JPH053144A - Exposure method - Google Patents

Abstract

PURPOSE:To eliminate an alignment error due to a fluctuation in a reference length between a projection optical system and a detection optical system. CONSTITUTION:A pattern of an alignment mark 32 of a reticle 9 is projected on a prescribed substrate 4 by a projection lens system 5, a latent image pattern 39 of the mark 32 is recorded in a photosensitive layer on the substrate 4, the substrate 4 is moved by a prescribed amount to send in the visual field of a detection optical system, a reference length between the lens system 5 and the detection optical system is decided using the position of the pattern 39 detected using the detection optical system and the movement of the substrate 4 and an off-axis alignment using the detection optical system is performed on the basis of this reference length.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure method used for manufacturing ICs, LSIs and the like, and more particularly to an exposure method involving a new alignment process.

[0002]

2. Description of the Related Art The basic performance required for a mask aligner is resolution performance and alignment accuracy. Another point would be processing capacity (throughput) because of its value as a production machine. In particular, with the miniaturization and higher integration of semiconductor elements, higher resolution performance and alignment product are required endlessly.

The mask aligner is roughly classified into a contact system, a proximity system, a 1: 1 mirror projection system, a lens projection system and the like depending on its exposure method, but at present, since a finer pattern can be printed, The reduction type lens projection system (so-called stepper) is becoming the mainstream.

The advantage of the reduction type lens projection system in terms of resolution performance is omitted because it has already been introduced in the literature and the like, but when viewed as a system considering alignment accuracy, this lens projection system causes a major obstacle. Have It's simply said that the projection lens has a certain wavelength, which is usually the exposure wavelength.
This is due to the fact that image formation and aberration are guaranteed only for

Among the alignment systems in the lens projection system which have been proposed or implemented so far, the above obstacles are overcome in the system using the light of the wavelength other than the wavelength guaranteed by the projection lens. However, all of them result in including an error factor newly generated by the method during the process from alignment to exposure. Classifying these errors from the aspect of character, A. Error due to indirect reference intervention, B. Movement of mask or wafer from alignment to exposure-error due to its feeding accuracy, C.I. Time required from alignment to exposure,
Movement error due to temperature and vibration of each element, D. It is the optical path difference between the exposure light and the alignment light between the mask and the wafer.

In the case of the system disclosed in Japanese Patent Laid-Open No. 59-79527, B and C errors are caused by Japanese Patent Laid-Open No. 55-135831.
In the case of the system using the additional lens of Japanese Patent Publication, C,
In the case of the system of Japanese Unexamined Patent Publication No. 58-145127, the error of D causes error factors such as A and D because the reticle and the wafer are not observed simultaneously.
Also, in the OFAXIS type stepper, all errors A to D occur. Although it is possible to reduce these errors, there is no doubt that stable dependence of the error on a low level will be a great burden on both the device manufacturer and the user.

[0007]

DISCLOSURE OF THE INVENTION Problems to be Solved by the Applicant
The system proposed in Japanese Patent No. 25638 has He / Cd having a wavelength very close to the exposure wavelength (g line = 436 nm).
The laser (442 nm) is used as an alignment light source, and the projection lens is corrected for these two wavelengths, thereby avoiding the occurrence of errors A to D.
However, in order to obtain higher resolution performance, projection lenses tend to use shorter wavelengths, in which case the benefits of the system cannot be exploited.
In addition, in order to use the resolution performance of the lens more effectively, it is necessary to reduce the constant pressure wave effect generated in the resist, and for that reason, the wafer surface is subjected to antireflection treatment, and the resist is filled with an absorber. Alternatively, the number of processes using multi-layer resists tends to increase in the future, and in such processes, it is more difficult to obtain a signal from a wafer in alignment by an exposure wavelength or a wavelength in the vicinity thereof. Is expected to become.

The present invention provides a new exposure method based on off-axis type alignment so as to be able to deal with a new process such as absorption resist or multi-layer resist while suppressing an error derived from the alignment system. The purpose is to

[0009]

According to the present invention, the position of an alignment mark on a substrate is detected by using a detection optical system arranged near the projection optical system, and the position between the projection optical system and the detection optical system is detected. In the exposure method of moving the substrate based on the reference length to align the substrate with the original plate and projecting the pattern of the original plate onto the substrate by the projection optical system, the original plate is projected by the projection optical system. The step of projecting the pattern of the alignment mark on a predetermined substrate and recording the pattern on the photosensitive layer of the predetermined substrate, and moving the predetermined substrate by a predetermined amount so that the photosensitive layer is in the visual field of the detection optical system. The method includes the steps of sending a recorded pattern and detecting the position of the pattern recorded on the photosensitive layer using the detection optical system to determine the reference length.

The concept of the present invention can be applied not only to conventional lens projection or mirror projection, but also to lens projection or mirror projection for further shortening the wavelength, and X-ray aligner.

The point of the present invention is to detect a partial change of the resist layer resulting from the exposure as a signal on the mask side. The resist causes an optical reaction by irradiation of light, which also causes a change in transmittance and refractive index in the optical sense, or in a special case, a step difference between the non-irradiated part and the surface due to expansion or contraction. .. Even in the commonly used OFPR or AZ resist, the result of selective exposure can be observed as a grayscale image under a white light microscope (this image will be referred to in the following description for convenience). The latent image of the mask printed by the regular exposure optical system is in a relationship that does not include an error in the relationship with the mask at that time, and the resist latent image and the wafer mark (step difference) below it. By reading the position error of (1), the relationship between images that are very close to each other on the same object is read, and extremely reliable detection that is not affected by the optical system or the like can be performed.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 2 shows a basic alignment exposure procedure which is the basis of the present invention. The errors that occur in this process are (a) detection error, (b) stage feed error, and (c) error due to wafer or mask movement within the time from pre-exposure to regular exposure. From the actual results of the current stepper, the feed amount of (b) is 2 to 3 μ at the maximum, and the time of (c) is about 1 second, and these errors can be suppressed to extremely small values.

Next, FIG. 1 shows an exposure apparatus to which the alignment exposure method according to the procedure of FIG. 2 is applied, and the function thereof will be described below. The entire system is assembled on the surface plate 1. (The structure is not shown). A wafer stage 2 is provided on the surface plate 1, and the wafer holding plate 3 and the wafer 4 sucked and held by the wafer holding plate 3 can be moved along a plane perpendicular to the optical axis of the projection lens 5. The position of the wafer stage 2 can be known by a known method of irradiating the optical mirror 6 provided on the wafer stage 2 with the light 7 of the laser interferometer, and the movement of the wafer stage 2 is controlled by a designated amount. Above the projection lens 5 is a reticle 9 held by a reticle holding table 8, and when the light is irradiated from the illumination optical system A above the reticle 9, the pattern incorporated in the reticle 9 is transferred to the wafer via the projection lens 5. 4 The structure is held so that it is transferred to the surface.

The illuminating optical system A is designed to irradiate the reticle 9 with the light emitted from the extra-high pressure mercury lamp 10 evenly.
The condenser lenses 11, 12 and 13 and the first and second mirrors 14 and 15 for bending the luminous flux.
The shirt 16 controls exposure. The second and third condenser lenses 12 and 13 and the second mirror 15 also show a surface which has an image forming relationship with the reticle pattern surface 17 in the same direction as the drawing B.
It is designed to be produced in a portion of the reticle 9, and by placing a masking on this portion, only a specific portion of the reticle 9 can be illuminated. On the surface B, the pattern exposure mask 19 held by the frame 18 and the alignment mark exposure mask 20 are switched and driven by the cylinder 21 so that they can be selectively inserted into the light beam.

An alignment optical system C is arranged between the projection lens 5 and the wafer 4 with a part thereof protruding. The light emitted from the halogen lamp 22 is condensed by the condenser mirror 23 and the condenser lens 24 and reaches the movable mirror 27 through the half prism 25 and the object lens 2. Movable mirror 27
When the light beam has a positional relationship of 45 ° with respect to the optical axis as indicated by the broken line, the light collected by the objective lens 26 irradiates the wafer surface. On the contrary, the light reflected from the wafer surface passes through the movable mirror 27, the objective lens 26, bends the optical axis upward by the half prism 25, passes through the relay lens 28, and forms a wafer image on the surface 30 of the image pickup tube 29. In the case of alignment, X, Y, O only by the planar alignment of the reticle and the wafer.
Since alignment of components in three directions is required, two objective lenses are required, and it is considered that there are two objective lenses in the embodiment.

FIG. 3 shows a flow chart as an example of the procedure of alignment and exposure in the system of FIG. In FIG. 3, the solid-lined frame is the part related to the present invention, and the broken-lined frame is within the range of the prior art.

4A and 4B show a reticle 9 used in the system in this system, alignment marks 31 and 32 on the reticle side, a circuit pattern area 33, and an area 34 corresponding to a scribe line.

5A and 5B, the same wafer 4, the alignment mark 35 on the wafer side, and the scribe line 3 are the same.
6, a portion 37 corresponding to the reticle circuit pattern is shown.

FIG. 6 is a plan view of a selective exposure mask corresponding to the reticle 9 shown in FIG. 4A. Reference numeral 38 is a window. In the step 4 of the flow of FIG. 3, the mask 20 on the left side of FIG. 6 is placed in the light beam, and the mask 19 on the right side is used in this step. Like the reticle, the mask may be an optical glass to which light-shielding chrome is selectively attached. However, it is preferable that the mask is transparent in consideration of foreign matters and the like being transferred onto the reticle and the wafer.

FIGS. 7A and 7B show patterns (reticle mark latent images) in which alignment marks 31 and 32 on the reticle 9 side are printed on the resist on the wafer surface and recorded. In the example of FIG. 7 (A), the reticle mark latent image 39 is imprinted with a slight deviation from the wafer mark 35. In this case, the wafer mark 35 is the seventh
This is a pattern formed by steps as shown in FIG. When this pattern is captured by the alignment optical system and imaged on the image pickup tube, one scanning line of the image pickup tube (for example, A → A)
The electric signal obtained from the signal is as shown in FIG. 7 (C). A method of separating the error in the x direction and the error in the y direction by using the secondary information obtained from the entire surface of the image pickup tube and outputting the error is already known, and a detailed description thereof will be omitted.

Alignment marks 31, 3 of reticle 9
The position error between the wafer pattern and the reticle pattern at the time of printing 2 is

[0022]

[Outer 1]

[0023] 2 in the case of a alignment mark places △ X _R as in this _{embodiment, △ Y R, △ X L} , △ Y Although four error amount as the _L is obtained, the four values If it is outside the allowable value, it is judged whether it is within the allowable value from

[0024]

[Outside 2] The stage on the wafer side is moved according to the value of. During these operations, the movable mirror 27 is retracted out of the light beam, and the mask is switched to the pattern exposure mask 19, so that the regular pattern portion can be exposed immediately after the stage is driven. As mentioned above, it was shown that the alignment method using the latent image can be sufficiently realized.

This method can be used not only for lens projection but also for other exposure methods. For example, even in an X-ray proxy exposure apparatus in which the printing result has a magnification error with respect to the mask size, the latent image is the "exposure result", so that the significance of this method can be exhibited.

The alignment optical system shown in FIG. 1 also functions as a pre-alignment detection optical system for coarse alignment, which is the role of the block 2 in the flow of FIG. 3, and is also used in a so-called off-axis alignment stepper. Alternatively, it may be used as a fine alignment detection optical system for detecting the position of the wafer alignment mark.

The present invention effectively utilizes the above-mentioned concept of latent image alignment for an off-axis microscope in this off-axis alignment stepper.

Among the error factors of off-axis alignment, the distance [reference length] between the optical axis of the projection lens and the optical axis of the off-axis microscope occupies a large proportion, but this reference length always depends on a constant value. Or, if it fluctuates, knowing the exact value at that time controls the accuracy of off-axis alignment. There are several proposals for measuring the reference length, but all of them are indirect methods and are likely to cause errors. The positional relationship between the image of the reticle pattern projected onto the wafer by the projection lens and the optical axis of the off-axis detection optical system is substantially the reference length when the pattern of the reticle to be printed is positioned.

Therefore, the mark as shown in FIG. 4 is moved into the visual field of the off-axis microscope by moving the stage of the latent image pattern printed on the resist on a certain wafer, and the position is detected by the off-axis microscope. By calculating the reference length from the detected value at that time, the movement amount of the stage, and the arrangement size of the mark on the reticle, it is possible to directly measure the reference length with a small error. In addition, according to this method, the reference length can be accurately measured, and the reference length can be confirmed even during the wafer processing routine without using a special tool or method, and the off-axis microscope is used. Highly accurate positioning becomes possible.

[0030]

According to the present invention, the pattern of the alignment mark of the original plate is projected onto the predetermined substrate by the projection optical system, the pattern is recorded on the photosensitive layer of the predetermined substrate, and the predetermined substrate is moved by a predetermined amount. The reference length between the projection optical system and the detection optical system is determined by sending it into the visual field of the detection optical system and detecting the position of the pattern recorded on the photosensitive layer using the detection optical system. Since the substrate is aligned using the detection optical system, it is possible to reduce the error due to the fluctuation of the reference length and always achieve accurate off-axis alignment.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of an exposure apparatus to which an alignment method which is the basis of the present invention is applied.

FIG. 2 is a flowchart for explaining an alignment method that is the basis of the present invention.

FIG. 3 is a flowchart of alignment one exposure by the apparatus shown in FIG.

4 is an explanatory view of a reticle used in the apparatus of FIG. 1, in which A shows an arrangement of various patterns on the reticle,
B indicates a reticle alignment mark.

5 is an explanatory view of a wafer exposed by the apparatus of FIG. 1, in which A shows arrangement of various patterns on the wafer,
B indicates a wafer alignment mark.

FIG. 6 is an explanatory diagram of a selective exposure mask.

FIG. 7 is an explanatory diagram of a latent image pattern and a wafer alignment mark, where A and B show a positional relationship between the latent image pattern and the wafer alignment mark, and C shows a latent image pattern detected through a detection optical system. A signal corresponding to the wafer alignment mark is shown.

[Explanation of symbols]

 A exposure optical system C detection optical system 4 wafer 8 reticle 32 reticle alignment mark 35 wafer alignment mark 39 latent image pattern

Claims (1)

Claim: What is claimed is: 1. A reference length between the projection optical system and the detection optical system for detecting the position of an alignment mark on a substrate by using a detection optical system arranged near the projection optical system. In the exposure method, in which the substrate is moved to align the substrate with respect to the original plate on the basis of, and the pattern of the original plate is projected onto the substrate by the projection optical system, the position of the original plate by the projection optical system. A step of projecting a pattern of alignment marks onto a predetermined substrate and recording the pattern on the photosensitive layer of the predetermined substrate, and moving the predetermined substrate by a predetermined amount to record the pattern on the photosensitive layer within the visual field of the detection optical system. And a step of detecting the position of the pattern recorded on the photosensitive layer using the detection optical system to determine the reference length.