JPH02247964A - Method and apparatus for cross-section shape measurement and comparative inspection - Google Patents

Abstract

PURPOSE:To compute cross-section shape independently of declining a sample surface by obtaining cross-section shapes by illuminance difference stereo method and stereo method according to the difference and sum of signals of a plurality of detectors responding to a plurality of tilt angles and composing waveforms by both methods and putting out the waveforms. CONSTITUTION:An x-y coordinate, tilt angle theta1, theta2, and magnification M are memorized in memory parts 20, 21, 22 respectively by a CRT 25 and key board 26. A stage 9 is transferred to the x-y coordinate by an driving part 12 and set at tilt angle theta1 and secondary or reflected electrons discharged by radiation of electron beam from an electron gun 3 to a sample 10 are received by detectors 7, 8 in right and left sides and sent to difference and sum signal detection parts 14, 15. The sum signal and different signal are memorized respectively in a stereo computing part 16 and illuminance difference stereo computing part 7 as one image. The same operation is carried out at the tilt angle theta2 and the cross-section shape computed result is displayed on a display screen 24 through a waveform composing part 23 after effect of a shadow of a pattern is eliminated using the difference signals at the tilt angle theta1, theta2.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to a cross-sectional shape measuring method and a cross-sectional shape comparative inspection method of fine patterns formed on semiconductor wafers, etc., and apparatus thereof, and particularly relates to an electron beam apparatus. The present invention relates to a cross-sectional shape measuring method, a cross-sectional shape comparative inspection method, and an apparatus thereof, which are suitable for non-destructively measuring the cross-sectional shape of a sample regardless of the shape of the object.

[Conventional technology]

Conventional cross-sectional shape measuring methods and devices for this type use a scanning electron microscope equipped with a plurality of detectors, as described in Japanese Unexamined Patent Application Publication No. 170840/1984, to measure the same part of an object at multiple locations. In a cross-sectional shape measurement method and apparatus for determining the cross-sectional shape using images obtained by observing specimens on sample stands set at different inclination angles, the above-mentioned method is used for sections with gentle inclinations on the sample. The cross-sectional shape is determined from the slope component of the surface element on the sample obtained based on the signals taken in from multiple detectors, and the above-mentioned sample stage is The cross-sectional shape is determined based on the principle of parallax from an image taken in an inclined state, and these are combined and output. According to this method, the cross-sectional shape can be calculated regardless of the inclination of the sample surface.

[Problem to be solved by the invention]

The above-mentioned conventional technology uses the sixth
As shown in the figure, no consideration has been given to the effects of shadows on areas that are shadowed by patterns of secondary electrons or reflected electrons, and the calculated cross-sectional shape will deviate greatly from the actual shape in these shadowed areas. There was a problem.

SUMMARY OF THE INVENTION An object of the present invention is to provide a cross-sectional shape measurement method and apparatus for accurately measuring changes in the shape of a sample surface while taking into account the influence of shadows caused by patterns on the sample.

Another object of the present invention is to provide a cross-sectional shape measurement method and apparatus for accurately calculating the cross-sectional shape regardless of the magnitude of the inclination of the sample surface.

Still another object of the present invention is to simplify the sample stage and increase positioning accuracy by replacing the tilt mechanism with a tilt of the charged particle beam, and to downsize the sample chamber to increase the degree of vacuum and prevent contamination. The object of the present invention is to provide a cross-sectional shape measuring method and an apparatus therefor.

Still another object of the present invention is to perform cross-sectional inspection by comparing cross-sectional shapes and find areas where the three-dimensional shapes do not match. An object of the present invention is to provide a shape comparison inspection method and an apparatus therefor.

[Means to solve the problem]

In order to achieve the above object, the cross-sectional shape measurement method and apparatus of the present invention are provided by using a sample tilted at at least two different tilt angles in order to capture changes in the shape of the sample surface taking into account the influence of pattern shadows. A charged particle beam or electromagnetic wave is scanned on the top of the sample, secondary electrons or reflected electrons generated from the sample are detected by at least two detectors, and the surface area is calculated using the difference signal of the outputs of the multiple detectors for multiple tilt angles. Find the slope of
This is based on the photometric stereo method, which is an improved method that obtains the cross-sectional shape by integrating this.

In addition, in order to achieve the above-mentioned other objects, the cross-sectional shape measuring method and apparatus of the present invention include a plurality of detectors for a plurality of tilt angles in order to accurately measure the cross-sectional shape regardless of the magnitude of the inclination of the sample surface. The stereo method uses the sum signal of the outputs of the two detectors to accurately measure sharp changes in shape, and the difference signal of the outputs of multiple detectors for multiple tilt angles is used to determine the slope of the pixel, and this is integrated. The cross-sectional shape is obtained by combining the waveforms of the calculation results of two methods: the photometric stereo method, which can more accurately measure gradual changes in shape.

In addition, in order to achieve the above-mentioned and other objects, the cross-sectional shape measuring method and apparatus of the present invention replace the tilt mechanism with a tilting method of a charged particle beam, and the objective A deflection coil is provided under the lens to tilt the charged particle beam.

In addition, in order to achieve the above-mentioned and other objects, the cross-sectional shape comparison inspection method and its apparatus of the present invention provide a pattern to be compared and a pattern to be compared, in order to perform inspection by comparing cross-sectional shapes and perform high-precision defect inspection. Calculate the cross-sectional shapes of two objects of the reference target pattern using a cross-sectional shape measurement method and its equipment, align the two calculation results, and determine whether or not there is a defect based on the size and volume of the difference. It was designed to do this.

(Operation) The above-mentioned cross-sectional shape measuring method and device are based on an improved photometric stereo method that takes into account the influence of the shadow part of the pattern as shown in Figure 6, and the tilt angle θ1.θ8
The intensities of the secondary electrons or backscattered electrons of the two left and right detectors when SL I +S□, SL! , 5I11, and the angle ψ that forms the shadow of the pattern of emitted secondary electrons or reflected electrons as shown in FIG.

Define ψ■, and let Φ be the inclination of the surface element with respect to the vertical axis (beam axis) that you want to find, then for the tilt angle θ, st*=cos−'Φ・(cosψ@−5inΦ)(2
)SIII-CO8-1Φ -(cosψa+sinΦ
)(3) holds, and in the conventional photometric stereo method, for example, the difference between SLl and S□, SLI s+tl""Co5-'Φ'(CO3ψA-
C059m 2sinΦ), without considering the influence of the shadow of the pattern, by setting ψ, = ψ1-〇, and setting SLI smt=cos-'Φ (2sinΦ), the slope Φ of the surface element is (SL-S□) However, in contrast, in the photometric stereo method of this improved method, it is calculated as a function of the other tilt angle θ. So, SLl-CO8-'(Φ+ΔΦ)(cosψa-sin
(Φ÷ΔΦ)) (4) Smt=cos-'(Φ+ΔΦ)
−(cosψ, +5in(Φ+ΔΦ)) (5) holds, where ΔΦ=08−〇, and this (2) to (5
) From the formula, middle 1. By eliminating ψ1 and finding Φ, the slope Φ of the surface element can be expressed as the difference between the outputs of the two detectors at two types of tilt angles θ, , θ8 (S□−3□), (SLI
S. ) and the difference between the tilt angle θ1°θ8 ΔΦ(-83-
It can be determined as a function of 〇, ) without being affected by the shadow of the pattern, and the cross-sectional shape can be obtained by integrating this slope Φ and multiplying it by a magnification M.

In addition, the cross-sectional shape measurement method and device described above combine and output the calculation results of two methods: the stereo method and the improved photometric stereo method. The feature points in the two images are extracted and the cross-sectional shape is measured by correlating them.
Although gradual changes in shape with unclear feature points cannot be measured, steep changes in shape with clearly defined feature points can be measured.On the other hand, with photometric stereo method, the incident angle is measured at multiple tilt angles θ1°θ2. Secondary electrons or backscattered electrons are detected by multiple detectors placed in positions facing the charged particle beam or electromagnetic waves, and the slope of the surface elements at each point is determined from the difference in output, taking into account the influence of pattern shadows. Since the cross-sectional shape is measured by determining Φ and integrating it, it is not possible to measure steep changes in the shape where the correlation between the difference signals of multiple detectors and the slope Φ of the surface elements breaks down, but it is possible to measure gradual changes in the shape where the correlation holds true. Since changes in shape can be measured, the cross-sectional shape can be calculated without depending on the inclination of the sample surface by combining and outputting the calculation results of the above two methods.

In addition, the above-mentioned cross-sectional shape measurement method and apparatus can irradiate the sample with a tilted charged particle beam instead of the tilt mechanism of the sample stage by providing a deflection coil under the objective lens for the incident charged particle beam. The stage and sample chamber can be downsized to improve positioning accuracy and vacuum, and to prevent contamination.

In addition, the above cross-sectional shape comparison inspection method and its device calculate the cross-sectional shapes of the measurement target pattern and the reference target pattern to be compared using the present cross-sectional shape measurement method and its device, and compare the two calculation results. Accurate defect inspection can be performed.

〔Example〕

Embodiments of the present invention will be described below with reference to FIGS. 1 to 13.

FIG. 1 is a block diagram showing a first embodiment of a cross-sectional shape measuring device using an electron beam according to the present invention. 1st
In the figure, this cross-sectional shape measuring device includes a scanning electron microscope 1 and a cross-sectional shape calculation section 2 connected to or built in.
Consists of. The scanning electron microscope 1 includes an electron gun 3 that emits electrons, a condenser lens 4 that reduces the electron beam emitted from the electron gun 3, a deflection coil 5 that scans the electron beam, and a deflection coil 5 that further reduces the electron beam and directs it to the sample. an objective lens 6 for illuminating the sample 10; two detectors 7°8 for detecting secondary electrons or reflected electrons emitted from the sample 10;
It consists of a sample stage 9 which can be moved, rotated and tilted in three directions (x, y and z) with fixed O, a sample 1O, a deflection control section 11 for the deflection coil 5, and a stage drive section 12 for the sample stage 9. Of these, two detectors 7 and 8 are located at left and right positions facing the incident electron beam of the sample 10,
And it is arranged in a plane perpendicular to the tilt axis of the sample stage 9.

The cross-sectional shape calculation section 2 also includes a control section 13 that controls the entire cross-sectional shape calculation, and outputs of the two left and right detectors 7.8, c.
A sum signal detecting section 14 which obtains a sum signal d of the two detectors 7 and 8, a difference signal detecting section 15 which obtains a difference signal e of the outputs of the two detectors 7 and 8, and a difference signal detecting section 15 which obtains a difference signal e of the outputs of the two detectors 7 and 8, and a stereo method from the output d of the sum signal detecting section 14. A stereo method calculation unit 16 that calculates the cross-sectional shape, and a photometric stereo method calculation unit 17 that calculates the cross-sectional shape from the output e of the difference signal detection unit 15 using an improved photometric stereo method that takes into account the influence of shadows due to the ° pattern. , a stereo method storage section 18 that stores the calculation result f of the stereo method calculation section 16 for one screen, and a calculation result g of the photometric stereo method calculation section 17.
a photometric stereo method storage unit 19 that stores one screen worth of θ, and a storage unit 20 that stores the tilt angle θ1 of the sample stage 9.
, a θ8 storage unit 21 that stores the tilt angle θ2 of the sample stage 9, a magnification storage unit 22 that stores the magnification M of the scanning electron microscope 1, and a calculation result f and illuminance difference stored in the stereo method storage unit 18. It consists of a waveform synthesis section 23 that synthesizes the calculation results g stored in the stereo method storage section 19, a display 24 that displays the synthesis results of the waveform synthesis section 23, a CRT 25 used for parameter input, and a keyboard 26.

FIG. 2 is a flowchart showing the flow of processing in the embodiment of FIG. The operation of the apparatus having the above configuration will be explained with reference to FIG. First, in step 30, parameters are entered, and here the x and y coordinates, tilt angle θ1°θ2, and magnification M are entered using the CRT 25 and keyboard 26.
Tilt angle θ1. The θ values are stored in the θ1θ2 storage units 20 and 21, and the magnification M is stored in the magnification storage unit 22. In step 31, the sample stage 9 is moved, and the stage drive unit Set x at 12,
Step 32: move to the y coordinate and further tilt to a tilt angle θ1, where θ1 is, for example, θ° (horizontal).
Then, an image is input, an electron beam is emitted from the electron gun 3, the emitted beam is reduced by a condenser lens, and the deflection control unit 11 performs rask scanning in the X-Y direction with the deflection coil 5. further reduce sample 10 with
irradiate. At this time, the secondary electrons or reflected electrons emitted from the sample 10 are shown on the left.

The right detectors 7 and 8 detect the signals and convert them into electrical signals, and the electrical signals (s) and (c) detected by the detectors 7 and 8 are sent to the sum signal detection section 14 and the difference signal detection section 15. The sum signal detection section 14 detects the sum signal d of the two electric signals S and c, and the difference signal detection section 15 detects the two electric signals S and C.

Detecting the difference signal e of C, the stereo method calculation unit 16 stores the sum signal d for one screen, and the illuminance stereo method calculation unit 17 stores the sum signal d for one screen.
Then, the difference signal e for one screen is stored.

In the next step 33, the sample stage 9 is moved in the same manner as in step 31, and this time it is tilted to a tilt angle θ2, where θ2 is, for example, 5''' to lO''. In step 34, the same image input as in step 32 is performed. Here, the stereo calculation unit 16 stores the sum signal d for one screen in a different location from the previous one, and the photometric stereo method storage unit 17 stores the difference signal e for one screen. Memorize it in a different place than before. In step 35, the stereo method calculating section 16 calculates the cross-sectional shape using the stereo method.The sum signal d is used here to balance the signal by removing the influence of shadows caused by the pattern, and the cross-sectional shape calculation result f is calculated using the stereo method. In step 36, the cross-sectional shape is calculated by the photometric stereo method in consideration of the influence of shadows caused by the pattern in the photometric stereo method calculating section 17, and the two tilt angles θ, , θ8 are stored in the method storage unit 18. The influence of the shadow of the pattern can be removed by using the difference signal e of The waveform synthesis section 23 performs waveform synthesis of the cross-sectional shape calculation result g using the photometric stereo method, and outputs the synthesis result.Finally, in step 38, the synthesis result is displayed on the display 24.

FIG. 3 is an explanatory diagram showing an example of waveform synthesis in the embodiment of FIG. 1 (FIG. 2). In FIG. 3, a in FIG. 3 is the cross-sectional shape of the sample 10 to be measured, b in FIG. 3 is the output signal waveform of the left detector 7, and C in FIG. 3 is the waveform of the output signal of the right detector 8. Output signal waveform, d in Figure 3 is the sum signal waveform of b and C, 3rd
In the figure, e is the difference signal waveform between b and C, f in FIG. 3 is the output signal waveform of the stereo method calculating section 16, g is the output signal waveform of the photometric stereo method calculating section 17, and h is a waveform composite signal waveform using f and g, and the symbol of these waveforms is

c, d, e, f, g are the signals in FIG.

It corresponds to the symbols e, f, and g. From FIG. 3, with respect to the cross-sectional shape of a in FIG. 3, the output signal waveform of the stereo method calculation unit 16 in r in FIG. Although the waveform of the output signal from the photometric stereo unit 16 shown in g in FIG. 3 captures a gradual change, it can be seen that there is a deviation from the actual cross-sectional shape in the portion of the steep change. Therefore, the output signal waveform of the stereo method calculation unit 16 shown in f in FIG. 3 is used for the steep change portion, and the output signal waveform of the photometric stereo method calculation unit 17 shown in g in FIG. 3 is used for the gradual change portion. If the waveform synthesis is performed by the waveform synthesis section 23 in the following manner, the cross-sectional shape can be calculated regardless of the shape of the measurement object, like the waveform synthesized signal waveform h in FIG. 3.

Here, in the waveform synthesis section 23 of FIG.
As described in Japanese Patent No. 0840, the calculation result r of the stereo method calculation section 16 and the calculation result g of the illuminance stereo method calculation section 17 are combined in the following three cases. First, the left edge of the screen is the left edge of the screen, the right edge is the feature point, and the altitude is determined by the stereo method.In this case, the right edge is matched by adding a constant to the calculation result g of the photometric stereo method. do it like this. The second case is when the right edge of the screen is on the right and the feature point is on the left side, and in this case, a constant is added so that the left edges coincide, as in the first case. The third case is that both ends are feature points and the altitude is determined by the stereo method.

In this case, the left end altitude is ZL and the right end altitude is 21'
When the altitudes of the left end and right end calculated by the stereo method are ZL and Zl, respectively, and there are N pixels in this section, the calculation result g of the photometric stereo method is converted as follows.

Zt=Z! +(Z'L Zt)+
((Z'*-Zt)-(Z'L-ZL)) Here, when i=0, it is the left end, and when i=N, it is the right end.

FIG. 4 is a block diagram showing the configuration of the stereo method calculating section 16 of FIG. 1. In FIG. 4, the stereo method calculation unit 1
6 is a sum signal detection unit 14 based on a switching signal from the control unit 13.
A switching unit 41 that switches and stores the sum signal d from the 01 image storage unit 42 or 02 image storage unit 43, an 01 image storage unit 42 that stores the sum signal d at the tilt angle θ, and a 01 image storage unit 42 that stores the sum signal d when the tilt angle θ2 02 image storage unit 43 that stores the sum signal d when Calculation unit 45, pattern feature points, θ, and tilt angle θ from θ8 storage unit 20.21
1. The calculation unit 46 calculates the height from θ2 and the magnification M from the magnification storage unit 22.

FIG. 5 is an explanatory diagram of a method for calculating feature points of the pattern shown in FIG. 4. The operation of the stereo method calculating section 16 will be explained with reference to FIG. First, θ3 of the sum signal d by the switching unit 41.
Assuming that θ, West image θ, θ2 have been stored in the image storage units 42 and 43, the feature point calculation unit 44 calculates the coordinates of the feature points of the pattern from the image at the tilt angle θ1. In FIG. 5, i in FIG. 5 indicates the cross-sectional shape a of the stepped portion of the pattern, and the feature points to be found here are the upper and lower ends of the stepped portion, and the points A and B in i in FIG. corresponds to

Therefore, point A in the waveform of the j-m sum signal d in FIG.

For the coordinate XA with respect to the upper end A of the 0-level difference portion for which the coordinates Xa and Xm corresponding to B are determined, for example, the coordinate X of the arrow point that takes the maximum value in the waveform of the sum signal d, as shown in 1 to m in FIG. Seek more. Regarding the coordinate X with respect to the lower end B of the stepped portion, there are, for example, four methods as shown in 1 to m in FIG. j in Fig. 5 is a method of determining from the coordinates Xl of the arrow point that takes the minimum value in the waveform of the sum signal d, and K in Fig. 5
is the maximum value and minimum value in the waveform of sum signal d, 100% between 0
The coordinates of the three arrow points that are n% from the minimum value when
The optimal value of n is determined in advance for the target sample 10. For example, 50% is a possible method for n, as shown in Figure 5! extracts the horizontal part and the slope part from the waveform of the sum signal d, and directly approximates each part to form a straight line shown by a dashed line, ,L! , and the coordinate X of the arrow point that is the intersection point.
is a method of determining from the coordinate X of the arrow point of the maximum slope in the waveform of the sum signal d, that is, the point of the maximum value of the differential waveform of the sum signal d of m' in FIG. Which of these methods is appropriate varies depending on conditions such as the material of the sample IO and the accelerating voltage of the scanning electron microscope 1, so the measurer selects the method that suits the measurement conditions. Note that the measurer may input the feature points using the CRT 25 and keyboard 26 while viewing the image on the display 24.

After calculating the feature points of the above pattern, the corresponding point calculation unit 45
In this step, the correspondence between a minute area including a feature point in the image at the tilt angle θ and a minute area in the image at the tilt angle θ2 is determined. In order to find this correspondence, correlation values are used as described in Japanese Unexamined Patent Publication No. 170840/1984, and the region where the correlation value is the largest is adopted. Therefore, the corresponding point calculation unit 45 consists of a means for calculating a correlation value and a means for calculating the maximum value of the correlation value, and by calculating this correspondence, the feature points in the 01 image and the 02 image are determined. The coordinates xAI+Xll of the feature point in the 01 image thus determined, and the 02
The coordinates Xag+Xll of the feature point in the image, the magnification M of the operating electron microscope 1, and the tilt angle θ8. Based on θ2, the altitude difference between the feature points is calculated using the following equation (1).

However, θ=02-01.

FIG. 6 is an explanatory diagram of the photometric stereo method of the improved method that takes into account the influence of pattern shadows, which is performed by the photometric stereo calculation unit 17 of FIG. In Fig. 6, in the conventional photometric stereo method, as shown in the figure, there are areas that are shaded by the pattern, and measurement errors from the actual cross-sectional shape occur near the patterns. Improvements are being made.

Here, for example, when the tilt angle is θ, the intensity of the secondary electrons or reflected electrons of the two left and right detectors 7.8 is S L
I h S I I, and the angle v that shadows the pattern of emitted secondary electrons or reflected electrons shown as a circle as shown in the figure,
'r' is defined, and the inclination angle of the surface element formed between the incident electron beam (vertical axis) to be determined and the perpendicular line N of the surface element of the sample IO is defined as φ. At this time, Scanning Microscopy Volume 1,
No. 3 (1987), pages 963 to 973 (Sc
Anning Microscopy, Vol. 1
, Na3 (1987) pp963-973), the following equation holds.

SLI"Co5-'Φ・(cos'i'4-sinΦ)
(2) sat"cos-'Φ-(coS's, +s
inΦ) (3) Here, in the conventional photometric stereo method, for example, the difference between the intensities SLl and S□ when the tilt angle θ is -, 5
LI-3ll"'cos-'Φ @ (cos'P
4-cos'Ps+-2sinΦ), without considering the influence of the shadow of the pattern, that is, as the angle 'Pa-'Pa'' O@ which is the shadow of the pattern, SL, -3,,=co
By setting s-'Φ・(-2sinΦ), the strength S
Considering the difference between Ll+ 311 as a function of only the inclination angle Φ of the plane element, find the relationship between 5LI-8□ and Φ in advance,
Φ was determined from the measured 5LI-3□. Therefore, errors from the actual cross-sectional shape occur in areas affected by the shadow of the pattern.

In this example, in order to take into account the influence of the shadow of the pattern, the secondary electrons or reflected electrons of the two left and right detectors 7 and 8 are changed from the tilt angle θ to θ8. SLI each strength! + sat, SL, -cos-'(Φ(ΔΦ)・(cosψ4-s
in(Φ+ΔΦ))(4)Sm*−(6g−1(
Φ+ΔΦ)−(cosψ,+5in(Φ+ΔΦ))(5
) holds, provided that ΔΦ=θ, 20.

By eliminating ψ, , ψ1 from equations C) to (5) and finding Φ, the following is obtained. Therefore, by equation (6), two types of tilt angles θ1. The outputs of the two detectors 7.8 when θ8.

C difference (SLI Sa+), (Stg Sl
and tilt angle θ1. It can be seen that the inclination angle Φ of the surface element can be determined without being affected by the shadow of the pattern from the difference ΔΦ (=08−〇,) between θ and θ. By integrating this inclination angle Φ and multiplying by a magnification M, the cross-sectional shape is obtained.

Figure 7 shows the photometric stereo method calculation unit 1 in Figure 1 (Figure 6).
7 is a block diagram of the configuration of FIG. In FIG. 7, the present photometric stereo method calculation unit 17 executes cross-sectional shape calculation using the improved photometric stereo method that takes into account the influence of the shadow of the pattern shown in FIG. a switching unit 61 that switches and stores the difference signal e from the difference signal detection unit 15 in the θ6.03 image storage unit 62.63 according to a signal;
The θ1 image storage unit 62 stores the difference signal e when the tilt angle θ, and the θ1 image storage unit 62 stores the difference signal e when the tilt angle θ:.
8 image storage section 63, a calculation section 64 that calculates the inclination angle Φ of the surface element by calculating equation (6), and an integration section 65 that calculates the cross-sectional shape of the sample 10 by integrating the inclination angle Φ of the surface element. Become. With this configuration, as explained in FIG.
1 of the 87.08 image of the difference signal e due to θ1. Assuming that the storage up to the θ image storage units 62 and 63 has been completed, the calculation unit 64 calculates the tilt angle θ using equation (6).

The intensity difference (SLI S□), (SL!S□) of the difference signal e from the difference signal detection unit 15 when θ2 and the tilt angle θ1
．． The difference ΔΦ (=08−〇,) between θ and the inclination angle Φ of the surface element
is obtained, and the inclination angle Φ of the pixel is integrated in the integrating section 65,
By applying the magnification M, a cross-sectional shape of the sample 10 from which the influence of the shadow of the pattern has been removed can be obtained.

FIGS. 8(a) to 8(f) are a partial block diagram showing a second embodiment of a cross-sectional shape measuring apparatus using an electron beam according to the present invention, and an explanatory diagram of an electron beam tilting method. In FIGS. 8(a) to (f), the same reference numerals as in FIG. 1 indicate corresponding parts, and the configuration of this cross-sectional shape measuring device is almost the same as that in FIG. 1, as shown in FIG. 8(a). However, the difference from Fig. 1 is that there is an additional lens under the objective lens 6 of the scanning electron microscope l.
The present invention is to provide two deflection coils 60 and to control them by the deflection control section 11.

With this configuration, the electron beam from the electron gun 3 can be irradiated onto the sample lO at an angle without having a tilt mechanism on the sample stage 9 or without tilting the sample stage 9 as shown in FIG. The stereo method calculation section 16 and the illuminance stereo method calculation section 17 of the cross-sectional shape calculation section 2 are
It is possible to calculate the stereo method and photometric stereo method. For example, as shown in FIG. 8(b), a deflection coil 60
If the tilt angle of the electron beam is ±3° with respect to the vertical axis, the error in the tilt angle is ±0. O1% (3°±0.03@”
), in order to obtain a field of view of 5μ-, the following equation holds true, where h is the working distance.

Solving this, we obtain h″#4.8s+a. Therefore, it is fully possible to arrange the deflection coil 60 in this space. Also, in the configuration of FIG. 8(a), it is possible to arrange the deflection coil 60 in this space.
Various electron beam deflection methods shown in (f) to (f) can be considered, and FIG. Figure (b) shows a method of deflecting an electron beam by a deflection coil 60 passing through the center of the objective lens 6 and below the objective lens 6.
Figure (e) shows the objective lens 6 without passing through the center of the objective lens 6.
A method in which the beam is incident parallel to the vertical axis and deflected by the deflection coil 5 above the objective lens 6. Fig. 8(f) shows a method in which the beam is incident parallel to the vertical axis into the objective lens 6 without passing through the center of the objective lens 6. For example, the beam may be deflected using a deflection coil 60 under the lens 6.

FIG. 9 is a block diagram showing an embodiment of a cross-sectional shape comparison inspection device using the cross-sectional shape measuring device according to the present invention. In Fig. 9, this cross-sectional shape comparison device is the cross-sectional shape measuring device shown in Fig. 1 (1 in Fig. 1 can be replaced with Fig. 8).
70, a storage unit 71 that stores the output waveform of the measurement target pattern for the sample IO, and a cross-sectional shape measuring device 70.
A positioning section 73 inputs the output waveform of the output waveform and the output waveform of the storage section 71 and aligns the two output waveforms, a difference detection section 73 takes the difference between the two waveforms, and a difference detection section 73 detects the difference between the two output waveforms. and a defect determination section 74 that determines a portion with a large volume difference to be a defect. With this configuration, the cross-sectional shape measuring device 7
The output waveform of the sample 10 of the measurement target pattern 0 is stored in the storage unit 71, the output waveform of the reference target pattern to be compared is inputted, and the alignment unit 72 aligns these two output waveforms. According to this embodiment, the difference detection unit 73 calculates the difference between the two waveforms, and the defect determination unit 74 determines a portion with a large volume of the difference as a defect and outputs it.
Compared to conventional optical cross-sectional shape comparison inspection equipment that uses differences in brightness, the defect size can also be determined in the depth direction, making it possible to conduct more rigorous inspections.

FIG. 1θ is a configuration block diagram showing a third embodiment of a cross-sectional shape measuring device using an ion beam according to the present invention.

In FIG. 10, the same reference numerals as in FIG. 1 indicate corresponding parts, and this cross-sectional shape measuring device can be used not only by electron beams but also by irradiating the sample 10 with other charged particle beams. An ion beam emitting device 8 using, for example, an ion source (ion gun) 80 instead of the electron source (electron gun) 3 of
1, and the other configurations are the same as in FIG.

With this configuration, an ion beam is emitted from the ion source 80 of the ion beam radiation bag 281, the emitted ion beam is reduced by the condenser lens 4, and the deflection coil 5
Rask scanning is performed in the X-Y direction, and the sample 10 is irradiated with further reduction by the objective lens 6. At this time, secondary electrons are generated from the sample 10 as shown in the following 11th example) and FIG.

FIGS. 11(a), 11(b), and 11(C) are schematic diagrams of interactions between electrons, ions, electromagnetic wave irradiation, and solid surfaces, respectively.

This schematic diagram is described in "Surface Analysis" by Iwao Amori and Dan Someno, published by Kodansha (1976), page 5. In response to the primary electron beam I'p in f), secondary electrons (Auger electrons) and backscattered electrons, as well as ions, X-rays, fluorescence, phosphorescence, etc. are generated from the interaction region surrounded by the broken line on the solid surface. In the 11th report), for the primary ion beam 1p in Fig. 10, in addition to secondary electrons, backscattered ions, reflected ions,
Secondary ions, neutral atoms, X-rays, light, etc. are generated, and in Figure 11 (C), photoelectrons and secondary electrons (
Auger electrons) are emitting faint X-rays.

FIG. 12 is a diagram showing the angular distribution of the ion beam and the emitted secondary electrons. In Fig. 12, the generation of secondary electrons as shown in Fig. 11 (b) due to ion beam irradiation in Fig. 10 is explained, for example, by microelectronic
Microelectronic Engineering 5 (1986) pp. 481-489
nering5 (1986) pp 481-489
), as shown in Figure 12, and in this example, the K, + ion beam W (polycr
,) is 60°, and the acceleration voltage is 5 key,
The angular distribution of secondary electrons emitted for 10 keV and 19 keV is shown. Therefore, as in the case of the electron beam shown in FIG. 1, the cross-sectional shape can be calculated by the cross-sectional shape calculating section 2 using the secondary electrons generated from the sample IO by irradiation with the ion beam shown in FIG. Further, as in the case of FIG. 8, an ion beam tilting method using a deflection coil 60 provided below the objective lens 6 can be used.

FIG. 13 is a block diagram showing a fourth embodiment of the cross-sectional shape measuring device using X-rays according to the present invention. In Fig. 13, the same reference numerals as in Fig. 1 indicate corresponding parts, and this cross-sectional shape measuring device is an example in which X-rays, which are electromagnetic waves, are used instead of charged particle beams, and in place of the electron beam system in Fig. 1. X-ray source 9
The X-ray device 91 is composed of a slit 92, a mirror 93, a slit 92, a mirror 93, etc., and the other components are the same as those shown in FIG. With this configuration, the sample 10 is irradiated with X-rays emitted from the X-ray source 90 of the XvA device 91 using the slit 92 and the mirror 93, and at this time, as shown in FIG. Processing similar to the case of electron beams is performed using secondary electrons. Here, as the mirror 93, Applied Optics Vol. 17, No. 14 (1978), pp. 601 to 603 (APPLIBD 0PTIC
3Vol, 17.11kL14 (197B) pp601
-603) or the Toroidal mirror described in
Solid State Physics J Vol, 20. Fill (1985), pages 865 to 870, a Wolter type mirror, etc., may be used, a zone plate may be used in place of the mirror, and a Schwarzschild objective lens may be used. However, in the case of X-rays, unlike an electron beam as shown in FIG. 8, it cannot be deflected, so the same operation as deflection is performed by moving the sample stage 9 by the stage drive unit 12 as in FIG. 1.

〔Effect of the invention〕

According to the present invention, since the slope of a surface element is determined by the photometric stereo method using the difference signals of a plurality of detectors for a plurality of tilt angles, an error in the slope of a surface element from the true value may occur due to the influence of the shadow of the pattern. By integrating this, gradual changes in the cross-sectional shape can be calculated regardless of the location.

In addition, for steeply sloped parts and stepped parts, the cross-sectional shape is determined using the stereo method using the sum signal of multiple detectors for multiple tilt angles, and for gently sloped parts, the cross-sectional shape is determined using the photometric stereo method using the difference signals of multiple detectors. Since the shape is determined and the waveforms from these two methods are combined and output, the cross-sectional shape can be calculated regardless of the magnitude of the slope of the sample surface.

Furthermore, by installing a new deflection coil under the objective lens and using this to irradiate the charged particle beam at an angle,
Since the sample stage can be simplified as a substitute for the tilt mechanism, positioning accuracy can be improved, and the sample chamber can be downsized to increase the degree of vacuum and reduce contamination.

Furthermore, by performing defect inspection by comparing cross-sectional shapes, it is possible to know the three-dimensional shape of the mismatched portion, which has the effect of enabling more accurate inspection than conventional defect inspection by comparing brightness.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a first embodiment of a cross-sectional shape measuring device using an electron beam according to the present invention, FIG. 2 is a flowchart of the process shown in FIG. 1, and FIG. An explanatory diagram of waveform synthesis, Fig. 4 is a block diagram of the configuration of the stereo method calculation section of Fig. 1, Fig. 5 is an explanatory diagram of the pattern feature point calculation method of Fig. 4, and Fig. 6 is an illustration of the illumination difference in Fig. 1. An explanatory diagram of the stereo method, FIG. 7 is a block diagram of the configuration of the photometric stereo method calculation unit in FIG. 1, and FIGS. 8 (a) to (f) are the second implementation of the cross-sectional shape measuring device using an electron beam An explanatory diagram of an electron beam tilting method showing an example, FIG. 9 is a block diagram showing an embodiment of a cross-sectional shape comparison inspection device according to the present invention, and FIG. 10 is a diagram of a cross-sectional shape measuring device using an ion beam according to the present invention. Third
11 (a) to (0
) is a schematic diagram of the interaction between electrons, ions, electromagnetic waves, and solid surfaces, Figure 12 is an angular distribution diagram of ions, electrons, and emitted secondary electrons, and Figure 13 is a cross-sectional shape measurement using X-rays according to the present invention. FIG. 7 is a configuration block diagram showing a fourth example of the device. 1...Scanning electron microscope, 2...Cross section water wiping calculation section,
3...electron gun, 4...condenser lens, 5...
Deflection coil, 6... Objective lens, 7.8... Detector, 9... Sample stage, 10... Sample, 11...
Deflection control section, 12... Stage drive section, 13... Control section, 14... Sum signal detection section, 15... Difference signal detection section, 16... Stereo method calculation section, 17... Illuminance stereo method calculation unit, 18... Stereo method storage unit, 19.
...photometric stereo method storage section, 20...θ, storage section,
21... θ2 storage section, 22... Magnification storage section, 23.
...Waveform synthesis section, 24...Display, 25...
CRT, 26...Keyboard, 60...Deflection coil, 70...Cross-sectional shape measuring device, 71...Storage unit, 7
2... Positioning section, 73... Difference detection section, 74...
- Defect determination unit, 80... Ion source, 81... Ion beam radiation device, 90... X-ray source, 91... X-ray device, 92... Slit, 93... Mirror agent Patent Attorney Tadashi Akimoto Actual Figure Figure 6 Figure Atsushido 1 White Figure (a) 60: aril'': J-hiki〆Figure (f) Seven-point 5P: Figure

Claims (1)

[Claims] 1. Scanning a charged particle beam or electromagnetic waves on a sample tilted at at least two types of tilt angles, and detecting secondary electrons or reflected electrons generated from the sample using at least two detectors; A cross-sectional shape measuring method characterized by calculating the inclination of the sample surface from the difference signal of the detector at each tilt angle and measuring the cross-sectional shape by integrating this. 2. The cross-sectional shape measuring method according to claim 1, wherein the two detectors are arranged in a plane perpendicular to a tilt axis for tilting the sample. 3. The cross-sectional shape measuring method according to claim 1, wherein instead of tilting the sample with respect to the charged particle beam, a deflection coil is provided under the objective lens and the charged particle beam is irradiated at an angle. 4. A charged particle beam or electromagnetic wave is scanned over a sample tilted at at least two different tilt angles, and secondary electrons or backscattered electrons generated from the sample are detected by at least two detectors, and a detector at each tilt angle is used. The cross-sectional shape of the sample is measured using the stereo method from the sum signal of A cross-sectional shape measurement method characterized by calculating a cross-sectional shape by combining shapes. 5. A cross-sectional shape comparative inspection method, characterized in that the cross-sectional shape of the pattern to be inspected is measured by the cross-sectional shape measuring method according to claim 1, and the measured cross-sectional shape is compared with a reference pattern. 6. A cross-sectional shape comparative inspection method, characterized in that the cross-sectional shape of the pattern to be inspected is measured by the cross-sectional shape measuring method according to claim 4, and the measured cross-sectional shape is compared with a reference pattern. 7. means for scanning a charged particle beam or electromagnetic wave on a sample tilted at at least two different tilt angles, and detecting secondary electrons or reflected electrons generated from the sample using at least two detectors; 1. A cross-sectional shape measuring device comprising means for calculating the slope of a sample surface from a difference signal of a detector and measuring the cross-sectional shape by integrating the slope. 8. The cross-sectional shape measuring device according to claim 7, wherein the two detectors are arranged in a plane perpendicular to a tilt axis for tilting the sample. 9. Regarding the charged particle beam, the detecting means is characterized in that instead of tilting the sample, a deflection coil is provided under the objective lens, and the charged particle beam is irradiated at an angle.
The cross-sectional shape measuring device described. 10. Scanning a charged particle beam or electromagnetic wave on a sample tilted at at least two different tilt angles,
means for detecting secondary electrons or backscattered electrons with at least two detectors, means for measuring the cross-sectional shape of the sample by a stereo method from the sum signal of the detectors at each tilt angle, and means for measuring the cross-sectional shape of the sample from the difference signal of the detectors at each tilt angle. A cross-sectional shape characterized by comprising means for measuring the cross-sectional shape by calculating and integrating the slope of the sample surface, and means for calculating the cross-sectional shape by combining the cross-sectional shapes of the above two means. measuring device. 11. A cross-sectional shape comparative inspection device, comprising means for measuring the cross-sectional shape of the pattern to be inspected using the cross-sectional shape measuring device according to claim 7 and comparing the measured cross-sectional shape with a reference pattern. 12. A cross-sectional shape comparative inspection device, comprising means for measuring the cross-sectional shape of the pattern to be inspected using the cross-sectional shape measuring device according to claim 10 and comparing the measured cross-sectional shape with a reference pattern.