JPH0625646B2 - Alignment method - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a highly accurate alignment position, and more particularly to a alignment method applicable to a highly dense alignment device for a semiconductor device (hereinafter referred to as an LSI).

Structure of Conventional Example and Its Problems Semiconductor devices have been increasingly densified recently, and the size of the fine pattern of each element reaches 1 micron or less.
The conventional alignment of the photomask and the LSI wafer during LSI manufacturing uses the alignment marks provided on the wafer to rotate the stage on which the wafer is mounted and move it in two axes in parallel. However, the alignment accuracy is about ± 0.3 microns, and when forming a submicron element, the alignment accuracy is poor and it is not practical. In addition, S Austin (Applied Physics Letters.vo
l.31 NO.7 P. 428, 1977) et al., The method of alignment using the interferometry causes the incident laser beam 1 to be incident on the photomask 2 and diffracted by the grating 3 formed on the photomask 2 as shown in FIG. Then, the diffracted light is diffracted again by the grating 5 formed on the wafer 4 to obtain diffracted lights 6, 7, 8 ... This diffracted light is represented by a binary representation of the diffraction order on the photomask and the diffraction order on the wafer. Diffracted light 6 is (0, 1), diffracted light 7 is (1, 1), and diffracted light 8 is ( -1, 2) ... The diffracted light is collected by a lens at one point and the light intensity is measured. The diffracted light has a light intensity at a position symmetrical with respect to the incident laser beam 1, and the photomask 2 and the wafer 4 can be aligned by matching the intensities of the diffracted light observed on the left and right. In this method, the alignment accuracy is set to several hundred Å.

However, in this method, since the alignment between the photomask 2 and the wafer 4 is greatly affected by the distance D between the photomask 2 and the wafer 4, the accuracy of the distance D is required. In addition, the photomask 2 and the wafer 4 must be brought close to each other and aligned with the accuracy of the distance D maintained, which complicates the apparatus, which is a problem in practical use.

Also, for alignment of elements with submicron line width,
Although there is a method of observing by secondary electron emission from the element, it cannot be handled in the atmosphere, so that the throughput in manufacturing an LSI becomes small and there is a practical problem.

SUMMARY OF THE INVENTION The present invention has been made in view of such problems in the related art, and an object of the present invention is to provide a positioning method capable of positioning a fine pattern in the atmosphere with a simple structure.

According to the present invention, coherent light is incident from two directions.
Interference fringes obtained by interference of light fluxes and light reflected or transmitted by a grating arranged in the optical path of the two light fluxes are guided to a light detecting means, and the light intensity changes when the incident angle of the two light fluxes is changed. Is measured, and the pitch P of the interference fringes of the two light beams is measured.
The incident angle of the two light beams is adjusted so that the relationship between _f and the grating pitch P _G is P _G ≈P _f × N (N is 1, 2, ... By detecting the relative position with respect to the grid, it is possible to perform the alignment with high accuracy.

Description of Embodiments FIG. 2 shows a position detecting device equipped with a holographic exposure device and a photodetector capable of implementing the position detecting method according to the present invention.

A coherent light 10 is made incident on a beam splitter (BS) from a laser generator (not shown), and is amplitude-divided into a reflected light 11 and a transmitted light 12 having almost the same intensity, and each is reflected by a reflecting mirror M.
BS, M1, so that both reflected lights are incident on the surface of the wafer W at substantially equal angles θ.
Place M2 and W. A grating G is formed on the wafer W, and reflected lights 13 and 14 diffracted by the grating G are formed.
Enters the photodetectors D1 and D2 through the slits S1 and S2. Assuming that the pitch of the interference fringes formed by the interference of the reflected lights 11 and 12 from λ, M1, M2 with the laser wavelength is P _f , the interference fringes formed on the wafer G are It is represented by.

From the grating G having a pitch P _G that is an integer multiple of the pitch P _f of the interference fringes, the light diffracted by the grating that splits the interference light of the two light beams 11 and 12 by the wavefront is obtained. Light intensity information indicating the parallelism between the interference fringes of the light flux and the grating G and the relative positional relationship in the pitch direction can be obtained.

3 and 4 show the relative positional relationship of the interference fringes of the two light fluxes and the grating G before the alignment. f is an interference fringe, α is an angle formed by the interference fringe f and the grating G, and x is a positional deviation between the interference fringe f and the grating G. By slightly rotating the wafer W around the normal to the surface having the grating G and guiding it to the photodetectors D1 and D2 through the slits S1 and S2, a change in the light intensity I as shown in FIG. 5 can be obtained. . The vertical axis represents the light intensity I and the horizontal axis represents the rotation amount α.
_Further , by slightly moving the wafer W in the pitch direction of the interference fringes f of the two light fluxes and detecting the light intensity I, a change in the light intensity I can be obtained as shown in FIG. The vertical axis represents the light intensity I, and the horizontal axis represents the periodic change of the light intensity I for each pitch P _f of the interference fringes f of the movement amount x ₀ 2. In addition, a slight fluctuation of the light intensity I is
This is a variation caused by feeding a fine pitch at a certain interval. At the peak value of the light intensity I, the positional deviation x between the interference fringe f and the grating G is x = 0.

In the present embodiment, the wafer W is moved, but the same alignment can be performed by moving the two light fluxes. The order of the alignment in the rotation direction α and the alignment in the pitch direction x may be performed first, and finally, in the peak value of the light intensity I, the alignment in both directions is completed, The closer to the peak value of the light intensity I, the higher the alignment accuracy becomes.

The pitch P _G of the grating G is made to be an integral multiple of the pitch P _f of the interference fringes. However, there is a possibility that the pitch P _G of the grating _G is deviated by ΔP _{1 as} the wafer W having the grating G goes through each structural process of the semiconductor. There is. Further, there is an error of ΔP _{2 in} the pitch P _G due to the manufacturing accuracy error of the grating G, and in practice, the grating pitch P _G and the interference fringe pitch P _f are P _G + ΔP ₁ + ΔP ₂ = P _f × N, that is, Since P _G ≠ P _f × N (N is 1, 2, 3 ... Integer), the alignment accuracy becomes poor. Therefore, the change varying the incident angle θ of the two light beams, whereas any reflection mirror M1, M2, the by pitch P _f of the interference fringes f is changed, or the incident angle θ of microspheroidal now two beams both The change in the light intensity of the reflected lights 13 and 14 when the
When measured with 1 and D2, the light intensity change as shown in FIG. 7 is obtained. When the vertical axis is the light intensity I, the horizontal axis is the incident angle θ, and the relationship between the interference fringe pitch P _f and the grating pitch P _G is P _G = P _f × N, the light intensities of the reflected lights 13 and 14 are It will be maximum.
Therefore, by slightly rotating one or both of the reflecting mirrors M1 and M2 so as to reach the light intensity peak position in FIG. 7, the relationship between the interference fringe pitch P _f and the grating pitch P _G becomes P _{Since G} ≈P _f × N, the alignment accuracy between the interference fringes f of the two light fluxes and the grating G is further improved.

Further, when the incident angle θ is changed, the pitch P _{f of the} interference fringes _f
The relative position between the interference fringe f and the grating G may also change at the same time with the change of, and the angle α and the positional deviation x formed by the interference fringe f and the grating G may be α = Δα, x = Δx, There is a nature.

Therefore, the incident angle θ, the angle α formed by the interference fringes f and the grating G, and the positional deviation amount x are alternately adjusted so that the light intensity is maximized.
The alignment of the interference fringe f of the light flux and the grating G can be realized with higher accuracy.

As described above, according to the present invention, the light reflected or diffracted by the interference fringes obtained by the interference of the two light beams and the grating is guided to the photodetector, and the incident angle of the two light beams is changed. By measuring the change in light intensity, the relationship between the pitch P _f of the interference fringes of the two light fluxes and the pitch P _{G of the} grating is P _G ≈P _f
By adjusting the incident angle of the two light fluxes so that XN (N is 1, 2, 3 ... Integer) and detecting the relative position between the interference fringes of the two light fluxes and the grating, the interference fringes of the two light fluxes can be obtained. It is possible to further improve the alignment accuracy with the grid.

[Brief description of drawings]

FIG. 1 is a principle diagram of a conventional alignment apparatus, FIG. 2 is a configuration diagram of an apparatus that realizes an embodiment of an alignment method according to the present invention, and FIGS. 3 and 4 are interference fringes of two light fluxes and a grating. FIG. 5 is a relational diagram showing the relative position before alignment with respect to the interference fringes of the two light fluxes, and FIG. 5 shows the interference fringes of the two light fluxes and the rotation direction of the grating when a slit whose longitudinal direction is arranged substantially parallel to the interference fringes is provided. FIG. 6 is a diagram showing the light intensity dependency, FIG. 6 is a diagram showing the light intensity dependency in the pitch direction obtained by the alignment method obtained by the present invention, and FIG. 7 is a diagram when the incident angle θ of two light fluxes is changed. It is a figure which shows the light intensity dependence of the reflected lights 13 and 14. 10 ... light, 11 ... reflected light, 12 ... transmitted light, 13,
14 ... Reflected light, W ... Wafer, G ... Lattice, S1, S
2 ...... slit, D1, D2 ...... photodetector, f ...... interference fringes, the pitch of _{P f} ...... interference fringes, the pitch of _{P G} ...... grating.

Claims (2)

[Claims] 1. Coherent light is incident from two directions,
A substrate having a grating formed substantially parallel to the interference fringes obtained in space by the interference of the two light beams is arranged in the optical path of the two light beams, and the light reflected or transmitted by the grating portion is The relative position between the interference fringes of the two light fluxes and the grating is detected by guiding the light to the light detecting means through the optical system and measuring the output change of the light detecting means, and the substrate is moved in the direction to eliminate the relative position shift. In the alignment method for aligning the interference fringes with the grating formed on the substrate by moving the mounted stage, the relationship between the pitch Pf of the interference fringes of the two light fluxes and the pitch PG of the grating is PG≈. A positioning method, wherein the incident angles of the two light fluxes are adjusted so as to be Pf × N (N is an integer of 1, 2, ...). 2. The adjustment of the incident angle of the two light fluxes and the relative position adjustment of the interference fringes of the two light fluxes and the grating are appropriately repeated so as to maximize the light intensity. The positioning method according to claim 1.