JPS6254436A - Exposure device - Google Patents

Abstract

PURPOSE:To enable detection with high precision by a method wherein the minimum values are successively detected until the change in physical properties of a photoresist layer in case of exposure to light is stabilized so that a means to assume the electric signal component corresponding to the edge of alignment mark to be the component with stabilized minimum value may be provided. CONSTITUTION:When an alignment mark 12 buried in a photoresist layer 23 is illuminated, the photoresist layer 23 covering the alignment mark 12 is exposed to light changing the physical properties thereof so that the alignment mark detection signal VDIN transmitted from alignment mark signal forming means 18, 19 may change the electric signal components correspondingly to the change of photoresist layer 23. The electric signal component corresponding to the edge face 12B of alignment mark 12 out of those components transmitted from the alignment mark signal means 18, 19 constantly represents the minimum value and the position represented by this minimum value can be assumed to be the position of alignment mark 12. Through these procedures, the alignment mark 12 can be detected under no substantial influence of the photoresist 23 to perform the alignment with high precision.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an exposure apparatus, and is particularly suitable for application to an alignment apparatus for a reduction projection exposure apparatus.

[Conventional technology]

A reduction projection exposure apparatus is used in semiconductor manufacturing equipment in the manufacturing process of exposing a photoresist coated on a semiconductor wafer to an exposure pattern. It is designed to be projected upward.

In such a semiconductor manufacturing process, it is necessary to repeatedly expose multiple patterns to one wafer.
An alignment mark is provided on the wafer, the alignment mark on the wafer is reflected onto the reticle through a reduction projection lens, and this reflected image is optically aligned with the alignment mark provided on the reticle. It is designed to align the exposure pattern of the reticle with the wafer on the stage, and this method has been conventionally used as a TTL (through the 1ens) method alignment device.

[Problem that the invention seeks to solve]

However, in this type of TTL alignment type exposure apparatus, when optically aligning the alignment mark provided on the wafer through the reduction projection lens,
Since the photoresist layer coated on the wafer covers the alignment mark so as to embed it, the photoresist layer covering the alignment mark is exposed to the alignment illumination light, and as a result, the edge portion of the alignment mark There is a problem that the detection accuracy of the position of is degraded.

In principle, detection of such alignment marks is possible as shown in Fig. 4 (
This is done using the method shown in A). When trying to detect reflected light LRF by irradiating the alignment mark 2 on the alignment mark 2 with the alignment illumination light LIN, the surface 2 of the alignment mark 2 out of the alignment illumination light LIN
Since the light irradiated onto A is reflected with a sufficient reflectance, the amount of reflected light LRF becomes considerably large.

On the other hand, since the edge portion 2B of the alignment ake 2 is sloped, the alignment illumination light L
Since IN is reflected in a direction that does not return to its original state, the light intensity of the portion of the alignment reflected light LRF corresponding to the edge surface portion 2B falls to a much lower level than the light reflected to the front portion 2.

Further, in a portion outside the edge surface portion 2B, the alignment illumination light LIN is reflected by the surface of the wafer 1 with a high reflectance. In reality, the light intensity of the portion of the alignment reflected light LRF that corresponds to the surface of the wafer 1 is:
The amount of light reflected by the surface portion 2A of the alignment mark 2 is comparable to that of the reflected light.

However, as shown in FIG. 4(B), when the photoresist layer 3 is coated on the wafer 1 and the alignment mark 2, the distribution of the amount of reflected light LRF with respect to the alignment illumination light LIN is different from that of the photoresist layer 3. The influence becomes even more complex.

First of all, when the photoresist layer 3 is illuminated with the photosensitive alignment illumination light LIN, the photoresist layer 3 is affected by the illumination light and its physical properties change over time. There is a time. Conventionally, in order to avoid being affected by this unstable change, this problem was avoided by waiting for the detection operation until the photoresist layer 3 changed to a stable state.

The second problem is that when the surface of the photoresist layer 3 is illuminated with the alignment illumination light LIN, the photoresist layer 3
There is light reflected at the surface of the wafer 1, and light transmitted through the photoresist layer 3 and reflected at the surface portion 2A of the alignment mark 2, the edge surface portion 2B, and the surface of the wafer 1, and these two types of reflected light coexist. The reflected light LRF for alignment is reflected.

However, the phase of the reflected light that is reflected through the photoresist layer has a phase relationship that is different from the phase of the reflected light that is reflected on the surface of the photoresist (Jii3), so the alignment reflected light LRF as a whole When viewed, it produces interference fringes. Therefore, since the brightness and darkness based on the interference fringes are superimposed on the brightness and darkness based on the reflected light described above with respect to FIG. This is unavoidable and may cause alignment errors.

The present invention has been made in consideration of the above points, and utilizes the fact that the interference fringes of the reflected light LRF for alignment change from the start of exposure of the photoresist layer to the time when they become stable. By removing this, alignment marks can be detected with high accuracy.

[Means for solving problems]

In order to solve this problem, in the present invention, a photoresist layer 23 is deposited on the wafer 11, and in an exposure apparatus that exposes an exposure pattern to the photoresist layer 23, the above-mentioned photoresist layer 23 is embedded in the photoresist layer 23. alignment mark detection means 8.15 and 6.17.9 for irradiating the alignment mark 12 provided on the wafer 11 with illumination light and transmitting the resulting reflected light;
Alignment mark detection means 8.15.16.17.9
Alignment mark 1 based on the reflected light obtained from
2, the edge surface 12B of the alignment mark 12, the surface of the wafer 11, and the photoresist layer 2.
Alignment mark signal forming means 18 for obtaining an alignment mark signal VDIN comprising an electrical signal component corresponding to the amount of reflected light reflected from each surface of 3.
19 and the alignment mark signal VDIN, extreme values that occur during the time period until the change in the characteristics of the photoresist layer 23 upon exposure is stabilized are sequentially detected over time, and the position of occurrence of the detected extreme values is detected. The electric signal component having an extreme value with no fluctuation is transferred to the edge surface 2 of the alignment mark 12.
A determination means for determining that the information corresponds to B is provided.

[Effect]

When the alignment mark 12 embedded in the photoresist layer 23 is illuminated with illumination light, the photoresist layer 23 covering the alignment mark 12 is exposed to light and its physical characteristics change, and the alignment mark signal forming means 18 Alignment mer obtained by .19.

The electrical signal component included in the detection signal VDIN changes in accordance with changes in the photoresist layer 23.

However, among the electrical signal components sent out from the alignment mark signal forming means 18 and 19, the component corresponding to the edge surface portion 12B of the alignment mark 12 always exhibits an extreme value. It can be determined that the position is the same.

As a result, the alignment mark 12 can be detected and the coordinate position of the alignment mark 12 can be determined without being substantially influenced by the photoresist layer 23.

〔Example〕

An embodiment of the present invention will be described in detail below with reference to the drawings.

In FIG. 1, 8 is a reticle, and its exposure pattern area 8A is reduced by a reduction projection lens 9 and projected onto a wafer 11 placed on a stage 10. At an outer position of the pattern 11A projected onto the wafer 11, an alignment mark 12 similar to that described above with respect to FIG.
(FIG. 3(A)), and the illumination generated in the illumination system 15 provided above the reticle 8 passes through the half mirror 16 and illuminates the reticle 8 with respect to this alignment mark 12. The alignment mark 17 provided outside the exposure pattern area 8A is irradiated.

The alignment marks 17 are slit-like marks having a wider interval than the alignment marks 12 on the wafer 11, and the illumination light that has passed through the alignment marks 17 passes through the reduction projection lens 9 and aligns the alignment marks 1 on the wafer 11.
Irradiate 2. In this way, the reflected light generated from the alignment mark 12 and the surface of the wafer 11 in the vicinity passes through the reduction projection lens 9 again, passes through the alignment mark 17 of the reticle 8, is reflected by the half mirror 16, and is reflected back at the half mirror 16. objective lens 1
8 and enters the imaging device 19 .

In this way, the imaging device 19, as shown in FIG. 2(A),
The wafer 11
A video signal V displaying a pair of slit-like marks 12A and 12B forming an alignment mark 12 of
DIN is supplied to the control device 20.

Here, as the coordinate position on the wafer 11, if the position of the alignment mark 17 of the reticle 8 is at PLI and ptz as shown in FIG. 2(B), the coordinate pto of the intermediate position is expressed by the following formula. This position pto is defined as the position of the alignment mark 17.

Further, if the coordinate positions of the alignment mark 12 on the wafer 11 are pwt and PW2, the coordinates of an intermediate position pwo between them are expressed by the following equation, and this coordinate position pwo is defined as the position of the alignment mark 12 on the wafer 11.

Defined in this way, the position PLO of the alignment mark 17 on the reticle 8 and the position P of the alignment mark 12 on the wafer 11. If the movement of the stage 10 is controlled so that the following equation % (3) is satisfied, the wafer 11 on the stage 10 can be aligned with the reticle 8.

Here, among the images regarding the alignment marks incident on the imaging device 19, the alignment mark 1 of the wafer 11 is
If we pay attention to the video signal component VDIN of 2, as described above with reference to FIG. As the photoresist layer 23 covering the alignment mark 12 and the wafer 11 around it is exposed to the illumination light, it is exposed at predetermined time intervals as shown in FIGS. 4(B) to (E). Sampled video signal VD I
The value of the light amount at each coordinate position N1 to N4 indicates the optical property (refractive index) of the photoresist layer.

The imaging device 2
The distribution of the amount of incident light No. 19 with respect to the coordinate axis position changes with the passage of time. The rate of change depends on the thickness of the photoresist layer, the sensitivity, and the illumination light intensity.

However, if we examine in more detail the changes in the light intensity distribution over time, we find that the amount of light reflected from the edge portion 12B of the alignment mark 12 is reflected in a direction that does not return to the reticle 8 side. Therefore, it will always be a low value at any point in time. Incidentally, no matter how the properties of the photoresist layer 23 change, the reflected light from this portion remains incident on the surface 12A of the alignment mark 12.

Since the reflected light from the surface of the wafer 11 is originally under such conditions that it would return to the imaging device 19 side as reflected light if the photoresist layer 23 was not present, it is greatly affected by changes in the properties of the photoresist layer 23, and the reflected light is The amount of light also changes irregularly in practice.

Based on such a phenomenon, the control device 20 samples the light intensity value corresponding to the coordinate position included in the alignment detection video signal VDIN over time, and then converts it into digital data by the analog/digital conversion device 33. and is stored in the storage device 31. When the property of the photoresist layer 23 enters a stable zone where there is no change in its properties, the control bag is released! 20 reads data at each coordinate position from the storage device 31, and uses the arithmetic processing device 32 to execute a calculation for calculating an average value.

When such a calculation is performed, the value after the averaging calculation converges to a substantially constant value for a position where the amount of reflected light fluctuates irregularly among the coordinate positions. On the other hand, for positions where only a low amount of light is always obtained, such as the edge 12B of the alignment mark 12, there is no canceling component, so the average value becomes an extremely lower value than the constant value. .

Therefore, the control device 20 determines that the coordinate position where the average calculation value becomes extremely low is the edge surface portion 12B of the alignment mark 12.

Thus, the edge surface 1 with respect to the alignment mark 12
When the coordinate position of 2B is known, the control device 20
) center position P based on the formula. . is calculated and calculated based on the above equation (1) to align the alignment mark 1 of the reticle 8.
If there is a difference from the intermediate value pto obtained for 7, a drive signal is given to the drive control device 34 to drive the stage 10, thereby eliminating the deviation and moving the wafer in a direction that satisfies the above equation (3). Correct the position of 11.

In this way, the control bag W20 can automatically align the alignment mark 17 on the reticle 8 and the alignment mark 12 on the wafer 11.

According to the above configuration, the photoresist layer 23
The alignment mark 12 embedded in the photoresist layer 23 is not affected by interference fringes whose generation position changes over time in the reflected light due to exposure to illumination light irradiated from the photoresist layer 23. The position of the reticle 8 and the wafer 11 can thus be accurately detected.

Incidentally, when we observed changes in the position and generation state of interference fringes using a certain type of exposure equipment, we found that 2.
During ~3 seconds, the interference fringes occur at a maximum of 2 to 3 positions.
It was observed that the position moved by about [μm] and then the position change gradually slowed down.

Further, depending on the conditions of the photoresist, the width and contrast of the interference fringes may change. For this reason, conventionally, a preliminary exposure for about 1 to 3 seconds was performed until the characteristics of the photoresist were stabilized, and then the video signal was read. On the other hand, according to the configuration of the above-described embodiment, one frame of the video signal (or one frame of the video signal) is divided into 10 equal intervals during the preliminary exposure (1 to 3 seconds). Since the scanning lines of Normally, one frame of a video signal is 1/60 second, and unless the intensity of illumination light is extremely high, changes in interference fringes will not become stable within 1/60 second.

In the above description, the data sampled until the properties of the photoresist layer 23 are stabilized is averaged for each coordinate position, and the coordinate position where the lowest value is obtained is taken as the position of the edge surface portion 12B of the alignment mark 12. Although the case where it is determined that the edge surface portion 12B of the alignment mark 12 exists is described above, the method of determining the position of the edge surface portion 12B of the alignment mark 12 is not limited to this, and various other methods may be used.

For example, at each sampling time point, all the coordinate positions where the minimum values occur are read, and it is determined whether the coordinate positions of the minimum values are moving while the properties of the photoresist layer are changing, and Alternatively, the coordinate position where a minimum value that is not the same as that in the above-mentioned area occurs may be determined to be the position of the edge surface portion of the alignment mark 12. Even in this case, the same effect as in the above case can be obtained.

Further, the same effect can be obtained by providing a scanning slit and a photomultiplier in place of the imaging device 19, and obtaining an image signal by photoelectrically converting the light transmitted through the slit.

〔Effect of the invention〕

As described above, according to the present invention, even if the characteristics of the photoresist layer change during the exposure time, the position data of the alignment mark embedded in the photoresist layer can be easily determined without being substantially affected by the change. In this way, alignment can be achieved with high accuracy.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the exposure apparatus according to the present invention, FIG. 2 is a route diagram showing the relationship between alignment marks and alignment detection video signals, and FIG. 3 is an explanation of the alignment mark judgment operation. The curve diagram shown in FIG. 4 is a route map used to explain the problems of the conventional exposure apparatus. 1.11...Wafer, 2.12...Alignment mark, 8...Reticle, 9...
... Reduction projection lens, 15 ... Illumination system, 19
...imaging device, 20...control device, 3
1...Storage device, 32...Arithmetic processing unit, 33...Analog/digital conversion circuit, 3
4... Drive control device.

Claims (1)

[Scope of Claims] An exposure apparatus that attaches a photoresist layer onto a wafer and exposes an exposure pattern to the photoresist layer, wherein an alignment mark provided on the wafer is embedded in the photoresist layer. alignment mark detection means for irradiating illumination light and sending out the resulting reflected light; and detecting a surface portion of the alignment mark and an edge surface portion of the alignment mark based on the reflected light obtained from the alignment mark detection means. , an alignment mark signal forming means for obtaining an alignment mark signal including an electric signal component corresponding to the amount of reflected light reflected from the surface of the wafer and the surface of the photoresist layer, respectively; The alignment mark signals are sequentially detected at discrete points until the change in the characteristics of the photoresist layer during exposure becomes stable, and from the detected alignment mark signals, an electric current having an extreme value with little variation in the position of occurrence is detected. an exposure apparatus comprising determining means for determining that a target signal component corresponds to an edge surface portion of the alignment mark.