JPS60206136A - Mask aligning process - Google Patents

Abstract

PURPOSE:To make mask aligning process feasible regardless of the groove depth of Frestnel zone target as a condition of incident ray interference by a method wherein the sectional shape of a target provided on a substrate to be processed is formed into tapered shape. CONSTITUTION:When a Frestnel zone target 8 is formed by etching the surface of substrate 5 in a sectional structure of target 8, the sectional shape of target grooves 12 is changed into tapered shape by etching process. Through these procedures, any reflected ray from tapered parts causing diffraction displays bright points 11 since the tapered part to not conform to the requirement of specific expression even if the depth of groove 12 conforms to the requirement of said expression causing interference. When the Frestnel zone target on a substrate surface is provided with tapered step difference, the conventional condition of aligning process infeasibility may be eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Technical Field of the Invention The present invention relates to a method for automatically aligning a substrate to be processed and a mask.

(b) Technical background Semiconductor devices such as ICs and LSIs are becoming smaller and more dense in order to achieve higher capacity and higher speed computers.
The pattern width and pattern spacing of the element used are extremely fine.

As a method for achieving this, electron beam exposure is used in place of conventional ultraviolet exposure, and ion beam exposure and X-ray exposure are being put into practical use.

If we take silicon (Si) as an example of a semiconductor material, the thickness of the semiconductor substrate (wafer) is about 500 μm and the diameter is 3.
There are various types, from 6 inches to 6 inches, and the maximum size of an IC chip is 10 mm square, so a large number of IC chips can be obtained from one semiconductor substrate.

Now, since an IC chip is formed with a multilayer structure, and the pattern width is reduced to about 1 μm, problems arise such as how to align a light source and a mask for exposing such a pattern.

In other words, in the conventional ultraviolet exposure method, the minimum processing line width is limited to approximately 1.5 μm due to the limitations of the wavelength used, so it cannot be used to expose such fine patterns, and the use of electron beams or X-rays with short wavelengths is not possible. Is required.

Furthermore, semiconductor elements are formed on semiconductor substrates using thin film formation technology and photolithography, but this requires multiple mask alignments, and as patterns become finer, alignment becomes more difficult. It becomes difficult.

The present invention relates to improvements in mask alignment methods.

(C) Prior Art and Problems Accurate mask alignment is required in the manufacturing process in which a substrate to be processed is subjected to selective oxidation, selective etching, and formation of a conductor pattern.

There are two types of exposure methods used for mask alignment: contact exposure method and projection exposure method, but the former has higher resolution performance and is also more suitable for methods that use non-charged rest, such as X-ray exposure. The contact exposure method is suitable for this.

The present invention relates to an alignment method used for such purpose.

Various methods have been put to practical use as alignment methods, and FIGS. 1(A) and 1(B) are typical alignment marks.

That is, a small alignment mark 1 is made on the substrate to be processed (substrate for short), and a large alignment mark 2 of a similar shape made on the mask is moved to this so that the centers of both coincide. Positioning is performed by this.

In this method, the substrate or mask is moved so that the margins between the small alignment mark 1 and the large alignment mark 2 are evenly spaced, and both operations are performed manually.

On the other hand, there is a method of performing optical alignment.

This method uses a Fresnel zone target, in which a pattern consisting of multiple concentric circles as shown in Figure 2 is created on both the substrate and the mask, and this is used as an alignment mark to align the bright spots of the diffracted light. Perform positioning by

Here, each concentric circle of the Fresnel zone target 3 whose plan view is shown in Fig. 2 is R・-(・AF)'' (1) However, R... Radius n...Positive number A...Wavelength F of the light used...There is a relationship between the focal length, and in the example, between R1 and R2, and between R3 and R4 A metal pattern such as chrome (Cr) is provided between the two.

If these conditions are met and the light is incident in parallel, the light will be diffracted and focused at a position F from the substrate surface.

FIG. 3 shows a method of using this Fresnel zone target 3 for positioning.The focal length is set to an arbitrary distance from the mask 4 and the substrate 5, and from this point R1, R2
+'R3......, mask 4 and substrate 5.
Create a Fresnel zone target above.

Here, the mask can be used by patterning a vapor-deposited film such as Or on a quartz substrate when ultraviolet rays are used as in the past, or by forming a pattern of a vapor-deposited film such as Or on a quartz substrate, or by using a boron nitride (BN) film when using X-rays. It is preferable to use a patterned gold (Au) or quintal (Ta) film.

Alignment marks for the Fresnel zone target on the substrate 5 are also made by selective etching.

Positioning the mask 4 and substrate 5 in this way, from above,
When the light 6 is projected in parallel, the reflected light from the Fresnel zone target 7.8 of the mask 4 and substrate 5 is diffracted and focused at a set distance, and a bright spot 9 appears.

Here, the intensity distribution of the diffracted light is distributed around the center position A of the Fresnel zone Kuget, as shown in Figure 5.
Bright spots 10 and 11 are the positions where the intensity of the diffracted light is maximum.

However, the bright spot position 10 of the diffracted light from the Fresnel zone target 7 of the mask 4 and the bright spot position 11 of the Fresnel zone target 8 of the substrate 5 are usually separated as shown in FIG. 5(A). In general, the positioning can be carried out by fixing the mask 4 and moving the substrate 5 so that they are aligned and the intensity of the diffracted light is maximized, as shown in FIG. 5(B).

If you use a Fresnel zone target and a photodetector in this way, you will not have to do it manually like in the past.
Automatic alignment becomes possible.

However, as shown in FIG. 4(A), the depth H of the groove of the Fresnel zone target 8 formed in the base Fi5 is H= (A/2 n,) xM ・-(2) However, Δ... ...Wavelength n of the incident light ...Refractive index M of the medium ...If the relation of positive numbers is satisfied, the reflected light from the Fresnel zone target 8 of the substrate 5 interferes with each other. There was a problem in that the diffracted light did not appear because of cancellation, and therefore alignment could not be performed.

(d) Object of the Invention The object of the present invention is to provide a method that enables alignment with a mask even if the depth of the groove of a Fresnel zone target formed on a substrate causes interference of incident light. There is something to do.

(e) Structure of the Invention An object of the present invention is to form a tapered cross-sectional shape of the target provided on the substrate to be processed when aligning a substrate and a mask for forming a fine pattern using a Fresnel zone target. This can be achieved by a mask alignment method characterized by the following.

(f) Embodiment of the Invention FIG. 4(B) shows a cross-sectional structure of a Fresnel zone target according to the present invention.

Here, when etching the substrate 5 to form the Fresnel zone target 8 on its surface, the etching depth H is accurately adjusted so that H=(Δ/2n)XM (2) In practice, it is difficult to maintain this relationship.

Therefore, the etching shape is changed so that the cross-sectional shape of the groove portion 12 of the target becomes tapered as shown in FIG. 3(B).

In this way, even if the depth of the bottom of the groove 12 satisfies the condition of equation (2) and interference occurs, a part of the tipper deviates from this condition, so the reflected light from this position will undergo diffraction and become a bright spot. Note that in order to create such a tapered groove portion 12, etching anisotropy of the crystal plane may be used.

For example, the 110 plane of a Si crystal is preferentially etched in the 100 direction, and by utilizing this property, a Fresnel zone cugett with a tapered cross section can be formed.

If the Fresnel zone target on the substrate surface is made to have a tapered step in this way, the S/N ratio will be lowered, but the condition that alignment cannot be achieved as in the conventional method can be solved.

Furthermore, by detecting the brightness of the diffracted light using a photodetector, highly accurate automatic positioning becomes possible.

(g) Effects of the Invention The present invention is based on the fact that in the manufacture of semiconductor devices such as LSIs, pattern widths have become finer and manual mask alignment has become increasingly difficult. Although desirable,
There was a problem in that it could not be used under conditions where light interference occurred.

However, by implementing the present invention, this problem can be solved and mass production efficiency can be improved.

[Brief explanation of the drawing]

Figures 1 (A) and (B) are explanatory diagrams of the conventional alignment method;
Figure 2 is a plan view of the Fresnel zone target, Figure 3 is an explanatory diagram of the positioning method using the Fresnel zone target, Figures 4 (A) and (B) are cross-sectional views of the target, and Figure 4 (A) and (B) are cross-sectional views of the target. Figure 5 (A). (B) is an intensity distribution diagram of diffracted light. In the figure, 3, 7.8 is the Fresnel zone target,
4 is a mask, 5 is a substrate to be processed, 6 is incident light, 9.10.
11 is a bright spot, and 12 is a groove.

Claims (1)

[Claims] When aligning the substrate to be processed and the mask for forming fine patterns using a Fresnel zone target,
A method for aligning a mask, comprising forming a cross-sectional shape of a target provided on the substrate to be processed into a tapered shape and aligning the target.