JPS5831728B2 - Electron beam exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a novel electron beam exposure method suitable for drawing special figures such as trapezoids and triangles.

BACKGROUND ART Recently, an electron beam exposure method has been proposed and is beginning to be put into practical use, in which the cross-sectional shape of an electron beam is varied depending on the pattern shape.

According to such an exposure method, high speed can be achieved while maintaining high precision, so it is highly anticipated as a means of manufacturing VLSI.

However, in actual exposure, although a sufficiently high finishing accuracy can be obtained for a rectangular cross section, a sufficient finishing accuracy is not necessarily obtained for special shapes such as a trapezoid or a triangle.

When drawing a trapezoid using a variable cross-section electron beam exposure device, as shown in FIG. 1, an electron beam 1 shaped into a long and narrow strip (length l x width W) is 1 a y 1 b.

A method of scanning and moving in the width direction while varying the length l in the order of 1 c t 1 d is used.

In this case, the movement of the electron beam is performed digitally, and the strip-shaped cross section is moved little by little (for example, 0.05 μm) and shot, and the width W of the strip-shaped cross section is, for example, about 0.5 μm. The electron beams are often shot while overlapping each other.

In this case, the shot time Tx of the electron beams shot while overlapping is To, the shot time required to expose a figure alone, and P is the movement pitch.

Letting the beam cross-sectional width be W, it is determined by the following equation.

Tx=ToX (P/W)・・・・・・・・・・・・
...(1) Therefore, now W-0.5μm, P=0.05
The shot time Tx in the case of .mu.m can be set to 1/10 of the time required for individually shot electron beams.

However, it has become clear that when a shot time is given using this method, a trapezoid may be exposed with a size significantly different from the predetermined size, or in extreme cases, a trapezoid may not be exposed at all.

The present inventor has made various considerations regarding this point and has found that the above phenomenon is caused by the measurement accuracy of the width W of the band-shaped electron beam cross section on the sample when drawing a trapezoid.

That is, the above shot conditions (W=0.5 μm, P=
0.05 μm), only the portion where the 10 band-shaped cross-section electron beams are overlapped has the desired electron beam current amount,
Since the sensitivity characteristics of the photoresist coated on the sample are adjusted so that a predetermined exposure is performed, as a result, a trapezoid with a predetermined dimensional accuracy is exposed.

However, if the width W is smaller than a predetermined value (0.5 μm),
A portion where the band-shaped cross-sectional electron beams should be overlapped a predetermined number of times (10 times) is overlapped only 9 times or less, a trapezoid smaller than the predetermined size is exposed, or a trapezoid with an exposure area is exposed, It has been found that in extreme cases, the total amount of electron beam current to the resist becomes insufficient and no trapezoids are exposed.

On the other hand, if the width W is larger than the predetermined value, the portion that would not normally be exposed by overlapping nine times or less will be exposed by overlapping 10 times, and at the same time the total electron beam current amount will increase significantly. As a result, it was found that a trapezoid having a size larger than a predetermined size was exposed.

The present invention was made based on such consideration, and the amount of change in the total electron beam current due to deviation of the cross-sectional width of the band-shaped cross-section electron beam from a predetermined value is determined by the width and pitch of the band-shaped cross-section electron beam. This method attempts to compensate for the shot time by controlling the shot time, which involves shaping the cross section of the electron beam emitted from the electron source into a long and narrow strip, and partially directing the shaped electron beam onto the sample surface in the width direction of the strip. When drawing the desired pattern by moving it while overlapping it,
The electron beam cross section is made into a square or rectangular shape in advance as a reference, and the current value of the electron beam in the standard cross section is measured.Then, the current value of the electron beam in the band-shaped cross section is measured, and the current value of the electron beam in the band-shaped cross section is measured. An ideal electron beam current value of the band-shaped cross section is determined from the product of the measured current value of the electron beam and the area ratio of the band-shaped cross section and the reference cross section, and the ideal electron beam current value and the measured electron beam current value of the band-shaped cross section are calculated. The present invention provides a novel electron beam exposure method in which the actual shot time of the strip-shaped cross-section electron beam is determined based on the ratio of the electron beam to the electron beam, and the exposure is performed according to the shot time.

Hereinafter, one embodiment of the present invention will be described in detail with reference to the drawings.

In FIG. 2 schematically showing an embodiment of the present invention, 1 is an electron gun, and an electron beam generated by the electron gun is focused by an irradiation lens 2 and enters a beam cross-sectional shape variable device 3.

The beam cross-sectional shape variable device 3 includes a first slit 4 having, for example, a rectangular electron beam passage hole H1. It is composed of a second slit 52 having a similarly rectangular electron beam passage hole H2, and a lens 6 and a deflector 7 placed between the two.

The lens 6 has a magnification of 1:1, for example, and is arranged so that the image of the first slit 4 is formed on the second slit 5.

The passage holes H1 and H2 of the slits 4 and 5 are formed to have the same size and shape, and when the deflector 7 is not activated, the image of H1 exactly overlaps with H2, and the electron beam with the largest cross section passes through the second slit 5. .

On the other hand, if the deflector 7 is operated, the image of the passage hole H1 of the first slit 4 projected onto the second slit 5 will move.
The cross-sectional shape and size of the electron beam passing through the slit 5 change.

The electron beam having a predetermined cross-sectional shape and size thus obtained is projected onto a material 9 by a projection lens 8, and is scanned by a deflector 10 so as to draw a desired figure.

Reference numeral 11 denotes a blanking deflector disposed between the focusing lens 2 and the beam cross-sectional shape variable device 3. When the blanking signal from the blanking circuit 12 is supplied to the deflector, the electron beam is Since the electron beam is largely deflected and irradiation of the electron beam to the sample 9 is prevented, the shot time is controlled by the blanking deflector 11.

13, 14, and 15 each store the shot time for drawing an independent square or rectangle, a trapezoid parallel to the X axis (abbreviated as X-1 trapezoid), and a trapezoid parallel to the Y axis (abbreviated as Y trapezoid). The signals representing the shot time from each of these registers are supplied to the discrimination circuit 16.

The discrimination circuit 16 is supplied with data for specifying a drawing pattern from the electronic computer 17, and the discrimination circuit selects the square or rectangular drawing shot time storage register 13. according to the content of the supplied data. X one trapezoid drawing shot time storage register 14. ￥1 trapezoid drawing Schiff 1
A signal representing the shot time from one of the time storage registers 15 is supplied to the blanking circuit 12 via the discrimination circuit 16.

The signal related to pattern data from the electronic computer 11 is D.
After being converted into an analog signal by the /A converter 18,
After being amplified by an amplifier 19, it is supplied to a deflector 7 to create an electron beam with a desired cross-sectional shape.

Further, drawing position coordinate designation data for designating coordinates at which a pattern should be drawn from the electronic computer 17 is sent to the D/A converter 2.
After being D/A converted at 0, the signal is amplified at amplifier 21 and supplied to deflector 10.

22 is a Faraday cup for measuring the electron beam current, and the output signal of the Faraday cup 22 is sent to an amplifier 23.
After being amplified by the A/D converter 24, the
It is converted into D and supplied to the electronic computer 17.

In the configuration as described above, for example, first, a signal related to pattern data for drawing an independent square of 125 μm in length and width as shown in FIG. An electron beam having a square cross section is taken out from the shape variable device 3.

At this time, the computer 17 supplies the deflector 10 with a signal carrying drawing position coordinate designation data for projecting the square pattern onto the Faraday cup 22.

As a result, an electron beam with a square cross section is projected onto the Faraday cup, and the detected current is amplified by the amplifier 23 and further converted into a digital signal by the A/D converter 24, which is then supplied to the electronic computer 17. Ru.

The electronic computer 17 stores in advance data related to the sensitivity of the photoresist applied on the sample 9.
The detection current (data related to this).

The electric power calculator 17 calculates the shot time To required for drawing an independent square based on the data regarding the square or the sensitivity. It is stored in the register 13.

Next, to draw an X trapezoid: As shown in Figure 3, the longitudinal direction is parallel to the
Pattern data for drawing a 2.5 μm band-shaped pattern is supplied from the electronic computer 17 to the deflector 7 via the D/A converters 18 and 19, and an electron beam having a band-shaped cross section is produced.]
r Love-cup 22.

Data representing the value of the delivery current of the Faraday cup 22 is supplied to an electronic calculator 17.

The electronic computer 17 performs the following calculations.

Detection beam current Io when the reference square cross-section beam is shot and area of the strip-shaped cross section shaped to draw the X trapezoid (12.5 x 0.5 μm2)
and the ratio (0.5/12.5) of the area of the reference cross section (12.5 x 12.5 μm2) (I).

The ideal beam current Ii of the strip-shaped cross-sectional beam is calculated from X O,5/12.5 ).

Then, based on equation (1) above, the shot time To required for exposing a figure alone is multiplied by the ratio of the moving pitch P and the width W of the band-shaped cross-sectional beam, and the band-shaped cross-sectional beams are superimposed and shot. Ideal shot time (Tx
) and the ratio (I i / I
) and (TxXIi/I) to calculate the actual shot time Tx' in which the band-shaped cross-sectional beams are shot while overlapping each other in order to draw a trapezoid of X.

Note that the reference cross-sectional beam may be square or rectangular, but it is sufficient to use a reference cross-section in which the change in electron beam current due to deviation in cross-sectional width is significantly smaller than that of a strip-shaped cross-sectional beam.

Data representing the shot time Tx' determined in this way is stored in the X-one trapezoid drawing shot time storage register 14.

In exactly the same manner, an electron beam having a band-shaped cross section with the longitudinal direction parallel to the y-axis as shown in FIG. Then, the shot time T y / for drawing the strip figure is calculated, and the data corresponding to the Ty / is stored in the ¥1 trapezoid drawing shift 1 hour storage register 15.

In this way, when actually drawing on the sample 9, the electronic computer 17 sends pattern data to the discrimination circuit 16 and the beam cross-sectional shape variable device 3, and also sends the drawing position coordinate designation data to the deflector 10. The circuit 16 determines the drawn figure based on the supplied pattern data,
If the figure to be drawn is, for example, a one-X trapezoid, the one-X trapezoid drawing shot time storage register 14 supplies a signal representing a shot time Tx' to the blanking circuit 12.

As a result, an X trapezoid is drawn by sequentially drawing band-shaped figures having a longitudinal direction in the X-axis direction at a shot time Tx' and shifting by a pitch P using a blanking signal from the blanking circuit 12.

Further, if the figure to be drawn is a Y trapezoid drawing, the discrimination circuit 1
At 6, a signal corresponding to the shot time of Ty/ from the Y trapezoid drawing shot time storage register is supplied to the blanking circuit 12.

As a result, the blanking signal from the blanking circuit 12 causes the \1 trapezoid to be drawn by sequentially drawing band-shaped figures having the longitudinal direction in the y-axis direction at a shot time T y/.

If the figure to be drawn is a square, data from the square or rectangle drawing shot time storage register 13 is supplied to the blanking circuit 12 via the discrimination circuit 16, and the square is drawn at the shot time TO.

According to the present invention, when drawing a trapezoid or a triangle while superimposing band-shaped cross-sectional beams, even if the cross-sectional width of the band-shaped cross-sectional beams deviates from a predetermined value, the portion that needs to be overlapped 10 times is changed to 9 times or 11 times. Even if the entire rectangle or triangle is irradiated with a predetermined amount of current, it is possible to expose a trapezoid or triangle with approximately predetermined dimensional accuracy. will not occur.

[Brief explanation of drawings]

Fig. 1 is a diagram for explaining the case of drawing a trapezoid using band-shaped figures, Fig. 2 is a diagram for showing an embodiment of the present invention, and Fig. 3 is a diagram for showing various basic drawing figures. This is a diagram. 1: Electron gun, 2: Focusing lens, 3: Beam cross-sectional shape variable device, 4, 5 Nislit, 6: Lens, 7: Deflector, 8
: Projection lens, 9: Sample, 10: Deflector, 11 Deflector for nib blanking, 12: Flanking circuit, 13: Square or rectangular drawing shot time storage register, 14: X one trapezoid drawing shot time storage register, 15: Y one-trapezoid drawing shot time storage register, 16: Discrimination circuit, 17:
Electronic computer, 18, 20: D/A converter, 19° 21. 23: Amplifier, 22: Faraday cup, 24
: A/D converter.

Claims (1)

[Claims] 1 The cross section of the electron beam emitted from the electron source is shaped into a long and narrow strip, and the shaped electron beam is moved on the sample surface while partially overlapping the strip in the width direction to draw a desired pattern. When performing this, the electron beam cross section is made into a square or rectangular shape as a reference in advance, and the current value of the electron beam in the reference cross section is measured, and then the current value of the electron beam in the band-shaped cross section is measured, An ideal electron beam current value of the strip-shaped cross section is determined from the product of the measured current value of the electron beam of the reference cross section and an area ratio of the strip-shaped cross section and the reference cross section, and the ideal electron beam current value and the strip-shaped cross section are measured. An electron beam exposure method characterized in that the actual shot time of the band-shaped cross-sectional electron beam is determined based on the ratio with the electron beam current value, and exposure is performed according to the shot time.