JP2003324061A - Electron beam equipment and method for manufacturing device using the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide electron beam equipment capable of forming an image with a large S/N ratio in a short period of time and precisely correcting the position of a wafer stage. <P>SOLUTION: In electron beam equipment 100 comprising an electron gun 104 having a thermionic emission cathode, an objective lens 109 for converging the electron beam emitted from the electron gun on a sample W such as a wafer, and a deflector for making the electron beam scan the surface of the sample W; the deflector is composed of at least two stages of deflectors 108 and 110 arranged between the electron gun 104 and the objective lens 109, an electromagnetic deflector 108 is arranged as the deflector on the side near the objective lens 109, and the secondary electron is detected also by the electron beam scan on a region which deviates from an optical axis OA<SB>1</SB>. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention has a minimum line width of 0.1.
The present invention relates to an apparatus for high-throughput evaluation of a sample having a pattern of μm or less, and the present invention relates to a device manufacturing method using these apparatuses.

[0002]

2. Description of the Related Art Conventionally, in an electron beam apparatus for detecting a secondary electron by applying a negative voltage to a sample, it is considered that an E × B separator deflects the secondary electron toward the detector. It was It is also known from theory that shot noise is significantly reduced when the hot cathode is operated in the space charge limited region.

However, even if the conventional electron beam apparatus as described above requires an E × B separator for deflecting secondary electrons, the E × B separator is not only complicated in structure, Since the adjustment takes time, there is a problem that it takes a long time to start up the electron beam apparatus.

[0004]

SUMMARY OF THE INVENTION The present invention has been proposed in order to solve these conventional problems, and one object of the present invention is to enable highly accurate wafer stage position correction. To provide a simple electron beam apparatus. Another object of the present invention is to provide an electron beam apparatus having a simplified structure by simplifying the E × B separator. Still another object of the present invention is to provide a device manufacturing method using such an electron beam apparatus.

[0005]

To achieve the above object, in one embodiment, the present invention provides an electron gun having a thermionic emission cathode and an electron beam emitted from the electron gun on a wafer. An electron beam apparatus including an objective lens that converges on a sample, and a deflector that scans the surface of the sample with the electron beam, wherein the deflector is provided between the electron gun and the objective lens. An electromagnetic deflector is arranged as the deflector on the side closer to the objective lens so that the electron beam scans a region deviated from its optical axis. A characteristic electron beam apparatus is provided.

In one embodiment, the electron beam apparatus includes a stage for mounting the sample, a movable mirror provided on the stage, a fixed mirror provided on the objective lens, and a laser. A laser oscillation / reception unit for transmitting and receiving light is further provided, and the relative position of the stage can be calculated.

It is desirable that the fixed mirror be fixed to the objective lens via ceramics having an expansion coefficient of substantially zero. In addition, the fixed mirror is formed by forming a SiC film on the surface of SiC ceramics and then mirror-polishing the reflection film, or by forming a film on the end surface of the ceramics stage with ceramics of the same material and then mirror-polishing. A reflective film formed by the above can be provided.

Further, in the electron beam apparatus, the deflector may include an electrostatic deflector, and the electrostatic deflector is formed by a ceramic member having an inner surface of a cylindrical shape and an axial direction of the ceramic member. A plurality of grooves, a plurality of deflection electrodes electrically formed on the end surface and the inner surface of the ceramic member and the surfaces of the plurality of grooves, and electrically insulated from each other,
Can be provided.

The electrostatic deflector may include a shield member that is disposed close to the opposing end faces thereof, at least a surface of which is a metal, and the end face and the face of the shield member that faces the end face. It is desirable that L / D> 4 where D is the distance and L is the shortest distance from the inner surface of the ceramic member to the groove.

It is desirable that the electron gun operates under a space charge limiting condition where the shot noise reduction coefficient is smaller than 1.0.

In order to evaluate the sample with the electron beam apparatus, the beam diameter of the electron beam is secondary when the electron beam is scanned in a direction orthogonal to the line-and-space pattern of the sample. S of signal obtained from electron beam
It is preferable to set the beam size so that the / N ratio is maximized.

The sample has a pattern formed by connecting visual fields having a size smaller than the chip size, and the evaluation is performed by aligning the boundary portion between the adjacent visual fields with the center of the visual field of the electron beam apparatus. It is desirable to do.

Further, the present invention provides a device manufacturing method characterized in that the electron beam apparatus is used to evaluate a wafer during the process.

In another embodiment of the present invention, a mask arranged in the vicinity of a wafer is scanned by an electron beam emitted from an electron gun having a thermionic emission cathode, and the pattern of the mask is changed to the above-mentioned pattern. An electron beam apparatus for transferring onto a wafer, characterized in that the electron gun is operated under a space charge limiting condition with a shot noise reduction coefficient of less than 1.0.

In this electron beam apparatus, a mark is provided on the dicing line of the wafer, a mark hole is provided in the mask so as to correspond to the mark, and the mark is formed by the electron beam passing through the mark hole. Marks can be detected by scanning. In this case, it is desirable that the electron beam is scanned while tilting with the mask as a fulcrum.
The detection of the mark is performed by the electron beam that passes through a detection hole provided near the mark hole.

Further, in this electron beam apparatus, when the wafer is placed on the wafer stage, information regarding the position of the wafer stage is fed back to the electron beam deflector to deflect the electron beam. It is desirable to adjust.

[0017]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1A is a diagram schematically showing one embodiment of an electron beam apparatus according to the present invention.
In the figure, the electron beam apparatus 100 includes a cathode 101,
It has an electron gun 104 with an anode 102 and a Wehnelt 103. The primary electron beam emitted from the electron gun 104 is axially aligned by the axial alignment device 105 and is incident on the condenser lens 106. An astigmatism corrector 107 for correcting astigmatism generated in the primary electron beam is arranged between the electron gun 104 and the condenser lens 106. The condenser lens 106 forms a crossover at a position slightly closer to the wafer W side of the deflector 108. This crossover is reduced by the objective lens 109, and a small-sized beam is focused on the wafer W.

The deflector 108 comprises an electrostatic deflector 108a and an electromagnetic deflector 108b, and for evaluation of the wafer W,
The primary electron beam emitted from the electron gun 104 is the electrostatic deflector 108a, the electromagnetic deflector 108b, and the electrostatic deflector 11.
The wafer W is scanned by 0. That is, as shown in FIG. 1B, the primary electron beam has an elongated visual field 111 extending in the X direction, and the electromagnetic deflector 108b and the electrostatic deflector 110 cause the primary electron beam to appear in FIG. As indicated by a thin solid line, the beam is deflected to a position deviated from the optical axis OA ₁ of the primary electron beam, and further electrostatic deflectors 108a and 110 cause X.
Scanned in the direction.

The scanning of the wafer W in the Y direction by the primary electron beam is performed by continuously moving the wafer stage 112 in the Y direction. Wafer stage 11
The movement of 2 is constantly monitored by the laser measuring system 113. In order to realize this, a laser moving mirror 114 is installed on the wafer stage 112, and a first laser fixed mirror 115 is installed on the outer surface of the objective lens 109, and laser oscillation / reception is performed at a position where the laser moving mirror 114 can be irradiated. The section 116 is installed. Further, a half mirror 117 and a second fixed mirror 118 are provided on the optical path of the laser light.

With this structure, the laser light emitted from the laser oscillation / reception unit 116 is transmitted to the half mirror 11.
It is bisected at 7, and a part of it is irradiated onto the laser moving mirror 114 and reflected there. The remaining laser light is reflected by the second laser fixed mirror 118 and the first laser fixed mirror 1
It irradiates 15 and is reflected there. Laser moving mirror 114
The laser light reflected by the laser beam passes through the half mirror 117 and returns to the laser oscillation / reception unit 116, and the laser light reflected by the first laser fixed mirror 115 is reflected by the second laser fixed mirror 118 and the half mirror 113. It is reflected and returns to the laser oscillation / reception unit 116. The laser oscillating / receiving unit 116 causes interference caused by a difference between the optical path length of the laser beam between the laser moving mirror 114 and the half mirror 117 and the optical path length between the first laser fixed mirror 115 and the half mirror 117. The relative position of the wafer stage 112 can be measured from the intensity of the received light intensity.

The laser movable mirror 114 is made of a material having a high rigidity, for example, SiC, so that the cross-sectional dimension can be achieved even when the laser movable mirror 114 is mounted on a wafer stage on which a large-diameter wafer such as 12 inches is mounted. Can be made smaller. Further, when the laser moving mirror 114 is made of ceramics, even if there is a bore in the ceramics, the laser moving mirror 114 becomes flat when the film is formed, so that it can be polished to a mirror surface.

The wafer W is scanned by the primary electron beam.
The secondary electron beam emitted from the secondary electron beam is accelerated by the objective lens 109 as shown by a dotted line in FIG.
The light enters the objective lens 109 at a position deviated from the optical axis OA ₁ and is refracted by the objective lens 109. Then, the secondary electron beam is further deflected by the electromagnetic deflector 108b, and proceeds along the optical axis OA ₂ secondary electron beam detector 11
9 is detected, converted into an electric signal having a magnitude proportional to the detected intensity, and image processed by the image data processing unit 120.
That is, the electric signal output from the secondary electron beam detector 119 is stored in the image storage device under the control of the CPU in the image data processing unit 120 and is displayed on the monitor, and the same pattern of the standard pattern or different wafers is displayed. Evaluation such as defect detection is performed by comparing with the image data of the die.

In FIG. 1A, the numeral 121
Indicates a lens barrel that houses the electron gun 104, the primary optical system, and the secondary optical system inside, and 122 indicates an electrode.

Hereinafter, the electron beam apparatus 10 shown in FIG.
Some of the components used for 0 will be specifically described. First, FIG. 2 shows the objective lens 1 shown in FIG.
The specific structure of the electrostatic lens that constitutes 09 is shown in detail. The objective lens 109 is formed in an axially symmetrical structure about the optical axis OA ₁ , and FIG. 2 shows a right half sectional view thereof.

The objective lens 109 is created as follows. First, a cylindrical part 20 brazed by embedding a metal rod 201 in a ceramic material that can be cut.
Create 2. Then, the ceramic material is lathe-processed so as to form the upper electrode portion 203, the central electrode portion 204, the lower electrode portion 205 and the axisymmetric electrode portion 206. Next, the exposed surface of the ceramic material is masked for insulation, and the remaining surface is subjected to metal plating by electroless plating to form the upper electrode 203 ′ and the central electrode 20.
4 ', a lower electrode 205' and an axisymmetric electrode 206 'are formed.

The upper electrode 203 'is supplied with a voltage from a lead wire 207 connected to its upper surface. The center electrode 204 'and the lower electrode 205' include a pair of metal rods 20.
The voltage from the lead wire 208 is supplied via 1. The metal rod 201 does not need to be vacuum-sealed. The lead wire 2 connected to the lower surface of the axisymmetric electrode 206 'is
The voltage is supplied from 09.

A ceramic cylindrical part having such a structure is made small and has a low coefficient of linear expansion outside thereof.
For example, a ceramic member 210 having a zero expansion coefficient (for example, NEXCERAN113 manufactured by Nippon Steel Corporation) is attached. Then, a flat laser fixing mirror 115 is adhesively fixed to the outside of the ceramic member 210. The laser fixed mirror 115 may be formed by polishing the side of the ceramics member 210 on which the laser beam is applied to a flat mirror surface.

The laser fixed mirror 115 is attached to the objective lens 109.
By integrating (bonding, fixing, or forming an integrated structure) with the wafer stage 112, not only the vibration of the wafer stage 112 but also the deviation of the primary electron beam due to the vibration even when the optical system vibrates in the XY plane direction. It is possible to correct the electron beam position by measuring with the laser length measurement system 113. That is, even if the objective lens 109 vibrates in the XY directions,
The change in the relative distance to the stage 112 is due to the laser measuring system 1
Since it is measured at 13, the primary electron beam can be corrected so as to cancel the change in the position of the objective lens 109. This allows the optical system and the wafer stage 112
It is possible to correct the relative minute vibration with respect to and to reduce the image distortion due to the vibration of the optical system.

The laser fixed mirror 115 installed on the objective lens 109 and the laser moving mirror 114 installed on the wafer stage 112 are manufactured by the manufacturing process shown in FIG. In FIG. 3, first, SiC ceramics is processed to have a cross section of 30 mm × 30 mm and a length of 35 cm (step 301). The laser reflection surface is a rough surface, but has a good flatness, and is ground into fine ground glass (step 302). Then, a CVD apparatus is used to form a SiC film to such an extent that the recesses on the reflecting surface due to the voids formed inside and the rough surface are sufficiently filled (20 μm in one example) (step 303). At this time,
It is desirable to form the film over a long period of time with the vertical line and the reflecting surface tilted at an angle of approximately 45 degrees as shown in the drawing so that the dents and the like can be efficiently filled. Then, mirror polishing is performed (step 304). Since the surface before the CVD is attached has a fine ground glass shape, the substrate and the CVD film are not peeled off during polishing. After mirror-polishing, a multi-layer film or a reflection film made of titanium, gold or the like is formed (step 305).

Next, FIGS. 4 and 5 will be described. As described above, in the electron beam apparatus shown in FIG. 1, the electrostatic deflector and the electrostatic lens are used as the elements forming the electron optical system. For example, in an electrostatic deflector, it has recently been proposed to use a film-like electrode (plating electrode) formed on an insulator by a surface treatment such as plating, instead of a metal electrode. The electrostatic deflector using the plated electrodes does not require screwing the electrodes, and has the advantage of reducing the number of parts and downsizing.

However, in an electrostatic deflector using a plating electrode, if the voltage application wiring is directly fixed to the plating electrode by screwing, there is a possibility that a hole may be opened on the surface of the plating electrode. When the plating electrode has a hole, the electrostatic field distribution in the space through which the charged particle beam passes is distorted, and the deflection control for the charged particle beam cannot be performed with high accuracy. Therefore, in the electrostatic deflector, the supporting portion of the insulator on which the plating electrode is formed is projected together with the plating electrode from the end portion of the insulating outer cylinder, and the electric field applying wiring is connected to this protruding portion. There are some that prevent holes from being formed on the surface of the plated electrode. However, this wiring structure is complicated, and the coating (insulator) of the voltage application wiring connected to the protruding portion may be visible from the gap between the adjacent plating electrodes.

It is desired that other electron beam control elements (electrostatic lens, etc.) be formed by using plating electrodes as in the above electrostatic deflector. It is desired to devise how to connect the voltage application wiring to the electrodes.

In response to such a demand, the voltage application wiring can be connected to this electrode with a simple structure while maintaining the surface of the electrode formed on the insulator with high precision by surface treatment such as plating. An example of an electron beam control element that can be used as an electrostatic deflector or an electrostatic lens is schematically shown in FIGS. 4 and 5. 4 is a top view thereof, and FIG. 5 is a vertical sectional view thereof.

4 and 5, an electron beam control element 400 such as an electrostatic deflector has a base 401 made of an insulating material. The base portion 401 has a cylindrical shape centered on the axis A, and has a structure defined by an outer surface 402, end surfaces 403 and 404, and a through hole 406 that forms an inner surface 405. In use, the axis A is aligned with the optical axis of the electron beam and the through hole 406
Is coaxial with axis A. Grooves 407 for separating the electrodes are radially formed in the cylindrical base portion 401 in a direction parallel to the axis A. As shown in FIG. 4, each groove 40
7 has the same bent shape, and a circular through hole 408 is formed at the end of each groove 407.

Further, as shown in FIG. 5, a conductor disc for shielding having a through hole having the same diameter as the through hole 406 above and below the electron beam control element 400, that is, near the facing end faces 403 and 404. 501 and 502 are installed, respectively.

In such a structure, the inner surface 405 and the opposing end surfaces 403 and 404 are metal-coated except for the non-coated surface 409 for insulation.
Specifically, a plurality of electrodes 410 separated from each other by a groove 407 are formed on the inner side surface 405, and a plurality of electrodes 410 electrically connected to the electrodes 410 are also formed on the respective end surfaces 403 and 404. The conductor portion 411 is formed. Furthermore, from the electrode 410 formed on the inner surface 405 to the through hole 4
The inner surface of the groove 407 extending to 08 is also coated with a metal, as indicated by the diagonal lines. However, the inner surface of the through hole 408 and the surface extending from the through hole 408 to the outer surface 402 are not metal-coated and are non-coated surfaces 409. Thus, each groove 407 and uncoated surface 409
A plurality of conductive portions are formed by the electrode 410 and the conductor portion 411 that are electrically separated by and electrically connected to each other.

In the structure shown in FIG. 4, the groove 407 has eight grooves.
Therefore, eight electrodes 410 and eight conductor portions 411 that are electrically connected to each other are also formed. It should be noted that, if necessary, the conductor portion 411 may also be provided on a part of the outer side surface 402.
You may make it form the conductor part electrically connected with.

Wiring 412 for each electrode 410
Is a thin wire, which is bonded to the outer surface 402 or one of the end surfaces 403 and 404. When the wiring 412 is taken out from the outer side surface 402,
The outer diameter of the electron beam control element 400 becomes large, and when the wiring 412 is taken out from the end face 403, an extra space in the direction along the optical axis A of the electron beam control element 400 is required. 4 and 5 show the wiring 412 on one end face 40.
The example taken out from 3 is shown.

In the surface including the axis A and the opposing through hole 408, the distance between the shield disks 501 and 502 and the end surfaces 403 and 404 of the base portion 401 is D, and the surface of the electrode 410 on the side closer to the axis A. L / D> 4., Where L is the radial distance between the surface of the through hole 408 on the side closer to the axis A.
It is desirable to set it to 0. As a result, when the inner surface of the through hole 408 of the base portion 401 is charged, the influence of the potential generated by the charging on the electron beam passing near the axis A can be suppressed to 1/1000 or less.

As can be seen from the above description, in the electron beam control element shown in FIGS. 4 and 5, the surface of the electrode formed on the insulator by the surface treatment such as plating is highly accurately maintained on the electrode. On the other hand, it is possible to connect the voltage application wiring with a simple configuration, and it is possible to realize not only downsizing and cost reduction of the element and the electron beam apparatus but also high accuracy of the trajectory control of the electron beam control element. .

FIG. 6 is a diagram for explaining a region for driving the electron gun 104 of the electron beam apparatus 100 of FIG. Electron gun 10
4 operates as a thermionic beam source, and an electron emitting material (emitter)
Is LaB ₆ . Other materials can be used as long as they have a high melting point (low vapor pressure at high temperature) and a low work function. Two methods can be used to obtain a plurality of electron beams. One is a method of obtaining a plurality of electron beams by emitting one electron beam from one emitter (one protrusion) and passing it through a thin plate (aperture plate) with a plurality of holes. The above method is a method of forming a plurality of protrusions on the emitter and emitting one electron beam from each protrusion, thereby emitting a plurality of electron beams as a whole. In either case, the property that the electron beam is emitted from the tip of the protrusion with high brightness is utilized. Other types of electron beam sources such as thermal field emission type electron beams can also be used.

The thermoelectron beam source is a system that emits electrons by heating an electron emitting material. The thermal field emission electron beam source is a system that emits electrons by applying a high electric field to the electron emitting material. Further, the electron emission is stabilized by further heating the electron beam emitting portion.

Most of the shot noise of the secondary electron beam is shot noise of the primary electron beam, and the shot noise of the secondary electron beam can be reduced by reducing the shot noise of the primary electron beam. Therefore, the electron gun 104 is configured so that the S / N ratio of the detection signal of the secondary electron beam can be set to a required magnitude even if the irradiation amount of the primary electron beam is small. The means for reducing the shot noise of the primary electron beam will be described below. When the electron gun is determined by the cathode temperature, that is, when the electron gun is operating in the temperature limited region, the shot noise i _n emitted by the electron gun is expressed by the following equation (P.471, "Communication Engineering Handbook" edited by The Institute of Electrical Communication).
(1957)).

[0044]

## EQU1 ## i _n ² = 2eI _p B _f (1) In the equation (1), i _n ² is the root mean square value of the noise current, e is the electron charge, I _p is the anode direct current, and B is _f is the frequency band of the signal amplifier. If the electron flow is in the space charge limited region,

[0045]

## EQU2 ## i _n ² = Γ ² 2eI _p B _f (2) In Expression (2), Γ ² is a shot noise reduction coefficient and is a value smaller than 1.

When the cathode temperature is sufficiently high, Γ ² is
The minimum is about 0.018, and the noise current is reduced to 13% of that in the temperature limited region. Assuming that the S / N ratio in this case is a secondary electron beam≈a primary electron beam,

[0047]

Equation 3] _{S / N = I p / {} Γ (2e · I p · B f) 1/2} = 1 / Γ · {I p / (2e · B f)} 1/2 = n 1/2 / (Γ · 2 ^1/2 ) (3) When Γ = 0.13, the following S / N ratio is obtained from the equation (3).

[0048]

## EQU00004 ## S / N = 7.7 (n / 2) ^1/2 (4) (n: number of secondary electrons / pixel) That is, the electron gun operating in the space charge limiting region is This is equivalent to detecting the number of secondary electrons per pixel by 59 times (= 1 / Γ ² = 1 / 0.13 ² ) more than that of an electron gun (TFE). Since the latter has a brightness that is about two orders of magnitude higher than the former, assuming the same optical system with the same beam diameter, the latter may be able to obtain a beam current that is two orders of magnitude higher than the former, but an optical system suitable for the former By newly designing, a beam current that is one digit different from the former can be obtained. S / N
The ratio is 1/55 of the former ratio. In other words, with the electron gun in the space charge limited region, the measurement time is 10 / 55≈0.18 times and the dose is 1/55 as compared with the electron gun in the temperature limited region.

Whether the electron gun is operating in the space charge limited region can be checked by the method described below with reference to FIG. FIG. 6A shows a curve 601 representing the relationship between the electron gun current and the cathode heating current,
In the figure, a region Q ₁ is a region where the electron gun current hardly increases even if the cathode heating current is increased, and this region Q ₁ is a space charge limiting region. Also, FIG. 6 (B)
Is a curve 6 representing the relationship between the electron gun current and the anode voltage.
02, the region Q ₂ is a region where the electron gun current increases rapidly when the assault voltage is increased, and this region Q ₂ is also a space charge limiting region.

From the above, the cathode heating current of the electron gun is increased to measure the electron gun current, and the region Q ₁ where the electron gun current is saturated or the anode voltage of the electron gun is increased. The electron gun current is measured, and if the area is where the electron gun current is changing rapidly, it can be determined that the electron gun is operating in the space charge limited area. Therefore, conditions for operating the electron gun in the space charge limited region can be set.

As described above, the electron gun 104 is set with the heating current or the anode voltage (the voltage applied to the anode 102) so as to operate in the space charge limited region.
Although not shown, the cathode 101 of the electron gun 104 is formed of single crystal LaB ₆ , and has, for example, nine protrusions whose pyramid-shaped tip is arranged on the circumference, although not shown. The tips of these protrusions have a radius of curvature of about 30 μm.
Since the primary electron beam is emitted only from the vicinity of the apexes of these pyramidal protrusions, with a relatively large electron gun current of about 1 mA, at a voltage of 1 kV, 1 × 10 ⁴ A / cm ² sr (1
A brightness of kV) is obtained.

Next, selection of the beam diameter will be described with reference to FIGS. 7 and 8. Conventional electron beam devices have difficulty in inspecting fine patterns, coarse patterns, and even when the pattern size of the inspection target changes, it is difficult to change the pixel size, so the pixel size is too small depending on the inspection target. Therefore, it takes a long time to perform the inspection, and on the contrary, the pixel size is too large to obtain sufficient resolution. Further, the conventional electron beam apparatus does not make much use of the advantage that the beam current increases when the beam diameter is increased when the beam diameter is increased, and the loss of the S / N ratio when the beam diameter is increased. Had the drawback of being large.

In order to solve this, when the primary electron beam is focused on the surface of the wafer W and the secondary electron beam emitted from the wafer W is imaged on the detector 119, the primary optical system is used. Particular attention should be given to minimizing the effects of distortion and field curvature that occur. Regarding the relationship between the intervals between the plurality of primary electron beams and the secondary optical system, by separating the intervals between the primary electron beams by a distance larger than the aberration of the secondary optical system, crosstalk between the plurality of beams can be prevented. It can be lost.

FIGS. 7A and 7B show layouts of standard marks 701 and 702 of a plurality of pattern sizes mounted on the wafer stage 112 for evaluation of the wafer W, respectively. In the figure, the standard mark 701 is an L & S (line and space) pattern of 0.05 μm, and the standard mark 702 is 0.
It is an L & S pattern of 1 μm. As described above, several kinds of standard marks corresponding to the line width of the evaluated pattern are set on the wafer stage 112. 7A and 7B show two typical standard marks among them.

Before performing inspections, etc., the wafer stage 1
12 is moved to a standard mark matching the size of the pattern to be detected on the wafer W, for example, 701 in FIG.
To match the optical axis OA ₁ of the primary optical system, and then
By changing the bias voltage applied to the Wehnelt 103 of the electron gun 104, the beam brightness from the electron gun 104 is changed, the beam diameter is changed, and the optimum beam diameter or beam current according to the size of the pattern to be detected is selected. If the bias of Wehnelt 103 is made shallow, the electron gun 10
The current of No. 4 increases, the brightness increases, and the current of the multi-beam increases. When the beam current of the multi-beam becomes large, the beam diameter becomes large due to the space charge effect.

As another method of changing the beam diameter, there is a method of operating a reduction lens or an objective lens as a zoom lens. Also in this case, since the beam reduction ratio is adjusted to approach 1 in order to increase the beam current, the beam diameter also increases. However, since the beam intervals of the multi-beams also change at the same rate at the same time, if it is not desired to change the beam intervals, it is preferable to adopt the method of changing the bias of the Wehnelt 103 as described above.

(A1) to (A3) and (B1) to FIG.
(B3) shows waveforms obtained by observing an electric signal output from the secondary electron detector 119 with a monitor (not shown) when the standard marks 701 and 702 are scanned with a multi-beam. (A3) shows a waveform when the standard mark 701 is scanned with various beam sizes changed, and (B1).
(B3) to (B3) show waveforms when the standard mark 702 is scanned while changing the beam size.

FIG. 8A1 shows an output waveform 801 when the beam diameter is made larger than the line width, and a beam having a size larger than the line width is used for scanning despite the large beam current. Therefore, the signal contrast S is not so large, and since the beam current is large, the noise N is a large value, and the S / N ratio is approximately 3.4.

FIG. 8A3 shows an output waveform 803 when the beam diameter is too small compared to the line width, and a faithful waveform (close to a rectangular wave) appears, but the beam current is Since it is small, the signal contrast S is not large. Further, the noise N is small corresponding to the small beam current, and S /
The N ratio is approximately 6.25.

FIG. 8A2 shows the output waveform 802 when the beam diameter is just right. In this case, the beam blur is a reasonable value, the beam current is relatively large, and the contrast is large. S signal is obtained and S / N
The ratio is approximately 12.2.

Which of the beam diameters (A2) and (A3) in FIG. 8 should be selected is large (contrast / noise).
You can choose the one who can obtain the ratio. In the example shown in FIG. 8, when the standard mark 701 is used, (A2) in FIG.
It suffices to select the beam diameter at which the output waveform of 1 is obtained.

When the standard mark 702 is used, the same calibration is performed, and the beam size or beam diameter suitable for this line width is selected. In the example shown in FIG. 8, the beam diameter at which the output waveform (B2) is obtained may be selected.

In this way, the beam is adjusted so that the S / N ratio of the electric signal output from the detector 119 and having an amplitude proportional to the intensity of the secondary electron beam is maximized in accordance with the pattern size to be evaluated. The diameter or beam current may be selected. More specifically, regular standard patterns having different pitches are arranged on the wafer stage 112, and a signal waveform obtained when the regular standard pattern is scanned by the primary electron beam is used to detect the electric signal. Calculate the amplitude (S). At the same time, the noise amplitude (N) is also obtained, and the S / N ratio is calculated. The electron beam apparatus 100 can set a plurality of beam diameters of the primary electron beam, and scans a regular pattern having a pitch twice the minimum line width of the pattern to be evaluated with these beam diameters. S / N ratio of
Select the beam diameter that maximizes the / N ratio. As a result, a high S / N ratio can be evaluated for all the patterns to be evaluated.

As the regular standard pattern, instead of the above standard mark, a standard pattern may be searched for from the wafer to be inspected and used, and the S / N ratio may be similarly investigated. . In the above method, the primary electron beam does not necessarily have to be a multi-beam, and can be applied to pattern evaluation when scanning with a single beam.

As a result, in the electron beam apparatus 100,
A desired S / N ratio can be obtained even if the scanning speed is increased, and a high S / N ratio can be obtained without averaging. Also, depending on the evaluated pattern, S /
Since the beam diameter or the beam current can be selected so that the N ratio becomes maximum, high resolution and high throughput can be realized regardless of the size of the evaluated pattern.

Next, one specific example of the method of inspecting the circuit pattern formed on the wafer W which is the substrate will be described with reference to FIG. FIG. 9 shows a typical example of forming a circuit pattern on a wafer using electron beam lithography. Wafer W has a dimension in the X direction of, for example, 5 m
It is divided into a plurality of stripes 901 of m width and extending in the Y direction, and is placed on the wafer stage 112. In that state, by continuously moving the wafer stage 112 along the stripe 901 in the Y direction,
The mask pattern is transferred to the wafer W.

One stripe has a dimension of 250 μ in the Y direction.
A main visual field 902 having a dimension of 5 mm in the m and X directions is divided, and the inside of one main visual field 902 is divided into a sub visual field 903 of 250 μm square, and transfer is performed in sub visual field units. That is,
One mask is prepared for each of the plurality of sub-fields 903 forming the main field of view 902, and the mask pattern is transferred by scanning the beam by one sub-field of view.

When the mask pattern is formed on the wafer W in this manner, the location where defects are most likely to occur is the boundary between two adjacent stripes 901, and the location where defects are likely to occur next is adjacent to each other. It is the boundary between the matching main visual fields 902, and the third defect is most likely to occur at the boundary between the adjacent sub visual fields 903. Similarly, the line widths are increased in the order of the boundary between stripes, the boundary between main fields of view, and the boundary between subfields of view.

Therefore, at the time of evaluation, the stripe 901 is used.
The boundary 904 between them is inspected with a width of, for example, 200 μm (7 places in FIG. 9). In order to perform a more accurate evaluation, it is preferable to inspect the joint between the main visual fields and further the joint between the sub visual fields. By carrying out the thinning-out inspection in this way, it is possible to carry out an evaluation capable of detecting most of the defects in a fraction of several tenths to several tenths of the time required for the entire inspection.

When the electron beam apparatus 100 is constructed so that the aberration of the optical system is small at the center of the field of view and the distortion of the optical system is small, the measurement at the center of the field of view 905 is more reliable. is there. Therefore, even when the area other than the joint area is evaluated, if the joint area is always evaluated at the center of the visual field 905, the probability of overlooking a defect becomes small. Further, it is less likely that a normal pattern is erroneously determined as a defect.

When performing a defect inspection, as described above, the image obtained by scanning the wafer W with the primary electron beam and the image having no defect stored in the image storage device in advance are compared and collated. By doing so, the defective portion is automatically detected.

Here, another embodiment of the electron beam apparatus according to the present invention will be described with reference to FIG. This electron beam apparatus is used to transfer the pattern formed on the mask onto the wafer. In FIG. 10, an electron beam apparatus 100
0 is LaB ₆ cathode 1001 and Wehnelt 1
An electron gun 1004 having an anode 002 and an anode 1003 is provided, and it operates under a space charge limiting condition with a shot noise reduction coefficient of less than 1.0. The primary electron beam emitted from the electron gun 1004 has an opening plate 100 having a hexagonal shaped opening.
Irradiate 5 with a uniform intensity. As a result, the hexagonal shaped primary electron beam forms a crossover 1006, and an intermediate image 1008 is formed by the condenser lens 1007. The intermediate image 1008 is converged by the objective lens 1009, and further two-stage electromagnetic deflectors 1010, 1010 ′ are provided.
Further, the traveling direction is controlled by the two-stage electrostatic deflectors 1011 and 1011 ′, and the image is projected onto the mask 1012.

In this embodiment, the mask 101
Reference numeral 2 denotes an equal-magnification mask having a pattern for forming one to several chips, which is fixed to the mask stage 1013. The mask 1012 is placed close to the wafer 1014. A wafer 1014 is fixed to a wafer stage 1015 below the mask, and the wafer stage 1015 is movably supported with respect to the mask 1012. When the wafer 1014 is placed on the wafer stage 1015, it is desirable to feed back information regarding the position of the wafer stage 1015 to the deflector to adjust the deflection amount of the primary electron beam.

In order to measure the relative distance between the mask stage 1013 and the wafer stage 1015,
A laser fixed mirror 1016 is fixed to the stage 1013, and the laser moving mirror 1 is fixed to the wafer stage 1015.
017 is fixed, and these mirrors are irradiated with laser light from the laser oscillation / reception unit 1018. At this time, the laser oscillation / reception unit 1018 and the laser fixed mirror 1
016, a half mirror and a fixed mirror are arranged between the laser moving mirror 1017 and the optical path length between the laser moving mirror 1017 and the half mirror of the laser light emitted from the laser oscillation / reception unit 1018, and the first laser. The relative distance between the mask stage 1013 and the wafer stage 1015 can be obtained from the intensity of the received light amount due to the interference caused by the difference in the optical path length between the fixed mirror 1016 and the half mirror. The relative distance thus obtained is the electromagnetic deflector 10
10, 1010 'and electrostatic deflectors 1011, 1011'
To the objective lens 100.
The electron beam converged at 9 is irradiated with the mask 1012 at an appropriate incident angle.

The electromagnetic deflectors 1010 and 1010 'are arranged along the optical axis of the primary electron beam and have substantially equal deflection sensitivities, but they are designed to deflect the electron beams in opposite directions, so that the deflection amount Is adjusted so that the primary electron beam always enters the mask 1012 vertically regardless of On the other hand, in the electrostatic deflectors 1011 and 1011 ′, which are also arranged along the optical axis of the primary electron beam, the ratio of the deflection sensitivities thereof is adjusted so that the deflection center is always on the surface of the mask 1012.

Above the mask 1012 and the mask 1012
A plurality of (for example, eight) backscattered electron detectors 1 for detecting electron beams reflected from the mask 1012 around the
019 are provided, and these backscattered electron detectors 1019 include circuits that perform signal processing independently of each other.

11A and 11B show the mask 10
12 is a diagram illustrating a configuration for detecting a relative position error between wafer 12 and wafer 1014 and performing registration between wafer 1014 and mask 1012. FIG. In the figure, (A) is an enlarged view of a part of the wafer 1014, showing one chip and its peripheral pattern,
FIG. 7B is an enlarged view showing a part of the mask 1012.

The pattern 1102 of the previous layer is already formed inside one chip 1101 formed on the wafer 1014, and the dicing
The line 1103 is formed, and the dicing line 110 is formed.
3, marks for alignment 1104, 1105, 11 for alignment are provided at positions corresponding to the four corners of the pattern 1102.
06, 1107, 1108, 1109, 1110, 11
11 is provided. These marks 1104-1111
May be, for example, regular steps or irregularities having a pitch of 100 nm.

On the other hand, the mask 1012 has the wafer 10
Pattern area 1 in which a pattern to be transferred to 14 is formed
112, and marks 1104 to 1 formed on the dicing line 1103 of the wafer 1014.
Corresponding to the convex portion of 111, the mark holes 1113, 11
14, 1115 and 1116 are provided. Further, detection holes 1117 to 1120 are formed on the opposite side of the mark holes 1113 to 1116 from the pattern area 1112.

Mark 11 formed on wafer 1014
04-1111 and mark holes 1113-1116 and detection holes 1117-1120 formed in the mask 1012.
Detects the marks 1104-1111 and detects the wafer 10
14 is provided for proper alignment between the mask 14 and the mask 1012, that is, registration.

The mask 1012 is scanned by the primary electron beam emitted from the electron gun 1104, and the pattern formed in the pattern region 1112 of the mask 1012 is transferred to the wafer 1.
At the time of transfer to 014, the primary electron beam is desirably scanned while tilting with the mask 1012 as a fulcrum, and the primary electron beam also passes through the mark holes 1113 to 1116 of the mask 1012 and irradiates the marks 1104 to 1111 of the wafer 1014. Therefore, the light is reflected by the marks 1104-1111 and emitted in the hemispherical direction as backscattered electrons. Some of the backscattered electrons are detected through the detection hole 1117 on the mask 1012.
Although it comes out upward from 1120, the detection holes 1117 to 11
The amount of backscattered electrons emitted from the
The marks and the mark holes formed at corresponding positions on the wafer 1014 and the wafer 1014, for example, are substantially proportional to the amount of the marks 1104 to 1107 and the mark holes 1113 to 1116 overlapping with each other or the amount of positional deviation. Therefore, the detection hole 1117
The electron beams that have passed through 1120 to above the mask 1012 are detected by the backscattered electron detectors 1019 (FIG. 10) arranged at the positions corresponding to the detection holes 1117 to 1120, respectively. Mark the amount of line 1
Comparison is made with the amount of backscattered electrons detected when the registration between 104 to 1107 and the mark holes 1113 to 1116 is proper, that is, when there is no relative error. As a result, the marks 1104-1107 and the mark holes 1113 are formed.
.About.1116, that is, the relative position error between the mark 1012 and the wafer 1014 can be obtained. The electrostatic deflector 10 calculates an amount corresponding to the obtained position error.
By feeding back to 11, 1011 ′, the electrostatic deflectors 1011 and 1011 ′ adjust the deflection angle of the electron beam so as to eliminate the position error.

Now, the mask 1012 and the wafer 1014
When the distance between the mask 1012 and the wafer 1 is assumed to be 50 μm, and the deflection angle of the primary electron beam irradiated by the electron gun 1004 is θ milliradian.
The relative error with 014 is measured as a positional deviation of 50 nm. Further, as a result of measurement using a laser length measuring machine, the mask 1-12 is Δx (nm) in the x direction and Δy (nm) in the y direction from the original position with respect to the wafer 1014.
When it is found that there is a deviation, the x component of the deflection angle with the surface of the mask 1012 of the primary electron beam emitted from the electron gun 1004 as the deflection center is θ′x (milliradian),
The backscattered electron detector 1019 with respect to the electrostatic deflectors 1011 and 1011 ′ with the y component being θ′y.

[0083]

[Mathematical formula-see original document] It is sufficient to perform feedback so that the primary electron beam is deflected by θ'x = -Δx / 50 (milliradian) θ'y = -Δy / 50 (milliradian).

When the energy of the primary electron beam is set to 2 keV, the resist sensitivity becomes as high as 1 μc / cm ² . At this time, the number of electrons incident on the 50 nm square pattern is

[0085]

[Equation 6] {(1 × 10 ^-6 c) / (1.6 × 10 ^-19 )}
× (5 × 10 ⁻⁶ ) ² = 156.25 pieces / pattern.

Therefore, the dose fluctuation ΔD due to shot noise is N = 15, where N is the number of electrons per pattern, when an electron gun operating in the temperature limited region is used.
Since it is 6.25,

[0087]

## EQU7 ## ΔD = 1 / (N / 2) ^1/2 = 0.113. That is, the dose fluctuates by 11.3%.

On the other hand, in the embodiment shown in FIG. 10, since the electron gun 1004 is used under the space charge limiting condition, when the shot noise reduction coefficient Γ ² is set to a value less than 1, the dose variation ΔD '. Is

[0089]

Equation 8] △ D '= Γ / (N / 2) 1/2 Γ <1.0 , and the variation of dose is greatly improved over the case of using a TFE electron gun. If Γ <0.5, the dose variation is about 5%. Further, if Γ <0.2, the dose variation is about 2%, which is ideal.

Further, in the embodiment of FIG. 10, since the transfer is performed using the mask 1012 of the same size, the device can be manufactured with high throughput. For example, under the following conditions, the throughput is 45 wafers / hour for a 12-inch wafer. That is, chip size 40 mm × 30 mm 12 inch wafer 300 mmφ wafer number of chips 50 beam current 20 μm dose 1 μc / cm ² beam size 50 μm square stage step time 0.5 sec / chip registration time 0.5 sec / chip transfer time / Wafer 80 seconds = (600 cm ² × 1 × 10 ⁻⁶ ) / (20 × 1 × 10 ⁻⁶ ) + 25 + 25 Throughput 45 wafers / hour, 12 inch wafer.

Here, a semiconductor device manufacturing method for evaluating a wafer in the process or a completed wafer by using the electron beam apparatus 100 shown in FIG. 1, including a general semiconductor device manufacturing method, will be described with reference to FIG. 13 and the flowchart of FIG.

As shown in FIG. 12, the semiconductor device manufacturing method includes a wafer manufacturing step S1 for manufacturing a wafer, a wafer processing step S2 for performing a necessary processing on the wafer, and a mask manufacturing step for manufacturing a mask required for exposure. S3, a chip assembling step S4 for cutting out the chips formed on the wafer one by one and making them operable, and a chip inspecting step S5 for inspecting a completed chip, and these steps are respectively divided. , Including some sub-steps.

Among the above-mentioned steps, the step which has a decisive influence on the manufacture of the semiconductor device is the wafer processing step S2. This is because, in this step, the designed circuit pattern is formed on the wafer, and a large number of chips that operate as memories and MPUs are formed. As described above, it is important to evaluate the processing state of the wafer in the sub-process executed in the wafer processing step S2 that greatly affects the manufacturing of the semiconductor device, and the sub-process will be described below.

First, a dielectric thin film that serves as an insulating layer is formed, and a metal thin film that forms a wiring portion and an electrode portion is formed. The thin film formation is performed by CVD, sputtering, or the like. Then, the formed dielectric thin film and metal thin film, and the wafer substrate are oxidized, and a resist pattern is formed in a lithography process using the mask or reticle created in the mask manufacturing process S3. And by dry etching technology etc.
The substrate is processed according to the resist pattern and ions and impurities are implanted. Then, the resist layer is peeled off and the wafer is inspected.

The wafer processing step S2 as described above is repeatedly performed for the required number of layers, and a wafer before being separated into chips is formed in the chip assembling step S4.

FIG. 13 is a flowchart showing a lithography process which is a sub-process of the wafer processing process S2 shown in FIG. As shown in FIG. 13, the lithography process includes a resist coating process S21 and an exposure process S2.
2. Includes a developing step S23 and an annealing step S24.

In the resist coating step S21, CVD
A resist is applied onto the wafer on which the circuit pattern has been formed by sputtering or the like, and the applied resist is exposed in the exposure step S22. Then, in the developing step S23, the exposed resist is developed to obtain a resist pattern, and in the annealing step S24, the developed resist pattern is annealed to be stabilized. These steps S21 to S24 are repeatedly executed by the required number of layers.

In such a semiconductor device manufacturing method, the electron beam apparatus 100 described with reference to FIG. 1 is used in the chip inspection step S5 for inspecting a completed chip to obtain a semiconductor device having a fine pattern. However, it is possible to obtain an image in which distortion, blurring, etc. are reduced, so that it is possible to reliably detect defects in the wafer.

[0099]

As described above, the electron beam apparatus according to the present invention and the semiconductor device manufacturing method using the apparatus have been described with reference to FIGS.
The present invention includes: 1. Since the E × B separator is not required, the structure of the optical system is simple. Since the fixed mirror of the laser measuring system that measures the position of the wafer stage is fixed to the objective lens, the relative position between the objective lens and the wafer is measured, and it is affected by temperature changes and optical system vibration. No, 3. By manufacturing the laser moving mirror of the laser length measurement system with a material having a large rigidity, the cross section can be reduced even when it is mounted on a stage on which a large diameter wafer such as 12 inches is mounted. In addition, when the laser moving mirror is made of ceramics, even if there is a bore in the ceramics, it will be flat during film formation, so it can be polished to a mirror surface. If the electrostatic deflector is made of one-piece ceramics, it can be processed precisely, so it is possible to provide a highly accurate electrostatic deflector. Also, install the metal plates above and below by an appropriate distance. Therefore, even if the exposed portion of the insulator surface of the electrostatic deflector is charged, its influence does not affect the optical axis. Since the electron beam emitted from the electron gun operating in the space charge limited region is used, an electric signal with a small shot noise and a large S / N ratio can be obtained from the secondary electron beam. 6. Since the beam diameter that maximizes the S / N ratio can be used as the primary electron beam, the beam will not be blurred and the current will not be too small. 7. The variation in dose due to the influence of shot noise can be reduced to about 1/5 at maximum compared to the conventional case. By detecting the mark on the wafer,
It is possible to perform accurate mask and wafer registration. 10. Even if there is mechanical vibration during exposure, the position of the electron beam can be corrected immediately by using the feedback to the deflector, so highly accurate transfer is possible. Even with a pattern having a minimum line width of 0.05 μm, it is possible to transfer at a high throughput, and for example, a wafer having a diameter of 12 inches can be transferred at a throughput of 45 wafers per hour, which is a remarkable effect.

[Brief description of drawings]

FIG. 1 is a diagram schematically showing an embodiment of an electron beam apparatus according to the present invention.

FIG. 2 is a diagram showing an example of a configuration of an objective lens provided in the electron beam apparatus shown in FIG.

FIG. 3 is a view showing a preferable manufacturing process of the laser fixed mirror shown in FIG.

4 is a diagram schematically showing an example of a configuration of an electron beam control element such as an electrostatic deflector provided in the electron beam apparatus shown in FIG.

5 is a sectional view of the electron beam control element shown in FIG.

6A and 6B are views for explaining a region for driving an electron gun of the electron beam apparatus shown in FIG.

7A and 7B are diagrams showing a layout of standard marks placed on a wafer stage of the electron beam apparatus shown in FIG.

FIG. 8 is (A1) to (A3) and (B1) to (B3).
FIG. 8 is a waveform diagram showing an electric signal contrast corresponding to the intensity of a secondary electron beam, which is obtained when the standard mark of FIG. 7 is scanned with various beam diameters using the electron beam apparatus shown in FIG. 1.

9 is a diagram for explaining evaluation of a wafer W using the electron beam apparatus of FIG.

FIG. 10 is a diagram schematically showing another embodiment of the electron beam apparatus according to the present invention.

11A is an enlarged view of a part of the wafer in the electron beam apparatus of FIG. 10, and FIG. 11B is an enlarged view of a part of the mask of FIG.

FIG. 12 is a flowchart of a method of manufacturing a semiconductor device by applying the electron beam apparatus according to the present invention.

13 is a flowchart showing a lithography process which is a sub-process of the wafer processing process in FIG.

[Explanation of symbols]

100: Electron beam device, 104: Electron gun, 105: Alignment device, 106: Condenser lens, 10
7: Astigmatism corrector, 108: Deflector, 109: Objective lens, 110: Electrostatic deflector, 111: Field of view, 11
2: Wafer stage, 113: Laser measurement system,
116: Laser oscillation / reception part, 119: Secondary electron detector, 120: Image data processing part, W: Wafer, 201: Metal rod, 202: Cylindrical part, 20
3: Upper electrode part, 204: Central electrode part, 205: Lower electrode part, 206: Axisymmetric electrode part, 210: Ceramic member, 400: Electron beam control element, 401: Base part, 406: Through hole, 407: Groove , 408: through hole, 409: non-coated surface, 410: electrode,
411: conductor part, 501, 502: shield disk, 70
1, 702: standard mark, 901: stripe, 90
2: Main visual field, 903: Sub visual field, 904: Boundary, 9
05: field of view, 1000: electron beam device, 1004: electron gun, 1005: aperture plate, 1006: crossover, 1007: condenser lens, 1008: intermediate image, 1009: objective lens, 1010, 101
0 ': Electromagnetic deflector, 1011, 1011': Electrostatic deflector, 1012: Mask, 1013: Mask stage, 1014: Wafer, 1015: Wafer stage, 1016: Laser fixed mirror, 1017: Laser moving mirror, 1018 : Laser oscillator / receiver,
1019: Backscattered electron detector, 1101: Chip, 11
02: pattern of the previous layer, 1103: dicing line, 1104 to 1111: mark, 1112: pattern area, 1113 to 1116: mark hole, 111
7 to 1120: detection hole,

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ Identification code FI theme code (reference) H01J 37/22 502 H01J 37/28 B 37/28 H01L 21/30 541D 541U 541B F term (reference) 5C001 AA01 CC04 5C030 BC09 5C033 FF03 UU03 UU04 5F056 BA08 BB01 CC05 EA02 EA06 EA14

Claims (10)

[Claims] 1. An electron gun having a thermionic emission cathode,
An electron beam apparatus comprising: an objective lens that converges an electron beam emitted from the electron gun onto a sample such as a wafer; and a deflector that scans the surface of the sample with the electron beam. An electron deflector, which is composed of at least two-stage deflector arranged between the electron gun and the objective lens, and which is disposed on the side closer to the objective lens. An electron beam device which scans an area off the optical axis. 2. The electron beam apparatus according to claim 1, wherein a stage for mounting the sample, a movable mirror provided on the stage, a fixed mirror provided on the objective lens, and a laser beam An electron beam apparatus, further comprising: a laser oscillation / reception unit for transmitting and receiving, and capable of calculating the relative position of the stage. 3. The electron beam apparatus according to claim 2, wherein the fixed mirror is fixed to the objective lens via ceramics having an expansion coefficient of substantially zero. 4. The electron beam apparatus according to claim 2, wherein the fixed mirror is formed by forming a SiC film on a surface of SiC ceramics and then mirror-polishing the same, or a ceramic stage. An electron beam apparatus comprising a reflection film formed by mirror-polishing after forming a film of ceramics of the same material on the end face. 5. The electron beam apparatus according to claim 1, wherein the deflector comprises an electrostatic deflector, and the electrostatic deflector comprises: a ceramic member having an inner surface of a cylinder; and an axial direction of the ceramic member. An electron beam comprising: a plurality of formed grooves; and a plurality of deflection electrodes, which are formed on the end surface and the inner surface of the ceramic member and the surfaces of the plurality of grooves, and which are electrically insulated from each other. apparatus. 6. The electron beam apparatus according to claim 5, further comprising: a shield member that is disposed close to opposing end surfaces of the electrostatic deflector and has at least a surface made of metal, and the end surface and the end surface. L / D> 4, where D is the distance from the surface of the shield member facing L and L is the shortest distance from the inner surface of the ceramic member to the groove. 7. The electron beam apparatus according to claim 1, wherein the electron gun operates under a space charge limiting condition in which a shot noise reduction coefficient is smaller than 1.0. 8. The electron beam apparatus according to claim 1, wherein the electron beam has a beam diameter in a direction orthogonal to a line-and-space pattern of the sample. An electron beam apparatus, characterized in that a beam size is set so that the S / N ratio of a signal obtained from a secondary electron beam becomes maximum when a line is scanned, and the sample is evaluated. 9. The electron beam apparatus according to claim 8, wherein the sample has a pattern formed by connecting visual fields having a size smaller than a chip size, and a boundary portion between adjacent visual fields is formed. An electron beam apparatus, characterized in that the evaluation is performed in conformity with the center of the field of view of the electron beam apparatus. 10. A device manufacturing method, characterized in that the electron beam apparatus according to any one of claims 1 to 9 is used to evaluate a wafer during a process.