JPH03108316A - Wafer periphery exposing method - Google Patents

Abstract

PURPOSE:To improve precision by making the output part of a light guide fiber move rectilinearly from a retired position to the rotary shaft of a wafer or the center axis of circumference. CONSTITUTION:The photo current of a photodetector 62 which receives a light from a light emitting device 61 is compared with a reference voltage by a comparator 81 via an amplifier 81. The reference voltage of the comparator 82 is previously set as the value of a photo current generated when a photo sensor 6 is so arranged that the optical axis H comes into contact with the periphery 1e of a wafer 1. An output signal of the comparator 82 delivered to a position controlling mechanism 7 makes a sensor head 63 retaining collectively the light emitting device 61 and the photodetector 62 move rectilinearly toward the rotary shaft 1t, and controls the sensor so as to detect always the periphery 1e. Position data of the sensor 6 detected by a position detecting means 9 and the rotation angle of the wafer 1 obtained by a rotation angle reading mechanism 5 are stored in a storing means. Thus the deviation amount between the center of the circumference of the wafer 1 and the center of rotation is calculated from the moving amount of a periphery detecting sensor, and exposure is performed while a correction angle is taken into account, thereby obtaining highly precise exposure position.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a wafer peripheral exposure method for selectively exposing predetermined portions of a semiconductor wafer around the wafer to light for use in ICs, LSIs, and other electronic devices. It is something.

[Prior Art] In the manufacturing process of ICs, LSIs, etc., when forming fine patterns, a zero-order process is performed in which a resist is applied to the surface of a silicon wafer, etc., and then exposed and developed to form a resist pattern. Then, using this resist pattern as a mask, processing such as ion implantation, etching, and lift-off is performed.

Usually, resist is applied by a spin code method. In the spin code method, the wafer is rotated while pouring resist at the center of the wafer surface, and the resist is applied fi/11 to the entire surface of the wafer by centrifugal force.

However, the entire wafer surface is not used for processing; in other words, wafers coated with resist are transported and stored in various processing steps and in various ways, and the entire periphery or a part of the wafer is used for holding. Because of this, it is not possible to use the peripheral parts very much, and generally, in the peripheral parts, circuit patterns are drawn distorted and the yield is poor.

By the way, if the resist is a positive resist, even after the exposure process for pattern formation, the resist in the peripheral area that is not used for pattern formation will remain, and the peripheral area of the wafer may be mechanically gripped and held, or the peripheral area of the wafer may be held. The parts may rub against the walls of storage containers such as wafer cassettes, creating a source of "dust". In the case of positive resists, unnecessary resist remaining on the periphery of the wafer becomes "dust" and reduces yield, which is a serious problem, especially now that integrated circuits are becoming more sophisticated and smaller. It's summery.

Therefore, in order to remove the unnecessary resist around the wafer that remains after such development, it is necessary to perform a separate exposure to remove the unnecessary resist around the wafer in the development process, in addition to the exposure process for pattern formation. It is being done.

FIGS. 9A, 9B, and 9C are schematic illustrations of a conventional wafer peripheral exposure method for removing such unnecessary resist.

The wafer l is placed on the rotation stage 2 by a transport system (not shown).
? : set up, rotate the rotation stage 2, detect the orientation flat if of the wafer l by the sensor 72, stop the rotation when the orientation flat If comes to approximately a predetermined position, and then rotate the orientation flat plate 74a and the bin. Centering and positioning of the orientation flat If are performed by 74b, 74c, and 74d.

After the above positioning, while rotating the wafer I, light is irradiated from the output end 75 of the light guide fiber to expose the circumferential portion. In the exposure by this rotational movement (arrow 76a), the portion along the orientation flat if is not completely exposed because it is exposed as shown in FIG. 9 (8), and the orientation flat If is stopped at a predetermined position.
Output end 75 parallel to orientation flat If
moving arrow 76b), it is necessary to expose.

[Problems to be Solved by the Invention] The conventional wafer peripheral exposure method described above exposes a constant width circumferentially from the edge of the wafer.

However, when considering the efficiency of wafer utilization, it is desirable to use as much of the periphery as possible without causing a decrease in yield, or to hold only a few points where the holding claws come into contact and use the entire wafer area. There is also a request.

Therefore, a technique is required to selectively expose only the portions of the wafer periphery that are in contact with the holding claws.

Here, it is quite technically difficult to select a predetermined location on the wafer periphery that is designated as a location to be contacted by the wafer holding claws.

In other words, when exposing a predetermined location on the periphery of the wafer as described above while rotating the wafer, the output end of the light guide fiber is brought to the irradiation position when the wafer rotates at a certain angle, or the shutter is turned on. Possible methods include controlling opening and closing.

This "certain angle" is determined by a predetermined angle based on the design of the quest, based on some singular point. In this case, it is assumed that the center of rotation of the wafer and the center of the circumference coincide with each other, so it is necessary to perform an operation to align the center of one rotation with the center of the circumference.

Therefore, selective exposure of a predetermined portion on the periphery of the wafer requires a wafer centering operation, which involves a fairly complicated operation mechanism, resulting in a long processing time. Furthermore, there is a limit to the accuracy of the mechanical centering operation (5), and it is not possible to improve the accuracy of the exposure position.

This invention was made to solve the above-mentioned problems, and includes a wafer centering operation, 41! It doesn't matter,'
It is an object of the present invention to provide a method for selectively exposing a predetermined portion of a wafer periphery in which the A light position is highly accurate.

[Means for Solving the Problems] In order to achieve the above object, the wafer peripheral exposure method of the present invention detects the rotation angle of the one-rotation stage while rotating the wafer with the rotation stage, and the rotation angle is determined by the exposure angle. θ9
This is a wafer peripheral exposure method in which the output part of the light guide fiber is moved linearly toward the rotational axis of the wafer or the central axis of the wafer's circumference to expose a predetermined location on the wafer's periphery. The periphery is detected by the periphery detection sensor, and the position is controlled by moving the periphery detection sensor linearly toward the rotation axis so that the periphery detection sensor always detects the periphery. A singular point on the wafer's periphery is detected based on the position information of the peripheral edge detection sensor, and the amount of deviation between the center of rotation and the center of the wafer's circumference is calculated based on this singular point, and correction is determined based on the calculated amount of deviation. Exposure is performed by determining the exposure angle θ based on the angle α and a predetermined angle φ.

[Operation] By performing the above method, even if the center of the wafer and the center of rotation are misaligned, it is possible to accurately expose a predetermined location on the periphery of the wafer.

[Example] Next, the present invention will be explained in detail.

Each step of the wafer peripheral exposure method according to the embodiment of the present invention will be schematically explained below.

Step 2: While rotating the wafer once (360 degrees), the position of the sensor for detecting the periphery is controlled to follow the periphery of the wafer to detect the periphery.

Step (2): A singular point on the wafer's periphery is calculated from the position information of the periphery detection sensor whose position is being controlled.

Step m: Calculate the amount of deviation between the center of rotation and the center of the circumference based on the singular point, calculate the correction angle α from the calculated amount of deviation between the center of rotation and the center of the circumference, and calculate the correction angle α. An exposure angle 0a is determined based on α and a specified angle φ input in advance.

Step (2): The wafer is rotated again, the shutter is controlled according to the determined exposure angle θ2, and a predetermined portion on the periphery of the wafer is selectively exposed.

FIG. 1 is a schematic explanatory diagram of an apparatus for carrying out a wafer peripheral exposure method according to an embodiment of the present invention. In Fig. 1, l is the wafer, ip is the wafer periphery, le is the wafer periphery, 2 is the rotation stage, 3 is the main controller, 4 is the stage drive mechanism, 5 is the rotation angle reading mechanism, and 6 is the sensor for detecting the periphery. 7 is a position control mechanism, 8 is a position control mechanism drive circuit, 9 is a position detection means for a peripheral edge detection sensor, IO is a sub-controller, 11 is a storage means, 12 is a calculation means, 13 is a shutter drive mechanism, 14
15 is a light guide fiber, 15 is an output section, and E is an exposure light source section.

First, step (2) will be explained.

FIG. 2 is an explanatory diagram of position control of the photosensor 6 in the apparatus of FIG. 1. In FIG. 1, the photosensor 6
is controlled so as to always detect the peripheral edge 1e of the rotating wafer. That is, as shown in FIG. 2, the photocurrent of the light receiver 62 that receives light from the light emitter 61 is amplified by an amplifier 81 and compared by a comparator 82. The reference voltage of the comparator 82 is
It is preset as the value of the photocurrent generated when the photosensor 6 is placed so that the optical axis H shown in FIG. 2 is in contact with the peripheral edge 1e. The output signal of the comparator 82 is sent to the position control mechanism 7, and the output signal between the positions v44i! Based on the signal from the comparator 82, the system 7 moves the sensor head 63 that holds the first light emitter 61 and the light receiver 62 together in a straight line toward the rotation axis It, and controls it so that it always detects the wafer peripheral edge le. do. Specifically, a servo motor is used as this position control mechanism 7.

The position of the photosensor 6 whose position is controlled by the position control mechanism 7 is detected by the position detection means 9.

As this position detection means 9, for example, a pulse counter that counts the pulses of the servo motor used as the position control mechanism 7 is used. The number of pulses counted by the pulse counter at the corner can be used as information on the position of the photosensor 6.

Information on the position of the photosensor 6 detected by the position detection means 9 and the rotation angle of the wafer I read by the rotation angle reading mechanism 5 are sequentially sent to the storage means 11 and stored therein. In this way, the wafer 1 is rotated once (360 degrees), and data on the position of the photosensor 6 at each rotation angle is stored in the storage means 11. As the storage means 11, an IC memory is used.

Next, step (2) will be explained.

In Figure 3, l is the wafer, OT is the center of rotation of the wafer, and 1f
indicates an orientation flat, and Or indicates the origin of the orientation flat as an example of a singular point.

The singular point is a reference point necessary for specifying a designated angle, determining an exposure angle, etc. Conventionally, an orientation flat If indicating the directionality of the crystal structure has been provided on the wafer l, and in this embodiment, this orientation flat If is detected and used as a reference point. However, since the orientation flat If has a length AB, select the intersection point 0 between the rotation center OT and the perpendicular drawn to the orientation flat 1f and the orientation flat 1f as the singular point, and set the point OF to the origin of the orientation flat, that is, Let it be a singularity. In addition, wafers that have small notches on their periphery and are used for alignment, etc. are also being used recently, and in this case, this notch can be used as a singular point. Good.

Figure 4 is an explanatory diagram of the relationship between the position of the photosensor and the rotation angle stored in the 1-position detection means, of which (a) shows the actual measured value and (b) shows the differential value. It is.

The information on the position of the photosensor sent from the position detection means is the number of pulses of a servo motor used as a position control mechanism. This number of pulses n is the actual value of the information on the position of the photosensor, and is taken as the vertical axis in FIG. 4(a). Further, d n /dθ, which is obtained by differentiating the number of pulses n with respect to the rotation angle θ, corresponds to the amount of displacement of the photosensor, and this is taken as the vertical axis in FIG. 4 (b).

Moreover, the horizontal axis is the rotation angle θ. Note that one rotation angle θ is obtained by inputting the data of the rotational angular velocity (ω) of the rotation stage 2 into the storage means 11 in advance without using the data from the rotation angle reading mechanism 5 shown in FIG. , dθ=ωdt may be calculated by the calculation means 12.

When the wafer starts to rotate and the photosensor moves to follow the change in the position of the periphery, the value of d n /dθ changes as shown by the dotted line in FIG. 4 (b).

Here, the region F where d n /dθ largely changes to the + side and the − side is considered to be the angular range in which the photosensor detects the orientation flat portion. Furthermore, θ0, which is dn/dθ=0 in the region F, is the displacement amount of the photosensor movement when detecting the orientation flat, so the photosensor 6
In this state, the optical axis H of the photosensor 6 is in contact with the periphery of the wafer at the origin Or, as shown in FIG. By determining °, it becomes possible to know the position of the origin 02. Therefore, in the calculating means 12 of FIG. 1, the rotation angle θ is such that d n /dθ=0 within the region F. Perform the calculation to find.

Next, step (2) will be explained.

FIGS. 5 and 6 are explanatory diagrams of calculation for determining the amount of deviation between the center of rotation and the center of the circumference from information on the position of the photosensor.

As shown in FIG. 5, from the rotation center OT to the origin 0 of the orientation flat. Based on the line segment drawn at θ1.
θ2. θ3. θ4. θ5. The points on the periphery of the wafer when rotated by an angle of θ6 are A, A, and A, respectively, as shown in FIG.
Let C, B, A', C', B'.

Specifically, θ1...45@02...90"θ3...135"04...225°08...270" 06...Conflict 315°, θ4=θ, $180', θ5=02+t
aO9θ6=θ, +180” and θ2=0.1)90
°, 01=01+90°, then θ8. θ1.
The angles of θ4 and 06 are not limited to the above angles. However, it is necessary to ensure that points A and B' do not lie on the orientation flat.

In FIGS. 5 and 6, for example, the rotation center 07
Consider a line segment AA' that is perpendicular to the line segment BB' that passes through the center Oc of the circumference. Let n(θ, ) be the position of the photosensor when it detects point A, and n(θ4) be the position of the photosensor when it detects point A', then n(θ4) - n( θ+)=A Or −A'
Or intersection P is the midpoint of line segment A A l', so AP=A
'If P=view, the amount of deviation is 0 aP.

0aP=AOT-see OT P = standing - A’ OT Therefore, 0a P= Suburb/AO Ding -A" OT = Repair・(n(θ,)-n(θ4)) becomes.

Therefore, information n(
By knowing θ, ), n (θ4), the OTP distance can be determined. Also, n(θ,)−n(θ
By checking the sign of I), the direction of the deviation can also be known. n(θI).

The value of n(θ4) is the first
Since it is in the data stored in the storage means 11 shown in the figure, the above θ1. The arithmetic means 12 may perform the above calculation by specifying θ4 and sampling.

Similarly, the distance and direction of OyQ shown in FIG.
From the data of θ3) and n(θ6), the distance and direction of the OTR are determined from the data of n(92) and n(Os).

FIG. 7 is an explanatory diagram of the correction angle α. lpD indicates a predetermined location on the wafer periphery, φ indicates a designated angle, and α indicates a correction angle.

The angle that can be read by the rotation angle reading mechanism 5 shown in FIG. 1 is the rotation angle of the rotation stage 2, that is, the rotation angle with respect to the rotation center OT. Therefore, for exposure in step (2) to be described later, it is necessary to specify the exposure angle 0a based on a line segment connecting the center of one rotation 0a and some singular point on the periphery.

Therefore, as shown in FIG. 7, first, the specified angle φ is specified in advance based on the line segment connecting the origin OF of the orientation flat as a singular point and the rotation center OT. This specified angle φ is specified according to the position of a predetermined portion on the wafer periphery where a wafer holder is held in a subsequent processing step, as shown in FIG.

It is input to the main controller in advance. Therefore, if two exposure angles 0a are expressed by the specified angle φ and the correction angle α, it becomes θt;φ+α.

The correction angle α is obtained from the data in step ① by the calculation shown below.

FIG. 8 is a diagram for explaining the calculation for determining the correction angle α. 0TP=a, OaQ=b, O,R=c in Figure 6
shall be.

In Figure 8, β=tan-'(c/y) y=Fiη stone T寥7・°・β= tan-"(c/"i-11]ko)
・,,,, (1) Also, x cosa −d cos(180”−(β+φ))
=r,”, x cos α+ d cos(β+φ)=
r -------(2)d 5in(180°-(β
+φ))=x sinαdsin(β+φ)=xsin
a:, x = d 5in (β + φ) /sin
a ” and (3) Substitute (3) into (2), (dsin (β + φ) / sin α) fist cosa + d
cos(β+φ)=r(dsin(β+φ)/lan a)+dcos(β+
φ)-r anα-5in(β + φ)b(r/
d)-cos(β+φ))・Fist α8 Lan-' (sin(β+φ)ba(r/d)-
cos(β+φ)))...(4) Therefore, the arithmetic means built in the main controller 3 in FIG.
After receiving the data, the calculation of equation (1) is performed, and then the calculation of equation (4) is performed to calculate the correction angle α.

Then, from the input φ and the calculated α, θ = φ + α
is calculated to obtain the exposure angle θ2.

Next, step (2) will be explained.

In FIG. 1, the exposure light source section E includes a light source lamp E1 such as a short-arc type mercury lamp, an elliptical condensing mirror E2 and a plane mirror E3 that converges the light from the light source lamp El into the light guide fiber 14, a plane mirror E3, and a shutter. Consists of E4 etc. Light from the exposure light source section E is guided by a light guide fiber 14, and is emitted from an output section 15 having a projection lens inside. Note that FIG. 1 shows the case where the emission section 15 is in the retracted position.

The emission section 15 is held by an emission section holder 16. The emission part holding base 16 is attached to the sensor head 63k,
It is movable together with the photosensor 6, and is also movable linearly toward the rotation axis It of the wafer on the sensor head 63 by the holder moving mechanism 17.

In FIG. 1, as the rotation stage 2 starts rotating again, the rotation angle is read by the rotation angle reading mechanism 5 and a monitor signal of one rotation angle is sent to the main controller 3. When the main controller 3 receives a signal indicating that one rotation angle has reached the exposure angle θ2, it sends a signal to the holding table moving mechanism 17, and directs the emission part 15 toward the rotation axis It by a predetermined distance at a predetermined speed. After moving in a straight line, return to the retreat position. As a result, a predetermined portion around the wafer is exposed. The speed and distance of movement of the emission unit 15 are inputted in advance to the main controller 3 according to the rotational speed of the rotary stage 2 and the shape of a predetermined area around the wafer to be exposed.
It is sent to the holding table moving mechanism 17 as an Mm signal.

Note that the retracted position of the emission section 15 at the initial stage of rotation is not necessarily on the straight line connecting the origin 02 of the orientation flat, which is the reference point, and the rotation center.

Therefore, the position of the origin 02 at the end of one rotation (360 degrees) in step ■ can be determined by calculating the singular point in step H, so specifically, the exposure angle θ2
and the above θ. When the rotation angle becomes 0° + 02,
The holding table moving mechanism 17 is driven.

Although the wafer l continues to rotate while the emission section 15 is moving, the main controller 3 may send a signal to the stage drive mechanism to stop the rotation.

It may be preferable to move the emission section 15 from the retracted position toward the central axis of the circumference in terms of accuracy of the exposure position, but it may be complicated from a mechanical point of view. If the amount of deviation between the rotation center and the center of the circumference is not large, movement toward the rotation axis It may be sufficient as in this embodiment.

As mentioned above, since the emission part 15 is movable together with the photosensor 6, the emission part 15 is fixed at a predetermined position on the sensor head 63 during rotation in step I, and the emission part 15 is moved together with the photosensor 6.
If the peripheral edge 1e is controlled integrally with the peripheral edge 1e,
A band-like exposure of the wafer peripheral portion at a predetermined distance from e can also be performed at the same time.

[Effects of the Invention] As explained above, according to the present invention, the amount of deviation between the center of the wafer's circumference and the center of rotation is calculated and determined from the amount of movement of the peripheral edge detection sensor, and the correction angle α is taken into account. Since exposure is performed, there is no need for a wafer centering mechanism or operation, and the wafer periphery exposure method provides high exposure position accuracy.

[Brief explanation of drawings]

FIG. 1 is a schematic explanatory diagram of an apparatus for carrying out a wafer peripheral exposure method according to an embodiment of the present invention. FIG. 2 is an explanatory diagram of position control of the photosensor in the apparatus of FIG. 1. FIG. 3 is a diagram for explaining the singularity in the embodiment of the present invention. FIG. 4 is an explanatory diagram of the relationship between the position of the photosensor and the rotation angle stored in the position detection means, of which (a) shows the actual measured values and (b) shows the differential values. It is. FIGS. 5 and 6 are explanatory diagrams of calculations for determining the amount of deviation between the center of rotation and the center of the circumference from data on the amount of movement of the photosensor. FIG. 7 is an explanatory diagram of the correction angle α. FIG. 8 is a diagram for explaining the calculation for determining the correction angle α. FIGS. 9A, 9B, and 9C are schematic illustrations of a conventional wafer peripheral exposure method. In the figure. l: Semiconductor wafer 2: Rotation stage 3: Main controller 5: Rotation angle reading mechanism 6: Photo sensor 7 as a peripheral edge detection sensor: Position control mechanism 9: Position detection means 14 of the peripheral edge detection sensor 14: Light guide fiber 15: Output portion lpD: Predetermined location on the wafer periphery le: Periphery lt: Rotation axis OF: Orientation as a singular point Origin of flat OC: Center of circumference OT = Center of rotation

Claims (1)

[Claims] While the wafer is being rotated by the rotary stage, the rotation angle of the rotary stage is detected, and when the rotation angle reaches the exposure angle θ_E, the output part of the light guide fiber is moved from the retracted position to the rotation axis of the wafer or This is a wafer peripheral exposure method in which the wafer is moved linearly toward the center axis of the circumference and a predetermined location on the wafer periphery is exposed. The position of the peripheral edge detection sensor is controlled by moving the peripheral edge detection sensor linearly toward the rotation axis so as to detect the peripheral edge, and the singular point on the wafer's peripheral edge is detected using the position information of the peripheral edge detection sensor that is performing this position control. Detect the singular point, calculate the amount of deviation between the center of rotation and the center of the circumference of the wafer using this singular point as a reference, and determine the exposure angle θ_E based on the correction angle α found by the calculated amount of deviation and the specified angle φ determined in advance. A method for exposing the periphery of a wafer.