JPS62160722A - Exposure method - Google Patents

Abstract

PURPOSE:To realize alignment of high accuracy, by forming alignment marks on the back side of a wafer which is not affected by a process or a resist, and performing the alignment using the marks. CONSTITUTION:A reticle R is set on a reticle holder 1, and a wafer W on the back surface of which a resist layer is formed is set on a glass plate 5. After the alignment of the reticle R to an apparatus, a pattern on the reticle is exposed on the wafer W. Then marks Wtheta, WX and WY are formed on the back surface of the wafer W by means of development. The wafer W is adsorbed with vacuum on the glass plate 5 in the manner in which the back side faces a stage side. After performing a global alignment to the wafer W, the alignment between the wafer W and the apparatus is performed using the marks Wtheta1, WX1 and WY1. Further, light irradiation is performed on the back side of the wafer W through an objective optical system 17 with a light source 18 and a half mirror 19.

Description

Detailed Description of the Invention [Technical Field of the Invention] The present invention relates to an exposure method for printing a necessary pattern on an object such as a resist layer on a semiconductor wafer, and particularly relates to an exposure method for printing a necessary pattern on an object such as a resist layer on a semiconductor wafer. The present invention relates to an improvement in the alignment method between a pattern and a pattern.

[Background of the invention]

In conventional exposure methods, alignment marks are generally formed on the surface of a semiconductor wafer to be exposed, that is, the surface on which a resist layer is formed and is irradiated with light. Then, the semiconductor wafer and the reticle (or mask) are aligned using this alignment mark and an alignment mark formed on the reticle or mask.

However, in this alignment method, the alignment marks are observed through a resist layer coated on the wafer surface, which may make it difficult to observe them clearly, resulting in an increase in alignment errors. .

Particularly in recent years, as the degree of integration of integrated circuits has improved, patterns have tended to become increasingly finer, and there has been a desire to reduce such alignment errors.

[Purpose of the invention]

The present invention has been made in view of this point, and an object of the present invention is to provide an exposure method with high alignment accuracy.

[Summary of the Invention] According to the present invention, the alignment mark is formed on the back side of the substrate, and the alignment optical system is fixedly arranged on the back side of the substrate.

Alignment marks are formed on the back side of the substrate (by conventional phosphorography means). Next, the surface of the substrate is exposed to light in a predetermined pattern.

Alignment at this time is performed using alignment marks on the back surface of the substrate. Since no resist layer is formed on the back side of the substrate except in the case of initial mark formation, alignment marks can be detected well.

〔Example〕

Embodiments of the present invention will be described below with reference to the accompanying drawings. First, a first embodiment of the present invention will be described with reference to FIGS. 1 to 6.

FIG. 1 shows the overall configuration of the first embodiment. In this figure, a reticle R on which a pattern to be exposed is formed is held by a reticle holder 1, and this reticle holder 1 is supported by a column 2 at an appropriate position. Alignment mark sx on reticle R
, sy, and sθ are shown in FIG.

Next, a projection lens 4 is arranged below the column 2, and the wafer W is arranged at an optical position that is conjugate with the reticle R of the projection lens 4.

This wafer W is placed on a glass plate 5, and the glass plate 5 is further supported by a θ table 6. This θ table 6 has a rotation center 75a'i
It is configured to be able to minutely rotate about the z center, and this drive is performed by the θ table drive unit 7, and the rotation angle is θ.
It is adapted to be read by an angle reading encoder 8.

Next, the θ table 6 is placed on two tables 9 that can be moved slightly in two directions, that is, in the up and down directions.

The movement of the two tables 9 is carried out by a two-table driving section 11 under the guidance of a two-direction movement guide roller 10.

Next, the two tables 9 described above are placed on the Y stage 12 via two-direction movement guide rollers 10. This Y stage 12 is capable of rectilinear movement in a direction perpendicular to the plane of the drawing, and is driven by a Y stage drive section 13.

Next, the Y stage 12 is provided on the X stage 14. This X stage 14 is driven by an X stage drive unit 15 in the left and right directions in the figure, that is, in directions perpendicular to the Z and Y directions.
It is configured to be movable. This X stage 1
4 is placed on a surface plate or column plate 16.

Secondly, an objective optical system 17 is fixedly provided at approximately the center of the upper surface of this column space 16 between it and the glass plate 5 and facing toward the projection lens 4 . This objective optical system 17 is for detecting marks formed on the back surface of the wafer W. Furthermore, this objective optical system 17 detects the mark Sθ of the reticle R when the wafer W is not on the glass plate 5.

The projection lens 4 is installed so that the projected images of sx and sy can also be observed.
It is arranged coaxially with the optical axis AX of.

Inside the column space 16, a light source 18 such as a laser is provided on the side, and the light is configured to enter the objective optical system 17 through a half mirror 19. Further, an objective optical system 17 is provided on the bottom side of the column space 16.
and an alignment sensor section 20θ corresponding to the half mirror 19.

20X and 20Y are provided respectively. These alignment sensor units 20θ, 20X, and 20Y each photoelectrically detect marks formed on the back surface of the wafer W or marks Sθ, sx, and sy on the reticle, and align these marks with a predetermined detection center. This is for detecting deviation. Since these alignment optical systems are all fixed to a surface plate, that is, the column space 16, they are hardly affected by vibrations of the stage or the like. Therefore, alignment can be performed with high precision.

Next, a movable mirror 30 is provided on the side of the two stages 9 mentioned above, and a fixed mirror 61 is fixed to the lower part of the lens barrel of the projection lens 4. The movable mirror 60 includes a mirror 32
The light from the interferometer 64 is incident on the fixed mirror 31 via the beam splitter 35, and the light from the interferometer 34 is incident on the fixed mirror 31 via the beam splitter 63. That is, the light from the interferometer 64 including the laser light source is incident on the movable mirror 60 and the fixed mirror 61, respectively, and the wafer W is moved by the Y#X stage 12, 14 using the interference of the respective reflected lights. Coordinate values are now measured.

Next, the alignment sensor section 20θ mentioned above.

20X and 20Y are each connected to the alignment processing section 40. The alignment processing unit 40 determines the amount of position correction of the wafer W in the θ, X, and Y directions based on alignment signals from the alignment sensor units 20θ, 20X, and 20Y.

Next, the drive unit 3, the θ table drive unit 7, the angle reading encoder 8, the z table drive unit 11, the Y stage drive unit 13, the X stage drive unit 15, the interferometer 34, and the alignment processing unit 40 are all It is connected to the control device 50. This main controller 50 (a) controls the drive unit 3 during alignment of the reticle R; (b) controls the θ table drive unit 7, angle reading encoder 8, and Y stage drive unit 15 during global alignment of the wafer W; , X stage drive unit 15, and control by interferometer 64 (c) 2-table drive unit 1 during alignment for each exposure shot (so-called die-pie-die alignment)
1. It controls the control by the Y stage drive section 13, the X stage drive section 15, the interferometer 64, the alignment processing section 40, etc.

Next, the alignment optical system provided below the wafer W will be described in more detail with reference to FIG. In this example, for ease of understanding, half mirror 19 is
The alignment sensor section 20 is also shown by only one.

In FIG. 2, a pattern PP of a reticle R is projected onto the surface of a wafer W. A mark Wθ is etched on the lower side of the wafer W, that is, the back surface.

WY is formed. The glass plate 5 on which the wafer W is loaded is composed of a first plate on the upper side and a second plate on the lower side.
A gap 5c is formed between these plates. Furthermore, on the first plate,
A plurality of intake holes 5d are formed, and the intake holes 5d are connected to the gap 5c described above. This gap 5c
becomes a vacuum passage when vacuum suctioning the wafer W.

As shown in the figure, the spacing between the intake holes 5d is sufficient so that alignment will not be hindered by this, and the structure is designed so that the area where alignment cannot be achieved is minimized. In addition, this vacuum suction also allows the wafer W to be
Flattening is also performed.

Next, the objective optical system 17 is configured telecentrically on both sides of the optical axis AX, and the light outputted from the output side lens 18a of the light source 18 and reflected by the half mirror 19 is transmitted to the mark on the back surface of the wafer W. The light focused on the positions wy and wθ and reflected at this part is sent to the objective optical system 1.
7 and a half mirror 19, and are formed on a predetermined imaging plane of the alignment sensor section 20 as enlarged images WYA and WθA.

Next, an example of the arrangement of marks formed on the wafer W will be described with reference to FIG.

Figure 3 shows the arrangement of each mark 1W' as seen from the surface of the wafer W.
In this example, three exposures of the pattern of the reticle R, that is, three shots are shown. First, shot SI is placed on the front side (photoresist side) of the wafer W.
Next, the Y stage 12 (see FIG. 1) is stepped to expose the shot S2, and then the Y stage 12 is further stepped to expose the next shot S3.

Among the marks shown in FIG. 3, mark Wθl.

WX, , wyl are marks formed in advance on the back surface of the wafer W for the shot S. Similarly, marks Wa2, WXt, WY2 Hashi Elt82
ivy kt) (D mark υ, mark Wθ3...
* is a mark for shot S3. The same applies to other shots not shown.

Next, among the above marks, marks Wθ and WY are for alignment in the illustrated X direction, and mark WX is for alignment in the X direction. Further, the marks Wθ and WY are arranged symmetrically with respect to a line parallel to the y-axis passing through the center of the corresponding shot.

Next, FIG. 4 shows an alignment sensor section 20θ.

The correspondence relationship between each of the detection windows 20X and 20Y and each mark on the back surface of the wafer W described above is shown. This figure shows the first
In the drawing, alignment sensor sections 20θ, 20X, and 20Y are viewed from the front side of the wafer W.

In FIG. 4, the field of view of the objective optical system 17, that is, the image field IF, is set to cover a shot area S on the surface of the wafer W, as shown by a broken line. Alignment sensor section 20θ.

20X, 20Y detection windows APθ, APX, APY
are set in the image field IP corresponding to the marks Wθ, wx, wy, respectively, and the detection directions are indicated by arrows FA, FB, F'C, respectively. That is, the detection window p of the alignment sensor section 20θ,
pf) The detection direction in v'c is the X direction, the detection direction in the detection window APX of the alignment sensor section 20X is the X direction, and the detection direction in the detection window APYK of the alignment sensor section 20Y is the X direction. . Detection window A
PQ, APX, and APYK are provided with reference marks for determining 8 among them.

As mentioned above, the alignment sensor section 20θ.

20X and 20Y are fixed to the column base 16, and the detection windows APθ, APX, and APY are fixed to each sensor section, so as a result, the detection windows AP
θ.

APX and APY are at fixed positions relative to the main body of the device. The marks Wθ, wx, and wy are configured to be imaged simultaneously within the respective detection windows APθ, APX, and APY. Here, the meaning that a part of the alignment sensor is fixed means that it does not move during the step-and-repeat process of exposing at least one wafer. Therefore, when the shot size changes, the detection window may be moved appropriately to match the mark.

Therefore, the mark W is placed at the center of the detection window APθ in the direction of the arrow FA.
The wafer W is moved by the Y stage 12 so that θ is located and the mark WY is located at the center of the detection window APY in the direction of the arrow FC.
by moving the wafer in the X direction, or by moving the wafer w using the θ table
The shock to be exposed on the wafer W is determined by moving the wafer W in the X direction using the X stage 14 so that the mark WX is positioned at the center of the detection window APX in the direction of the arrow FB force. ) S can be made to coincide with the optical axis AX of the projection lens 4 (or the pattern center of the reticle R).

Next, a description will be given with reference to FIGS. 5 and 6 in addition to the above-mentioned FIGS. 1 to 4. Note that FIG. 5 shows how the light incident on the back surface of the wafer W from the alignment optical system is reflected by marks WX, etc., and FIG. 6 shows how the alignment sensor unit 20X detects marks. An example waveform of an image signal is shown.

First, as shown in FIG. 1, the reticle R is set in the reticle holder 1. This reticle R may be one for first printing having a circuit pattern for exposure, or may have only marks sy, sx, and sθ. In addition to being used for such a reticle R', a wafer W is made capable of forming a pattern by performing appropriate processing on both sides, and has a resin layer formed on one surface (this surface is referred to as the "back surface"). , orientation flat (
0'F) etc., and set it on the glass plate 5 with pre-alignment accuracy. The wafer W is fixed by vacuum suction as explained in FIG.

Next, after properly aligning the reticle R with the device, the pattern on the reticle R is exposed and printed onto the wafer W, and then it is developed to form the mark W as shown in FIG.
θ, wx, wy i are formed on the back surface of the wafer W.

These marks are preferably made by etching etc.
A recess is formed on the surface so that adsorption of the wafer to the glass plate 5 is not hindered.

Next, the wafer W is turned upside down so that the back side on which marks have been formed is on the stage side, that is, the alignment optical system side, and the orientation flat is set in the same direction as in the first baking process. 5 is vacuum-adsorbed. Therefore, the mark corresponding to mark SY on reticle R is Wθ, and the mark corresponding to mark Sθ on reticle R is WY. However, as described above, since the marks Wθ and WY are symmetrical, no particular disadvantage or inconvenience occurs. Furthermore, if a first-baked reticle is used as the reticle R, a circuit pattern will also be formed, but this does not cause any particular inconvenience in this embodiment.

Next, after performing appropriate Groogle alignment on the wafer W, first mark Wθ(+WXHr WYH(
(see FIG. 6), the wafer W and the apparatus are aligned. First, θ table 6, Z table 9,
The X stage 12 and the X stage 14 are driven by drive units 7.11.
13.15, based on the command from the main controller 50, so that the image field IF of the objective optical system 17 includes the shot area SR'1'rL as explained in FIG. 4, and each mark Wθ1 *WXI, WYI
are the detection windows APθ and APX, respectively.

Make it possible to request from APY.

On the other hand, a light source 18 and a half mirror are provided on the back surface of the wafer W.
19, light irradiation is performed via the objective optical system 17 as shown in FIG. In this example, since the mark WX is formed as a single linear concave shape, two parallel
Light scattering occurs at the two edges, reducing specularly reflected light. The specularly reflected light having a light amount distribution corresponding to the shape of the mark WX is transmitted to the objective optical system 17 and the nof mirror 19.
, detection window APX'a-, respectively, and enter the alignment sensor section 20X.

Next, as described above, since the reference mark is provided on the detection window APX, the specularly reflected light is also fully influenced by the light amount distribution. The specularly reflected light incident on the detection window APX is detected by scanning in the direction of arrow FB by an optical sensor (not shown). FIG. 6 shows an example of an image signal, where the horizontal axis represents the scanning position and the vertical axis represents the amount of reflected light. In Figure 6, bottom E, , g2
The center positions xa*xb are the centers of the reference marks formed in the detection window APX, and these positions xa,
The center position of xb is Xe. This center position xc is the detection center position, and can be considered to be fixed on the device. Next, bottom B1 * B2 is mark W
This is due to light scattering at the edges of
l.

The center position x3 of x2 becomes the position of the center line of mark WX. These center positions Xc* Since the amount of deviation of X3 is an alignment error, these center positions xc.

The X stage 14 is moved so that x3 matches. This movement is performed by the alignment processing section 40 determining the alignment error from the output of the alignment sensor section 20X, and based on this, the main controller 50 instructing the X stage driving section 15 to drive. This performs alignment in the X direction.

Similarly, alignment errors in the θ direction and the X direction are determined by the alignment processing unit 40 based on the outputs of the alignment sensor units 20θ and 20Y. Based on this, the main controller jJt50 controls the θ table drive unit 7,
A drive instruction is given to the stage drive unit 16, and the θ table 6 and the X stage 12 are driven to perform alignment in the θ and y directions.

Note that the second table 9 is used to move the wafer We up and down so that the mark on the wafer W (DW surface) comes to the focal point of the objective optical system 17.

As described above, shot S shown in FIG. 6! mark W
θ1. WXI + wyl is aligned to the device. In other words, the center of Sl is the optical axis AX.
This means that the wafer W has been aligned so as to match the .

In this state, the projection lens 4 exposes the pattern on the reticle.
The pattern of shot S1 is printed.

Hereinafter, in the same manner, marks Wθ2, WX2, W
The wafer W is aligned using Y2 so that the center of shot S2 matches the optical axis AX,
The printing of the pattern in S2 is performed. shot S3
．． The same applies to...

In this embodiment, if the reticle R used for first printing is used to form marks on the back surface of the wafer, the reticle R't'' can be attached as is to perform the first printing on the front surface of the wafer.

Furthermore, the alignment when attaching reticle R is marked S.
This can be done by detecting the projected images of θ, SX, and SY through the glass plate 5 with the alignment sensor sections 20θ, 20X, and 20Y. This problem tends to occur in systems where the reticle R and wafer W are aligned together using a common alignment sensor, so for example, in conventional systems where the alignment sensor for the reticle and the alignment sensor for the wafer W are separate systems. This means that there is basically no offset (or drift between systems), and highly accurate alignment is achieved. Of course, the projected images of the marks Sθ, sx, sy are the surface of the glass plate 5 (
The objective lens 17 also needs to be focused on the surface of the glass plate 5.

Next, a second embodiment of the present invention will be described with reference to FIGS. 7 and 8. Note that the same reference numerals are used for the same components as in the above-described embodiment. In this embodiment, in addition to the die-to-die alignment in the above embodiment, the entire alignment of the wafer W, that is, global alignment is also performed.

FIG. 7 is a view of the alignment optical system viewed from the glass plate 5 side. FIG. 8 is a diagram showing the arrangement of marks on the back side of the wafer W when viewed from the front side.

7 and 8, the optical system for global alignment is a Y microscope 60, a θ microscope 60, and an X microscope 60.
Microscope 62f! : each of which is arranged radially with respect to the optical axis of the projection lens 4. The detection center of the Y microscope 60 in the X direction is 60Y, the detection center of the θ microscope 61 in the X direction is 60θ, and the detection center of the X microscope 62 is 60Y.
It is 0X.

These microscopes 60, 61, 62 perform global alignment of the entire wafer in the X*7*θ direction.

Next, mark Wθ for alignment for each shot S
, wx, and wy are also arranged radially with respect to the center SC (see FIG. 8), and the detection windows APθ of each alignment sensor section 20θ, 20X, and 20Y.

APl and APY are also arranged correspondingly. Alignment for each shot is performed as described above using the marks Wθ, wx, and wy as described above.

Note that the laser interferometers 64 shown in FIG. 1 are provided for both X11 directions, but the intersection of the laser beam for detecting the position in the X direction and the laser beam for detecting the position in the X direction is the objective optical When the optical axis of the system 17 (or projection lens 4) is determined to pass through, the detection center of the microscope 60, 62 (5QY, 60X
'! It is best to make each distance equal. Also, the detection center 60
It is preferable to make Y and 60θ symmetrical with respect to the center SC of the objective optical system 17. As mentioned above, each mark is first formed on the back surface of the wafer W using an appropriate reticle, and then the wafer W is turned over and placed on the glass plate 5.
This is because a pattern of 14 beams of light is applied to the surface of the wafer W when the wafer is set on the top.

Also, as shown in FIG. 8, marks Wθ, WX.

If the marks Wθ and WY' of WY are arranged in the same manner as shown in FIG. 6, and the mark WX is provided outside the shot S, the arrangement of the marks Wθ, wx, wy and the marks SY, SX, Sθ on the reticle will be different. The arrangement of the projected image of wafer W
The front and back sides of the reticle are aligned, and the reticle can be aligned using the alignment sensor units 20θ, 20X, and 20Y even when the wafer W is not on the glass plate 5.

Next, a third embodiment of the present invention will be described with reference to FIG. In the embodiment described above, even if the size of each shot changes depending on the device to be created, it is devised to be able to respond well to this change.

First, a case where the shot is large will be explained.

Marks WθL and WYL are formed on the back side of the wafer W (WXL is omitted). This mark position is indicated by FP and coincides with the focal plane of the objective optical system 17.

A mirror 70 is provided on the output side of the reflected light of the objective optical system 17.
a and 70b are provided, whereby the optical axis is bent in the left-right direction. Mirrors 70a and 7Db
The reflected lights are reflected by the first objective lenses 71 m and 7 l, respectively.
b, aperture play through the second objective lenses 72a, 72b) 73a, 73b detection windows APY, AP
It is made to be incident on θ.

The detection windows APY and APθ of these aperture brakes) 73a and 73b are at positions conjugate with the position F'P. Furthermore, the first objective lenses 71a, 71b and the second objective lenses 72m, 72b are connected in an afocal system, and the aperture plates 73m, 73b are provided with the reference marks explained in FIG. (not shown).

As described above, alignment in the θ* The state shown in FIG. 9 is the state at the time of alignment of a large shot, and the image features θL and IYL of the marks WθL and WYL are observed.

When aligning a next small shot, images IθS and IYS of the marks Wθs and wys are observed.

At this time, the first objective lenses 71a, 71b and the second objective lenses 72a, 72b are moved together and fixed at appropriate positions.

It should be noted that the present invention is not limited to the above-described embodiments, but may be modified in various ways to achieve the same effect.

〔Effect of the invention〕

As explained above, by the exposure method according to the present invention, tl,
Since alignment marks are formed on the back side of the wafer, which is not affected by the Id process or resist, and are used to perform alignment, highly accurate alignment can be performed in a short time.

In particular, when applied to Ueno that incorporates multilayer resist, it is possible to obtain unprecedented alignment accuracy.

[Brief explanation of drawings]

FIG. 1 is an overall configuration diagram showing a first embodiment of the present invention, FIG. 2 is a configuration diagram showing an enlarged portion of the alignment optical system, FIG. 6 is an explanatory diagram showing an example of a mark configuration, and FIG. 4 5 is an explanatory diagram showing the relationship between the alignment optical system and marks, FIG. 5 is an explanatory diagram showing how illumination light is reflected by the wafer, FIG. 7 is an explanatory diagram showing a second embodiment of the present invention, FIG. 8 is an explanatory diagram showing an example of mark arrangement in the second embodiment, and FIG. 9 is an explanatory diagram showing a configuration of a third embodiment of the present invention. be. [Explanation of symbols of main parts] 4... Projection lens, 5... Glass plate, 17.
... Objective optical system, 18... Light source, 19... Half mirror 120θ, 20X, 20Y... Alignment sensor section, 40... Alignment processing section, R... Reticle, W... Wafer. Agent Patent Attorney Tadashi Sato Figure 4

Claims (3)

[Claims] (1) By moving a substrate on which a photosensitive layer is formed relative to a mask means having a pattern to be transferred, the pattern is sequentially transferred onto the photosensitive layer on the surface of the substrate by an exposure means. In an exposure method that performs multiple exposures, a first step of sequentially forming a large number of alignment marks arrayed line-symmetrically in the pattern on the back surface of the substrate by moving the substrate relative to the mask means; a second step of detecting the alignment mark by a mark detection means arranged on the back side; and a third step of aligning the substrate with the mark detection means based on the position of the alignment mark detected by the alignment means. An exposure method comprising: (2) The second step includes a step of vacuum suctioning the back side of the substrate by optically transparent fixing support means provided at the detection position of the mark detection means, flattening it, and fixing it. The exposure method according to scope 1. (3) The exposure method according to claim 1 or 2, wherein the second step includes a step of detecting a pattern image projected by the exposure means.