JPS63128639A - Prealignment method - Google Patents

Abstract

PURPOSE:To contrive improvement in overlapping accuracy by a method wherein a mask and a wafer are set at the optimum position by providing a wafer stage with which a vacuum-attracted mask is transferred to a mask stage. CONSTITUTION:The wafer chuck 7 equipped with a mask 5 is pushed up in the Z-direction by a Ztheta stage 4, and after the mask 5 has been vacuum-attracted to a mask chuck 11, the vacuum state of the wafer 7 is removed. When, the Ztheta stage 4 is returned to the normal position, the mask 5 is vacuum-attracted to the mask chuck 11 in the state wherein the mask 5 is accurately positioned against a television camera 13. A wafer 6 is vacuum-attracted to the wafer chuck 7, the prealignment marks 18a and 18b are observed by the television camera 13, and a prealignment operation is performed. As a result, the mask 5 and the wafer 6 can be prealigned in highly overlapping accuracy.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a pre-alignment method, and particularly to a pre-alignment method applicable to an X-ray exposure apparatus.

[Technological environment]

With the trend towards higher integration, as exemplified by the recent RIDRAM, the minimum line width of VLSI patterns is moving from micron to submicron. I'm in a situation like this.

In conventional optical semiconductor exposure equipment using ultraviolet G-line and I-line, the resolution limit due to RT and light wavelength is 0.5 μm.
I was told that it was a degree? There is a strong desire for next-generation exposure equipment that can handle patterns of 0.5 μm or less. As this next-generation exposure device, research and development is currently underway on the XII Infinite Device.

[Common technology]

In general, XAM optical devices use divergent X-rays and cannot constitute an optical system such as a projection lens, so the exposure method is gloximity light.

The professional shimity gears and holes for masks and wafers are as small as K10 to 50 μm in order to reduce the penumbra, and it is necessary to strictly control the gear settings to reduce run-off errors.

Therefore, the alignment in the X-ray non-photographic device) ri,
In addition to aligning the mask and wafer in the lateral direction, gaps and tilts must be set accurately, and various alignment methods have been proposed for this purpose.

[Conventional technology] As a conventional technology, for example, 1 Nikkei Microdevice 1
There is an X 11M station, PA ``MX-16oo'' by Mike O-Kuss in the US, which was introduced in the April 1986 issue.

[Mask and wafer alignment in the MX-1600J uses a condensing lens that utilizes light diffraction called linear 7-renel zone play (L↑ZP) as the mask mark, and a linear wafer mark as the wafer mark. This alignment method using a diffraction grating is described in B.
Fay Ranyoshi Journal of Vacuum
8science Technology Vol, 16
(6) pp, 1954-1958. Now/Dec,
1979's "0ptical Alignment"
System for SubmicronX-ray
Reported in Lithography @. Here, the principle will be explained with reference to the drawings.

FIG. 3 is an explanatory diagram showing an alignment method using LFZP. A diffraction grating 2° is engraved on the wafer 19, and a mask 21 is opposed to the wafer 19 with a predetermined gap therebetween. An LFZP 22 whose focal length is equal to the gap between the mask and the wafer is drawn on the mask 21.

FIG. 4 is an explanatory diagram showing the structure of the LFZP of the mask mark. LFZPti has a structure with stripes of various widths and intervals, and the stripes are represented in the marks. Here, the fri focal length and λ are the wavelengths of the lasers used for alignment. FIG. 5 is an explanatory diagram showing the diffraction grating of the wafer mark. The zero diffraction grating has a structure in which rectangles of equal size are arranged at equal intervals.
The diffraction angle is determined by the angle d.

As shown in FIG. 3, the laser beam enters from above the mask 21, is focused by a nine-parallel laser beam 23riLFZP22, and is focused on the surface of the wafer 19 to form a slit-shaped image. When this imaged slab and the diffraction grating Zθ on the surface of the wafer 19 overlap in a straight line, the laser beam is diffracted and returns to the LFZP.
22 becomes parallel light and is detected as a positioning signal.

The aforementioned X-ray stepper "MX1600J" uses the mask mark LFZP to perform closed-loop automatic positioning.
The lateral angle of incidence of the laser beam 23 onto 22 is varied. As a result, the imaged slit scans the linear diffraction grating 2o, and the amount of deviation of the alignment mark can be detected in a "realistic manner".At this time, the scanning range of the slit on the surface of the wafer 19 is approximately Since it is 2 μm, the detection range for the amount of deviation of the alignment mark is also about 2 μm.

[Problem that the invention seeks to solve]

In the conventional alignment method described above, the detection range for positional deviation is as narrow as about 2 .mu.n1, so when mounting the ask and wafer on the stage, it is necessary to perform pre-alignment to achieve a certain degree of overlay accuracy.

Conventional mask and wafer pre-alignment) 1'i.

Since positioning is performed by mechanically abutting and positioning based on the respective external shapes, the overlay accuracy of pre-alignment is greatly influenced by the external shapes of the mask and wafer and the positional accuracy of the pattern.

Therefore, when manufacturing a mask, it is necessary to improve the positional accuracy of the mask pattern relative to the external shape.
The drawback was that it made it difficult to make masks. Therefore, if the overlay accuracy of pre-alignment is 2 μm or more, the positional deviation signal cannot be detected, and the stage is moved so that the mask and wafer are overlapped within the detection range of the 1-positional deviation signal, and the 1-positional deviation signal is found. This method has the disadvantage that it requires several steps and takes a long time for alignment.

[Means for solving problems]

Pre-alignment method of the present invention ri, xy stage and z
A process of attaching a mask with a pre-alignment mark to a wafer stage consisting of a zero fine movement stage and a wafer chuck capable of vacuum suctioning both UNO 1 and the mask, a process of moving the UNO stage in front of a television camera, and a process of moving the UNO stage in front of a television camera. A wafer stage is installed in front of the mask stage, which consists of a Zl, zl, and zs fine movement stage that can adjust tilt and height, and a mask chuck that can vacuum hold the mask. , a step of transferring the mask from the wafer stage to the mask stage by controlling the state of the Z stage of the wafer stage, the wafer chuck, and the mask chuck, and mounting a wafer having a pre-alignment mark on the wafer stage. The process includes a step of moving the wafer stage in front of the television camera, a step of observing the pre-alignment mark on the wafer with the television camera and performing alignment in the XYθ directions, and a step of moving the wafer stage to the mask stage. configured.

〔Example〕

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

FIG. 1 is a perspective view showing an embodiment of the present invention.

The X-ray exposure apparatus shown in Fig. 1 consists of a base 1, an X stage 2 mounted on the base 1 and capable of moving in the X-axis direction, an X stage 3 mounted on the 3 and a 2θ stage 4 that can move in the Z-axis direction and rotate around 2 axes;
A wafer chuck 7 is mounted on the wafer chuck 7 which can vacuum hold both the mask 5 and the wafer 6, and an X-ray chuck 7 is mounted below the
a helium chamber 9 for guiding the X-rays from the radiation source 8;

Zl, zl, Z3 fine movement stages lO are installed below the helium chamber 9 and can be adjusted in height.
l, z, Z3 The mask chuck 11 attached to the fine movement stage 10, the alignment optical system 12 attached to the side of the helium chamber 9 for performing final fine alignment, and the mask chuck 11 are located near the mask chuck 11. Two television cameras 13 that can observe the pre-alignment marks of the mask 5 and wafer 6 mounted on the reed wafer, and a right-angle mirror 14 mounted on the X stage 3.

It is configured to include a laser length measuring device 15 that measures the position of the X stage 3.

FIG. 2 (al to Th1ri) is an explanatory diagram showing a process of pre-alignment of a mask and a wafer in the X-ray exposure apparatus shown in FIG. 1.

FIG. 2(a) rt, the mask 5 is vacuum-adsorbed to the wafer chuck 7, and the X stage 2 and the X stage 3 are placed in a position where pre-alignment is performed by observing the two pre-alignment marks 16a and 16b with the television camera 12. It shows where you have moved. The distance between the two television cameras 13 is equal to the distance between the pre-alignment marks 16a and 16b.

Prisoner 2 (b) rt, pre-alignment mark 16a.

16b is shown on the television monitor when observed with the television camera 13. The amount of deviation in the X direction between the observed pre-alignment mark 16a and the reference mark 17a on the monitor is assumed to be Δxa, the amount of deviation in the Y direction is 67 m, and the amount of deviation in the X direction between the pre-alignment mark 16b and the reference mark 17b is - b-x-.

When the punch weight Δxb and the displacement amount in the Y direction are Δyb, the displacement amount of the mask 5 in the XY19 direction ΔXrm, ΔY − tΔθ
m is Δx=(ΔX−+ΔX1l)/2. Δy=(Δya
+Δyb)/2. It is expressed as Δθ=(Δya−Δyh)/D. However, Dji pre-alignment mark 16
The distance is a, 16b. Based on these ΔX, Δy, and Δθ, K.

X stage 2. X stage 3. Move the Zθ stage 4 and mark the pre-alignment mark 16” e 16 on the monitor.
Pre-alignment of the mask is performed by overlapping the reference marks 17a and 17b.

FIG. 2(C) ri, after completion of mask prealignment.

This shows the state when the X stage 2 and the X stage 3 are moved below the mask chuck 11. The position of the stage is measured by using the position where pre-alignment is completed as the origin and measuring the displacement from that position with the laser length measuring device 15.

FIG. 2(d) shows the state when the mask 5 is transferred from the wafer tray 7 to the mask tray 11. When the Zθ stage 4 pushes up the wafer chuck 7 on which the mask 5 is attached in two directions to bring the mask 5 into close contact with the mask chuck 11, the mask chapter 1iri mask 5 is vacuum-adsorbed. After the mask 5 is vacuum-adsorbed on the mask chuck 11, the vacuum of the wafer chuck 7 is turned off and the mask 5 is transferred.

FIG. 2(e) shows the state when the transfer of the mask 5 is completed and the 2θ stage 4 has returned to its normal position. At this time, the mask 5ri is vacuum-adsorbed to the mask chuck 11 while being accurately positioned with respect to the television camera 12.

FIG. 2(f)rj, the wafer 6 is vertically attracted to the wafer chuck 7, and the X stage 2 and the X stage 3 are placed in a position where pre-alignment is performed by observing the two pre-alignment marks 18a and 18b with the television camera 13. It shows where it has been moved. Pre-alignment mark 18 on wafer 6
The distance between a and 18b is also made equal to the distance between the pre-alignment marks l 6 a and 16b on the mask 5.

FIG. 2(g) rt, pre-alignment mark 18a.

It shows the state of the television monitor when 18b is observed with the television camera 13.

11. - As in the case of the mask described above, the deviation amounts ΔXw, Δy v and A e w of the wafer 6 in the XYθ directions are determined, and the X stage 2. X stage 3. Move Zθ stage 4.

Pre-alignment is performed by superimposing the pre-alignment mark 183.18b and the reference marks t'ya, t'yb on the monitor.

FIG. 2(h)d, after completion of glial implantation of wafer 6,
This shows the state when the X stage 2 and the X stage 3 are moved to the initial exposure position. The position ri of the wafer 6 with respect to the television camera 12 can be accurately measured with a laser length measuring device as in the case of the mask 5. Since the mask 5 has already been accurately positioned with respect to the television camera 13, it is possible to pre-align the mask 5 and the square 6 with high overlay accuracy.

After the pre-alignment of the mask and the sea urchin is completed, final fine alignment is performed by the alignment optical system 12 and exposure is started.

〔Effect of the invention〕

The pre-alignment method of the present invention uses vacuum suction to create nine masks! By providing a wafer stage that can be transferred to the second stage, the mask and wafer can be pre-aligned on the wafer stage, which has the effect of allowing good pre-alignment with high overlay accuracy. As a result, there is an effect that the burden of final fine alignment is reduced.

In addition, instead of performing pre-alignment based on the mask outline, the mask is mounted on a wafer stage and the wafer stage is moved for alignment, so there is no need to provide positional accuracy between the mask outline and the mask pattern. This has the effect of making production relatively easy.

[Brief explanation of the drawing]

FIG. 1 is a perspective view showing one embodiment of the present invention, and FIG. 2 (al.
~(h)ri The process diagram of the pre-alignment method shown in Fig. 1, Fig. 3 is an explanatory diagram showing the conventional alignment method, and Fig. 4 shows the linear Fresnel mark which is a mask mark in the alignment method shown in Fig. 3. - A plan view showing a zone putto; FIG. 5 is a plan view showing a diffraction grating which is a mark for a wafer in the alignment method shown in FIG. 3; 1...Base, 2...X stage, 3...
・...Y stage %4...2θ stage, 5...
...-Mask, 6...Wafer, 7...
・Wafer chuck, 8...-X-ray source, 9...
・Helium chamber, 10...z1* Z goose*
Z3 fine movement stage, 11...mask chuck. 12... Alignment optical system, 13...
・TV camera, 14...Right angle mirror, 15・
...Laser length measurement W, 16a, 16b...
Mask pre-alignment mark, 17a, 171)・
...Reference mark, 18as18b...Wafer pre-alignment mark. 19...Wafer, Zθ...Diffraction grating,
21...Mass/, 22...Linear 7
Renel zone plate, 23...Laser beam, ΔXa, Δxb-...Amount of deviation in the X direction between pre-alignment mark and reference mark, Δya, Δyb.
...The amount of deviation in the Y direction between the pre-alignment mark and the reference mark, d...The pitch of the diffraction grating.呵 １ Concave Z diagram

Claims (1)

[Claims] A first mounting step of mounting a mask having a pre-alignment mark on a wafer stage consisting of an xy stage, a Zθ fine movement stage, and a wafer chuck capable of vacuum suctioning both the wafer and the mask, and moving the wafer stage in front of a television camera. a first movement step, and observing the pre-alignment mark of the mask with the television camera
The wafer stage is moved in front of the mask stage, which includes a first adjustment step in which adjustment is performed in the θ direction, and Z_1, Z_2, and Z_3 fine movement stages that can adjust tilt and height, and a mask chuck that can vacuum-hold the mask. a second movement step; a transfer step of controlling the vacuum states of the Z stage, wafer chuck, and mask chuck of the wafer stage to transfer the mask from the wafer stage to the mask stage; and providing a pre-alignment mark on the wafer stage. a second mounting step in which the wafer is mounted; a third moving step in which the wafer stage is moved in front of a television camera; and a second step in which the pre-alignment mark on the wafer is observed with the television camera and alignment is performed in the xyθ directions and a third moving step of moving the wafer stage to the mask stage.