JPS6316687B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an apparatus capable of extremely accurately detecting the height of an object, such as the height of a mask blank or wafer surface in a charged particle beam exposure apparatus.

For example, when drawing a fine circuit pattern on a semiconductor wafer etc. using an electron beam exposure device, if the wafer surface deviates from the set height, the position and size of the exposed circuit may differ from the predetermined one. Especially when multiple exposure is performed on a semiconductor wafer, the drawing accuracy is significantly reduced. Therefore, accurately measuring the height of the exposed material is extremely important for highly accurate drawing.

In conventional height measuring devices, an electrode is placed opposite the surface of the material to be exposed, and the capacitance of a capacitor formed between this surface and the electrode changes as the surface moves up and down. There are two types of laser beams in use: one that irradiates the surface of a material to be exposed with laser light and uses interference fringes between the reflected light on the surface and the irradiated light.

However, in the former case, an electrostatic field is generated, which adversely affects the electron beam during measurement.
Therefore, the height measurement point must be located significantly away from the most important electron beam irradiation point, and high measurement accuracy cannot be expected. In addition, since the latter is an optical measurement, it is possible to match the electron beam irradiation point and the measurement point, but since it uses the integration of the number of interference pulses, it is difficult to traverse irregularities on the surface of the exposed material. If there is an interruption in the light, such as when the light is interrupted, subsequent measurements will be completely unreliable.

Recently, devices have been proposed that can solve these drawbacks. As shown in FIG. 1, this device projects light from a light source 1 obliquely onto the surface of a material to be exposed 2, irradiates this projected light onto a member 3 having an aperture, and directs the passed light through a lens. 4 to form an image near the surface of the material to be exposed, and a lens 5 is placed in the traveling direction of the light reflected from the material surface to form the image on the photoelectric detection surface of the image dissector tube 6. Instead, a signal corresponding to the position of the image is generated, and the height displacement is calculated from the signal.

In such an apparatus, if the material 2 changes by a height h from 2a to 2b as shown in FIG.
When the distance between the virtual images p' and p'' of the aperture image p is L, the magnification of the lens 5 is M, and the angle of incidence and reflection of light is θ, the amount of deviation Δ of the aperture image on the detection surface is Δ=M・It is given by Lcosθ=M・2hcosθ.Since the above M and θ are known,
If Δ is found, the height displacement h can be easily found.

This device is a non-contact, optical type and can measure the surface height at the point of irradiation of the electron beam without affecting the electron beam in any way, and does not use integration of interference pulses, so it is possible to measure the surface height at the point where the electron beam is irradiated. This has the effect that accurate height measurement is possible even when there is a problem.

However, the image dissector tube is composed of a photoelectric conversion surface, an aperture plate, a secondary electron multiplier, a collector electrode, an electrostatic lens, a deflection coil, a power source for each part, etc., so the structure is complex and large.
Another problem is that it is very expensive. especially,
The large size of the device makes it difficult to install it in a narrow exposure room, making it impossible to utilize the device, which has many advantages. Therefore, if a semiconductor array sensor is used in place of the image dissector tube, the above structural problem will be solved, but a new problem will arise. In other words, a semiconductor array sensor uses a semiconductor photodetector element with a thickness of 10 to 30μ.
Although a large number (for example, 2048) are arranged at a pitch of m, there is a limit to the measurement accuracy (absolute accuracy).
It is difficult to obtain an accuracy of 1 μm or less in terms of height. Now, if the magnification M of the imaging lens system is 10 times and the pitch P of the semiconductor element is 25 μm, even if there is no blur in the image formed on the array sensor, in principle P/M = 2h cos θ 25/10
= measurement accuracy of only 2.5 μm can be obtained. Moreover, since the light passing through the aperture plate is imaged by a lens and projected onto the material surface, and the reflected light is further magnified by the imaging lens and projected onto the detector, the image on the detector is considerably blurred. , and the sensitivity of each semiconductor element is usually
Since there is a variation of about 20%, the above accuracy becomes even worse.

The present invention has been made to solve all of the above problems, and proposes an apparatus that performs highly accurate height measurement using a semiconductor array sensor that has a simple structure and is inexpensive. The configuration of the present invention includes an irradiation optical system that irradiates light onto the surface of an object at a constant angle θ and forms a plurality of bright and dark images at minute intervals near the irradiation point, and condenses the light reflected from the irradiation point. and an imaging optical system for forming the plurality of bright and dark images, a semiconductor array sensor disposed on the imaging plane of the imaging optical system and consisting of a large number of semiconductor photodetecting elements, and each sensor on the sensor. a calculation means for determining the distances of the bright and dark images from a reference position and averaging them; The present invention is characterized by a distance measurement device.

An embodiment of the present invention will be described below based on the drawings.

In FIG. 3, numeral 7 indicates an electron gun, and an electron beam 8 emitted from the electron gun is focused by an electron lens system 9 and projected onto the material 2 to be exposed. 10 is a deflector for deflecting the electron beam 8 and moving it on the exposed material 2 to draw a pattern;
A pattern signal is sent from a computer 12 via 1. Between the light source 1 and the lens 4 of the irradiation optical system, a plurality of lenses, for example, three
A slit plate 3' having a number of holes is placed, and the light passing through the slit plate is imaged by a lens 4.
After forming its bright and dark image at point p, it is projected onto the material 2 at an angle θ. The light reflected by the material 2 is imaged by an imaging lens 5, and is magnified and imaged onto a semiconductor array sensor 13 in which a large number of semiconductor photodetecting elements are arranged. In other words, a virtual image p' of point p is projected onto the sensor. The position of the image of the slit plate 3' by the lens 4 is not limited to the front of the material irradiation point as shown in the figure, but may be at or behind the irradiation point. The signal from the semiconductor array sensor 13 is amplified by the amplifier 14 and sent to the arithmetic circuit 1.
Sent to 5. In this arithmetic circuit, the distances of each bright and dark image of the slit image from the reference position are determined and averaged. The averaged height signal of the material 2 is sent to the display device 16 and displayed as a height position. Further, the signal is sent to the amplifier 11 of the deflector, the objective lens, the beam shift deflector, and the field rotation lens, etc., so that the drawing position of the drawing pattern and the exposure field are adjusted regardless of the height displacement of the exposed material. Control them so that the size, focusing, etc. are constant.

FIG. 4 is a diagram illustrating the operation of the present invention, and a
The figure shows the detection surface of the semiconductor array sensor 13, in which a large number of semiconductor photodetecting elements are arranged at regular intervals. Further, Figure b shows the light intensity distribution of the image (bright and dark image) of the slit plate 3' projected onto the detection surface. In this figure, a slit plate 3' having three slits is used, and A, B, and C are the light intensity distributions of the image. The light image of A produces outputs from elements a ₁ and a ₂ , which correspond to both edges of the light image _of A.
Reference position of a ₂ (Here, the left end was used as the reference device)
The distances A ₁ and A ₂ from are found. Depending on the optical image B _{, outputs are obtained from the elements b 1 , b 2 , and b 3 , and the distances} B ₁ ,
B ₂ is required. Furthermore, by the optical image of C, c ₁ , c ₂ ,
Outputs are generated from elements c ₃ and c ₄ , and distances C ₁ and _{C 2 from} the reference positions of c ₁ and c 4 are determined.

A ₁ , A ₂ , B ₁ , B ₂ , C ₁ , obtained in this way,
C ₂ are added together and averaged in the arithmetic circuit 15. This averaged distance signal (therefore,
A signal corresponding to the height of the material 2) is sent to a display device 16, where the height is displayed, for example numerically. or,
This distance signal is compared with a reference value, and the difference signal is sent to, for example, an amplifier 11, and the pattern signal from the computer 12 is corrected. Material to be exposed 2
When the height of the optical images A, B, and C changes, the positions of the optical images A, B, and C on the semiconductor array sensor 13 change (shift to the right or left in FIG. 4), so that the averaged signal value changes. The display changes, and the correction signal changes accordingly. In addition, Fig. 4 a, b
As can be seen, there are three slit optical images (A, A, and A, respectively).
(corresponding to B and C) array length (length from A to C)
is made shorter than the detection surface of the array sensor 13 by an amount corresponding to the normal height displacement of the material 2.
If it is not shortened like this, the displacement of material 2 will cause 3
One or more of the slit light images A, B, and C may be projected off the array sensor detection surface. If such a thing occurs, the distance signal cannot be obtained from the deviated slit light image, so the calculated value in the calculation circuit 15 is based on the distance signal from the three slit light images A, B, and C. As described above, in the present invention, a plurality of optical images A,
Find the distance of both edges B and C from the reference position,
Since these are averaged and determined as the height of the material surface, it is possible to overcome the limits of measurement accuracy due to variations in the array spacing of semiconductor elements, variations in sensitivity of each element, and blurring of the image, and use a single aperture image. The accuracy can be improved several times (approximately 5 times) compared to the conventional case, and therefore height measurement can be performed with submicron accuracy.

Note that the above is one embodiment of the present invention, and many modifications are possible. For example, the number of slits is not limited to three as shown in the figure, but may be any number as long as two or more images of the slits are projected onto the sensor 13. However, the array length of the optical images of the plurality of slits is made shorter than the detection surface of the array sensor by an amount corresponding to the normal height displacement of the material. Of course, the greater the number of slits, the better the accuracy. Also, the fourth
In the figure, the position of the element in a bright part, ie, a part irradiated with light, is detected, but conversely, the position of the element in a dark part, ie, a part not irradiated with light, may be detected. Furthermore, although FIG. 3 shows the case where the present invention is applied to an electron beam exposure apparatus, there is no particular restriction on the applicable apparatus. Furthermore, the width of each slit, that is, the width of the optical images A, B, and C, may be the same or may be different widths as shown in FIG. Furthermore, although the case where the intervals between the slits are constant is shown, the intervals between the slits may be different.

[Brief explanation of the drawing]

Fig. 1 and Fig. 2 are diagrams for explaining conventional height measurement, Fig. 3 is a block diagram showing an embodiment of the present invention, and Fig. 4 is a diagram for explaining the operation of the present invention. . 1: light source, 2: exposed material, 3': slit plate,
4, 5: optical lens, 7: electron gun, 8: electron beam,
9: Electronic lens system, 10: Deflector, 11: Amplifier, 12: Computer, 13: Semiconductor array sensor, 14: Amplifier, 15: Arithmetic circuit, 16: Display device.

Claims (1)

[Claims] 1. An irradiation optical system that irradiates the surface of an object with light at a constant angle θ and forms a plurality of bright and dark images at minute intervals in the vicinity of the irradiation point, and condenses the light reflected from the irradiation point and an imaging optical system for forming bright and dark images, a semiconductor array sensor that is arranged on the imaging plane of the imaging optical system and is composed of a large number of semiconductor photodetecting elements, and a reference for each bright and dark image on the sensor. An apparatus for measuring the surface height of an object, comprising calculation means for determining distances from a position and averaging them, and in which the array length of a plurality of bright and dark images on the semiconductor array sensor is sufficiently shorter than the length of the sensor. .