the base surface of the truncated cone magnetic pole D as a fulcrum. ) Changes like a seesaw. As d / D becomes smaller, the B (z) distribution becomes steeper and the distribution width becomes narrower, and the aberration coefficients Cs and Cc become smaller as shown in FIG. From the standpoint of actual arrangement, d / D needs to be 4 or less in order to make the magnetic pole cross-section of the outer magnetic pole 13b thinner at the inner magnetic pole side and to make closest contact toward the inner truncated cone pole base. In order to make use of the low aberration characteristic of the truncated cone magnetic pole, D needs to be 6 mm or less. This is because, in the region D> 6 mm, (Cs, Cc) increases to (1.1, 2.0) even under the condition d / D = 2.5, which is equivalent to the conventional objective lens.

Next, FIG. 8 shows the z _L dependency of the B (z) distribution. However, d = 2 and other parameter values are the same as those in FIG. When z _L is in the range of −1 to +2, it is preferable that z ( _L ) distribution becomes steeper as z _L increases. When z _L = + 4, a shoulder peak occurs around z = −d of the B (z) distribution curve and the B (0) value also decreases, which is not preferable as a lens magnetic field. In the region of z _L > 0, the base surface of the truncated cone magnetic pole recedes from the outer tapered magnetic pole and becomes the bottom of the concave space, so that it is difficult to arrange the sample 12 close to the base surface of the truncated cone magnetic pole. Furthermore, as the arrangement of the sample 12 moves away from the base surface of the truncated cone magnetic pole, there is a disadvantage that the effective lens area of the B (z) distribution becomes smaller. In the region of z _L <0, the outer tapered magnetic pole is closest to the conical side surface of the inner frustoconical magnetic pole, so that B increases near the closest portion of both magnetic poles, and magnetic saturation is likely to occur. FIG. 9 is a representative example in which the B distribution in the outer magnetic pole 13b and the inner magnetic pole 13a with a conical hole is displayed in gray luminance when the lens coil 13c is 1000AT (product of coil turns N [T] and current I [A]). B increases in the vicinity of the closest part of both magnetic poles, and its maximum value Bmax is about 2.3T. In a permender often used as a magnetic pole material, the saturation magnetic flux density is about 2.4T. Since a strong excitation lens that exceeds the magnetic flux saturation of the material can be made, it is most preferable to set z _L to almost zero when the sample size and arrangement operability are balanced with the B intensity from a practical viewpoint.

A second embodiment will be described with reference to FIG. The inner magnetic pole 13a of the objective lens 13 is separated from the conical end side portion 13a1 and the base portion 13a3 through the inner magnetic pole electrical insulating portion 13a2. Similarly, the outer magnetic pole 13b is separated from the sample-side tapered magnetic pole 13b1 and its base portion 13b3 in terms of DC potential via the outer magnetic pole electrical insulating portion 13b2. However, the conical end side portion 13a1 and the tapered magnetic pole 13b1 are magnetically strongly coupled to the respective base portions. Both bases are set to ground potential, while a potential Vdecel is applied to the conical end side portion 13a1 and the tapered magnetic pole 13b1. Further, a potential Vdecel is applied to the sample 12 through a sample holder (not shown), and a mesh-like or plate-like potential shield plate 17 in which a hole for an incident electron beam is formed in the vicinity of the electron beam incident side of the sample 12. Is usually set to ground potential. With this potential arrangement, if the electron acceleration potential is Vacc, the electron beam is decelerated between the potential shield plate 17 and the sample 12, and the incident voltage on the sample 12 becomes Vacc-Vdecel. The lens action works even with this deceleration potential, and the aberration coefficients Cs and Cc have a feature that can be reduced. For example, when the potential shield plate 17 is placed 5 mm away from the cone electrode base surface and the sample 12 is placed 0.5 mm away from the objective electrode with z _L of 0, the aberration coefficients Cs and Cc are Vdecel / Vacc as shown in FIG. Behave like For example, when the incident energy of the electron beam to the sample 12 is 200 eV, (Cs) when the potential shield plate 17 does not work, that is, when (Vdecel, Vacc) = (0, −200) [unit: V]. , Cc) = (0.54, 1.1) [unit: mm], but when (Vdecel, Vacc) is (−1000, −800) by operating the potential shield plate 17, (Cs, Cc) is (0.30, 0.39) and (56, 35)%, respectively. In such a low-acceleration SEM observation, since this deceleration mode operation does not deteriorate the beam performance, high-resolution observation is possible.

  In order to achieve high resolution, the electron gun 2 needs to have high brightness, a small light source size, and a small energy width of emitted electrons, and in either of the first and second embodiments, the field emission type It was adopted. The field emission type is further classified into cold cathode field emission (CFE), Schottky, and carbon nanotube (CNT). The CFE was used for the highest resolution, while the Schottky was used for the highest temporal stability. CNT is currently under development as a combination of both, and practical reliability is still insufficient.

The figure explaining schematic structure of the scanning electron microscope of 1st Example of this invention. Sectional-structure schematic in the scanning electron microscope of a prior art well-known example. Sectional-structure schematic in the scanning electron microscope of a prior art well-known example. The figure explaining the means employ | adopted in order to solve the subject of this invention. The figure explaining the parameter of the truncated cone magnetic pole structure objective lens of this invention. The figure explaining the dependence of the outer side taper magnetic pole internal diameter d of the axial magnetic flux density B (z) in a truncated cone magnetic pole structure objective lens. The figure explaining the dependence of the outer side taper magnetic pole internal diameter d of the aberration coefficients Cs and Cc in a truncated cone magnetic structure objective lens. The figure explaining the _zL dependence of the axial magnetic flux density B (z) in a truncated cone magnetic pole structure objective lens. FIG. 10 is a gray luminance display diagram of B distribution in an outer tapered magnetic pole and an inner truncated cone magnetic pole with a conical hole in a truncated cone magnetic objective lens. The figure explaining the objective lens of a truncated cone magnetic pole structure in the scanning electron microscope of 2nd Example of this invention. The figure explaining the Vdecel / Vacc dependence of the aberration coefficients Cs and Cc.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Electron beam, 2 ... Electron gun, 3 ... Extraction electrode, 4 ... Acceleration electrode, 5 ... 1st focusing lens, 6 ... Objective lens aperture, 7 ... 2nd focusing lens, 8 ... Deflection coil, 9 ... ExB Filter, 10 ... Secondary electron detector (or backscattered electron detector), 11 ... First objective lens, 11a, 13a ... Inner magnetic pole, 11b, 13b ... Outer magnetic pole, 11c, 13c ... Lens coil, 12 ... Sample, DESCRIPTION OF SYMBOLS 13 ... Objective lens, 14 ... Secondary electron (or reflection electron), 15 ... Transmission electron, 16 ... Transmission electron detector, 17 ... Potential shield liquid, 20 ... Switching means for every mode, 21 ... Electric system control part, 22 ... Image display device.

Claims (11)

In a scanning electron microscope provided with an electron gun and an objective lens that focuses an electron beam emitted from the electron gun,
A sample stage for disposing the sample between the electron gun and the inner magnetic pole; and the objective lens has a magnetic pole end surface along the ideal optical axis of the electron beam and toward the electron gun. An inner magnetic pole to be formed and an outer magnetic pole for forming a magnetic field for focusing the electron beam between the inner magnetic pole and a direction perpendicular to the ideal optical axis of the electron gun side magnetic pole end of the inner magnetic pole Scanning electrons characterized in that the diameter D is 6 mm or less and the relationship between the electron beam passage opening diameter d of the outer magnetic pole and the diameter D is d / D ≦ 4 microscope. In a scanning electron microscope comprising an objective lens for focusing an electron beam from an electron gun onto a sample, and observing the sample using a signal generated from the sample,
Both the inner magnetic pole and the outer magnetic pole constituting the objective lens 1 are opposite to the electron gun side with respect to the sample,
The inner magnetic pole has a truncated cone shape at the sample side end,
On the other hand, in the outer magnetic pole, the cross section of the magnetic pole including the lens axis has a tapered shape toward the inner frustoconical magnetic pole, and the vicinity of the sample side end of the frustoconical magnetic pole is the closest location. microscope.   When the outer diameter and inner diameter of the sample side end of each of the inner magnetic pole and the outer magnetic pole of the objective lens 1 are represented by D and d, D is 6 mm or less and d / D is 4 or less. The scanning electron microscope according to claim 2.   The scanning electron microscope according to claim 2, wherein a distance on the optical axis between both end faces of the inner magnetic pole and the outer magnetic pole is equal to or smaller than an outer diameter D of the end face of the inner magnetic pole. It also has an objective lens 2 that is different from the objective lens 1,
The scanning electron microscope according to claim 2, wherein both the inner magnetic pole and the outer magnetic pole constituting the objective lens 2 are on the electron gun side with respect to the sample.   3. The scanning electron microscope according to claim 2, wherein when the sample is thin, the first inner magnetic pole of the objective lens 1 has a hole for transmitting a scattered scattered beam from the sample.   In the hole for transmitting a scattered scattered beam from the sample provided in the first inner magnetic pole of the objective lens 1, the hole shape in the vicinity of the sample side of the hole is a cone, and the apex angle of the cone is 60 ° to 90 °. The scanning electron microscope according to claim 5, wherein the scanning electron microscope is in a range.   6. The electron beam irradiation mode for irradiating the sample is divided into two, and switching means is provided to operate the objective lens 1 and the objective lens 2 in accordance with each mode. Scanning electron microscope.   The scanning electron microscope according to claim 5, wherein a detector for secondary electron signals generated from the sample is installed on the electron gun side with respect to the outer magnetic pole of the objective lens 2.   The scanning electron microscope according to claim 2, wherein the sample side portions of the inner magnetic pole and the outer magnetic pole of the objective lens 1 are separated from their base portions in terms of DC potential.   The scanning electron microscope according to claim 2, wherein the electron gun is a field emission type.

Images