JPH03241821A - Thin film removing device - Google Patents

Abstract

PURPOSE:To manufacture the title thin film removing device capable of removing the local part of a thin film such as resist layer, etc., within a short time without deteriorating the washability of a substrate surface such as wafer, etc., by a method wherein a specimen is held on a turntable; the position of the local part of the thin film on the specimen is determined by a polar coordinate system; and then the position is irradiated with the energy beams conforming to the coordinate values. CONSTITUTION:The title thin film removing device receiving a specimen W wherein a thin film is evenly formed on the surface of a substrate whereon a specific pattern is formed and then removing a local part of the thin film on a previously specified position for the pattern shall be provided with a turntable 7 holding said specimen W to be turned on almost central part thereof, detectors 2, 3 to detect the specific parts of the pattern on the specimen W, a measuring means to determine the position of the local part to be removed on the thin film by the polar coordinate system making reference to the turning center 0 of the turntable 7 conforming to the detection results of the deterctors 2, 3 and the coordinate positions on the pattern design as well as a beam irradiating means 1 to irradiate the local part on said thin film with the energy bean for removing the thin film.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to an apparatus for improving the performance of a lithography apparatus (particularly an exposure apparatus) used in the manufacture of semiconductor integrated circuits, and in particular a device having a pattern on the surface, The present invention relates to an apparatus that locally removes a thin film from a substrate (wafer, etc.) whose entire surface is coated with a photosensitive thin film.

[Conventional technology]

In recent years, with the increase in the degree of integration of VLSI patterns, the minimum line width of circuits has also been reduced. Next-generation memory is expected to have a minimum line width of around 0.35 μm on a semiconductor chip.
For the production of 64Mbit D-R'AM, far ultraviolet (D
eep UV) light source, e.g. KrF with wavelength 248n-
It is expected that an excimer stepper using a (krypton fluoride) excimer laser as a light source will be mainly used.

In order to improve the alignment (especially wafer alignment) accuracy of this type of excimus timber,
Through-the-lens (T
It is desirable to perform alignment measurement (mark detection) using the TL method.

In the TTL method using such a reticle, illumination light is used to detect alignment marks on the wafer.
Good mark position measurement cannot be performed unless light of the same wavelength as the exposure light is used.

The reason for this is that the projection lens system used in the excimus beam has strong chromatic aberration, and the mutual benefit relationship between the reticle and the wafer cannot be maintained with non-exposure light, such as visible light.

For this reason, it has been considered to use excimer laser light branched from the same light source as the exposure light as illumination light for TTL alignment of the excimer beam. However, when detecting marks on a wafer using excimer laser light, the excimer laser light is often strongly absorbed by the resist layer coated with a uniform thickness on the wafer. In this type of resist, a layer coated to a thickness of 1 μm has a low transmittance of about 5 to 8%, and the amount of light that is reflected by the mark, passes through the resist layer, and returns to the projection lens system is approximately 1 μm in thickness. The transmittance can be expected to be about 0.64% at most.

With such a small transmittance, separate alignment measurement cannot be performed, and even if mark detection itself is possible, the position measurement accuracy and reproducibility will be extremely poor.

Therefore, as one countermeasure, it has been considered to remove the resist layer only at the alignment mark portion before the alignment operation.

FIG. 11 is a diagram showing a conventional method in which a resist layer at a mark portion is removed by exposure and development by irradiation with a light beam.

FIG. 11(A) shows a mercury lamp HG, a condenser lens CL, and a reticle R1 projection lens PL.

This figure shows how a resist removal device made up of a tube TB and the like is incorporated into a stainless steel bar made up of a stage ST for moving a wafer W in the X and Y directions in a step-and-repeat manner. On the surface of the wafer W, an alignment mark M is formed as shown in 111ffl (B), and furthermore, a film AL to be photo-etched and a resist layer Pr are deposited on the alignment mark M. The tube TB is movable in the vertical direction in the state shown in FIG. 11(A), and when removing the resist layer Pr, the tube TB is moved in the state shown in FIG.
), it descends so as to come into contact with the resist NPr via the backing PK. Now, when the resist layer Pr is positive type, the exposure light IL is irradiated from the lens GL to the mark M via the fiber FB.Then, a developer is supplied to the resist layer Pr through the inner tube IP,
This developer passes through the space between the inner tube IP and the outer tube OP and is discharged toward an absorption device (not shown). Subsequently, rinsing liquid is supplied from the inner tube IP, and the backing PK
Clean the developing area surrounded by . Further, N (nitrogen) gas is supplied from the inner tube IP to perform drying.

When the resist removal is completed in this way, the tubes TB etc. are raised from the wafer W, and the normal wafer alignment process is performed.
Entering the exposure process.

In addition, in order to omit the development process, a high-energy laser beam may be irradiated to the resist layer Pr near the mark M to evaporate (vaporize) the resist layer Pr. In that case, N and gas are removed from the inner tube IP. It is desirable to vent residual particles while feeding.

[Problems to be Solved by the Invention] In the above-mentioned conventional method, the part (tube TB, backing PK) from which the resist layer Pr is peeled off comes into contact with the wafer surface, or there is a high possibility that this will cause defects on the resist layer. There was a high risk of leaving scratches, etc., and it could not be put to practical use.

Furthermore, with the conventional method, it has not been possible to remove the resist at an economically viable speed while maintaining the cleanability of the wafer surface.

In conventional equipment, it is necessary to place the wafer on the XY stage ST in a pre-aligned state.
A clear alignment station based on the wafer outline and a transfer device (loader arm, etc.) to the xY stage ST are required, which inevitably slows down the processing speed. Part of the removal device (tube TB) for the local area (marked area) to be removed
Since the XY stage ST is used to position the
When the stage needs to be moved and positioned, and many parts are to be removed, a reduction in throughput is unavoidable.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide an apparatus that can remove only a local portion of a resist layer or the like in a short period of time without impairing the cleaning performance of the surface of a substrate such as a wafer.

[Means to solve problems]

In the present invention, a substrate such as a wafer is fixed on a rotating table, and a local portion of a thin film (resist layer) on the substrate to be removed (
The position of the alignment mark (alignment mark portion) is treated in a polar coordinate system as the rotary table rotates.Normally, alignment marks, etc. on a wafer are arranged according to a rectangular coordinate system, but the detection means of the present invention [alignment sensor] The mark position or other mark positions are converted into polar coordinate values based on the rotation angle of the rotary table when the mark etc. is detected and the distance from the rotation center of the alignment sensor.

Thereby, as long as the rotation angle of the rotary table is determined, the local area on the wafer is always positioned in the polar coordinate system with respect to the beam of the beam irradiation means.

This positioning is performed by two methods: controlling the angular position of the rotary table and controlling the relative position of the beam irradiation means and the rotary table in the radial direction. If the beam irradiation means has a light source such as a pulsed laser beam, extremely high-speed removal processing can be achieved by emitting pulsed light in synchronization with the rotation angle of the rotary table while it is being rotated.

Furthermore, by sharing the rotary table for removing localized portions of the resist layer with the spinner of a coater/developer that is a resist coating device, a developing device, or both integrated, it is possible to remove a dedicated resist layer during the photolithography process. There is no need to provide a partial removal device, and at the same time, time loss due to wafer transfer can be completely eliminated.

[Operation] In the present invention, a wafer is placed on a rotary table, the pattern position shift on the wafer is measured in a rotational coordinate centering on the rotation axis of the rotary table, and the resist is peeled off while the wafer is placed on the same table. Since the wafer is irradiated with light, there is no need to transfer the wafer, and processing can be performed in a short time. In addition, since the wafer is placed on a rotating table, even if the resist is blown off with pulsed light such as a laser beam and some resist droplets are produced, if the resist is removed while rotating, the centrifugal force will cause the droplets to fly away towards the outer circumference of the wafer and leave the resist on the wafer. Does not easily remain on the surface.

Furthermore, since a rotary table is used, it can also be used as a coater/developer spinner, and even if droplets remain, they can be cleaned with fluid.

When partially exposing a positive resist, there is no possibility of dust generation if development is performed on a rotary table.

〔Example〕

FIG. 1 is a plan view of a thin film removing apparatus according to a first embodiment of the present invention, and FIG. 2 is a front view thereof. The wafer W is the rotation axis 8
The wafer holder (rotary table) 7 is fixed by vacuum suction. 9 is a rotational drive unit such as a motor, 10
is an angle detection section of the rotating shaft 8. An alignment mark is marked on the wafer W together with a circuit pattern, and a photosensitive resist is applied thereon to a varying thickness. The rotating shaft 8 is held by a precision bearing, such as an air bearing, and the radial play during rotation is less than 1 μm, and the central axis of the rotating shaft 8 is made perpendicular to the adsorption surface of the wafer holder 7. There is. Further, the rotation angle detection section 10 uses a detector such as an optical rotary encoder or a magnetic rotary encoder, and can detect angles with resolution and accuracy of 1 second or less. Although not shown, a loader that automatically loads and unloads the wafer W onto the wafer holder 7;
In conjunction with this, a mechanism for facilitating the transfer of the wafer W is incorporated, but since these techniques are well known, they will not be specifically explained. Of course, it is desirable that a mechanism for centering the center of the wafer and the center of the rotating shaft 8 based on the outer shape of the wafer W be incorporated.

WY and Wθ are alignment detection sensors, which are installed at the tips of the Y sensor optical system 2 and the θ sensor optical system 3, respectively. Y sensor optical system 2 and θ sensor optical system 3
are connected by a connecting part 6 and can move linearly in the X direction along the rails 4 and 5. The position in the X direction is measured by a coordinate measuring device (not shown), such as a laser interference system or a linear encoder, and a motor enables positioning and constant speed movement. The Y sensor WY and the θ sensor Wθ are each movable in the X direction, and their moving positions are measured by a coordinate measuring device such as a linear encoder, and the positions can be automatically set by a motor.

1 is a light irradiation optical system, with a light irradiation part EM at the tip;
The light irradiation unit EM is movable approximately in the radial direction (X direction) about the rotation center 0 of the table 7 along a rail (not shown). The moving position can be measured using a precision coordinate measuring device such as a laser interferometer, and can be automatically set using a motor. Further, the light irradiation unit EM or the entire optical system 1 can also be slightly moved in the X direction.

The internal optical arrangement of the WY sensor optical system 2 and the Wθ sensor optical system 3 is the same, and this is shown in FIG. , 12 is a beam expander, 13 is a cylindrical lens, 14 is a relay lens, 15 is a beam splitter, 116 is a dichroic mirror, 17
is ξ, 18 is an objective lens, and the surface W○ of the wafer W is
An elongated laser spot is formed above.

This laser beam forms an alignment mark W in the form of a diffraction grating.
When incident on M, a diffracted light is generated, and the diffracted light is passed through the objective lens I8.
, Mirror] 7, Gikroic Stagnation-16, Beam Splitter-15, Relay Lens 19, Mirror 20,
The light enters the spatial filter 22 through the relay lens 21 . The spatial filter 22 selectively passes only the diffracted light,
The diffracted light is condensed by a condensing lens 23, enters a detector 24, and is photoelectrically converted. Such an alignment detection system is described in Proceedings of 5PrE Vol.
, 53 B (1985), pp. 9-16, and has already been used in the production of LSIs.

27 is an optical fiber 26 that guides illumination light for image observation, is a condenser lens, 25 is a beam splitter, 128 is a relay lens, and 29 is an image pickup device. The illumination light for image observation uses a wavelength range that is not sensitive to the resist. It is assumed that an image processing circuit is connected to the image sensor 29, which processes an image signal obtained when a mark on the wafer W is observed and automatically measures the mark position.

FIG. 4 is a diagram showing a specific configuration of the light irradiation optical system l. 31 is a laser light source, KrF, ArF, Xe
A pulsed laser beam with a large peak power such as an excimer laser such as C1, a YAG laser or its high frequency, a semiconductor laser or its high frequency, a copper vapor laser or its high frequency, etc. is output. The concave lens 32 and the convex lens 33 work to irradiate the variable aperture 34 with laser light. The lengths of the two sides of the variable aperture 34 can be set independently with respect to the center, and furthermore, the variable aperture 34 can be rotated about the aperture center by the rotating section 35.

This rotational angular position can be read by a rotary encoder or the like, and can be set to any desired angular position by a motor.

The variable frontage 34 is reduced and projected onto the surface WO on the wafer W by the action of the relay lens 36 and the objective lens 39. 47
is a beam splitter disposed between the relay lens 36 and the objective lens 39, and together with the condenser lens 40, the reflected light from the wafer surface WO of the laser beam is sent to the photoelectric detector 41.
It has the function of introducing the wafer surface W by laser light.
It is possible to monitor whether the resist on O has been removed. , 42, a beam splitter 137 is a dichroic mirror, and the illumination light that is insensitive to the resist is introduced through an optical fiber 44, passes through a condenser lens 43, a beam splitter 42, and is guided by a dichroic ink mirror 37, and is directed to the wafer surface WO. An enlarged image of the wafer surface WO is formed on the image sensor 46 by the objective lens 39 and the relay lens 45. The dichroic mirror 37 slightly reflects the wavelength range of the light emitted by the laser light source 31, and the position of the irradiation point can be observed using the imaging device 46.

Y sensor WY and θ sensor Wθ in Figures 1 and 2
To change the position of the optical system in the X direction, the entire optical system shown in FIG.
This requires afocality on the image side.

Also, when changing the position of the light illumination part EM of the light irradiation optical system l with respect to the rotation center O, moving only the objective lens 39 and mirror 38 in parallel requires only a small movable part.
In this case as well, the image side of the objective lens 39 must be afocal.

FIG. 5(A) is a block diagram of this embodiment, in which a rotation axis control section 5 that controls a rotation axis 8 that rotates a wafer holder 7 is shown.
2 is to turn on the vacuum chuck of wafer W from CPU 5 L.
10 F F command, a command regarding the rotation angle positioning angle and its execution, a command regarding the rotation speed and its execution in the case of constant speed rotation, a wafer W (Table 7 ) and commands for focusing, and also sends rotation position information and other information regarding the status of the rotation axis 8 and wafer holder 7 to the CPU 51.Y detection block 56 and θ detection block 57 are blocks corresponding to the WY sensor optical system 2 and the Wθ all-sensor system 3, respectively, and the x of the Y sensor WY and θ sensor Wθ is
It receives the information on the 1y set coordinate position and the movement command, and outputs the information on the x and y coordinates and the alignment mark measurement position actually measured by a laser interferometer or linear encoder.

Reference numeral 55 denotes a light irradiation block corresponding to the light irradiation optical system 1, which receives setting position information and positioning commands of the light irradiation unit EM from the CPU 51, information on the size of the variable aperture 34 and setting commands, and rotation angle and setting commands of the variable aperture 34. It receives a setting command, a laser output command to the laser 31, and an irradiation amount (energy amount) instruction signal, and also sends information regarding the state of the light irradiation block 55 to the CPU 51.

Reference numeral 53 denotes an external shape detection block for the wafer W, which measures the position of the outer periphery of the wafer W such as the lattice portion, the circumference, the non-cut portion, the circumferential position, the amount of eccentricity with respect to the rotation center 0, etc., and sends the information to the CPU 51 .

54 is a loader/unloader block for the wafer W;
The wafer W is loaded onto or unloaded from the cassette or truck to or from the wafer holder 7 according to a command from the CPU 51.

FIG. 5(B) is a circuit block diagram showing the blocks of FIG. 5(A) more specifically, and the same components as those shown in FIGS. 1 to 4 are designated by the same reference numerals. It's a dike. After the wafer W is almost centered on the rotary table 7 and adsorbed, it is necessary to align the orientation flat (or notch) of the wafer W in one direction while driving the motor 9. For this purpose, a photoelectric edge sensor 7
0 and a wafer outer shape measuring section 71 are provided. The wafer outer shape measuring section 71 includes an angle detecting section (hereinafter referred to as an encoder) 10
The central position of the flat (or notch) of the wafer is detected based on the angle pulse signal from the edge sensor 70 and the signal change from the edge sensor 70, and the information is sent to the CPU 51.

Further, the pulse signal from the encoder 10 is sent to the drive circuit 72 of the motor 9 and the rotation angle detection counter 73.

The drive circuit 72 uses the pulse signal for speed control, and the counter 73 digitally counts the pulse signal to measure the rotation angle of the table 7 with a resolution of 1 second or less. The count value DSθ of the counter 73 is sent to the drive circuit 72, CPU 51, and trigger synchronization circuit 87. The count value DSθ is from O″ to 360 in Table 7.
The drive circuit 72 uses the count value DSθ as position deviation information (current value input of the servo system) when rotating and positioning the table 7.

The trigger synchronization circuit 87 compares the angular position designated by the CPU 51 with the count value DSθ, and triggers the laser light source 31 for pulse oscillation when the table 7 reaches a predetermined rotational angular position.

On the other hand, a photoelectric signal from the alignment detector 24Y in the Y sensor optical system 2 (its objective lens is 18Y) is input to the Y mark detection circuit YAC. Detection circuit YA
Inside C, there is an amplifier 300, an analog-to-digital converter (
An ADC (ADC) 302, a memory 304, an address counter 306, etc. are provided, and the waveform of the photoelectric signal is stored in the memory 304 in correspondence with the position.

The same applies to the θ mark detection circuit θAC, and the signal waveform from the alignment detector 24θ in the θ sensor optical system 3 (its objective lens is 18θ) is stored in the internal memory. The address counter 306 in the detection surface IYAc (θAC) counts position pulse signals from the position sensor 84 that detects the positions of the Y sensor optical system 2 and the θ sensor optical system 3 in the X direction. The position sensor 84 also receives information X2 representing the position in the X direction of the alignment system WY at the tip of the Y sensor optical system 2, and information X representing the position in the X direction of the alignment system Wθ at the tip of the θ sensor optical system 3. is output to the CPU 51.

The position sensor 85 detects the position of the light irradiation part EM at the tip of the light irradiation optical system 1 in the X direction, and transmits the position information to the CPU 5.
Output to 1. The drive motor 86 moves the light irradiation part EM in the y direction in response to a command from the CPU 51.

Further, the variable aperture diaphragm 34 in the light irradiation optical system 1 is rotated by a drive unit (motor) 35, and its angle is detected by an encoder 89. The motor 35 is driven by an aperture diaphragm drive circuit 83 that reads angle information θa from an encoder 89. The drive circuit 83 servo-controls the motor 35 so that the set target angle commanded by the CPU 51 and the angle information from the encoder 89 match.

The image processing circuit 88 electrically processes image signals of the image sensor 29Y in the Y sensor optical system 2, the image sensor 29θ in the θ sensor optical system 3, and the image sensor 46 in the light irradiation optical system 1. , detects the one-dimensional or two-dimensional positional deviation of marks or specific linear patterns existing within the field of view of each objective lens 18Y, 18θ, 39, and converts the results to C.
Output to PU51.

The mark position detection circuit 80 processes the signal waveforms stored in each of the detection circuits YAC1θAC, determines the position of the mark on the wafer W in the y direction as a value on the orthogonal coordinate system, and converts the position information into a coordinate transformation calculation section. 82. Arithmetic unit 82
inputs information on the mark position on the wafer from which the resist should be removed from the mark position design data section 81, and sends the information to the detection circuit 80.
The mark position to be removed from the resist is converted into a polar coordinate format (diameter r and angle θ) based on the measured value of the mark position detected in .

These polar coordinate values are output to the CPU 51, parameters related to the diameter r are used to control the drive motor 86, and parameters related to the angle θ are used to control the aperture diaphragm drive circuit 83 and set command values to the trigger synchronization circuit 87. be exposed.

Next, a positioning method in the embodiment will be explained.

In this embodiment, the resist removal beam is positioned using a polar coordinate system 0-rθ centered on the rotation center O of the wafer holder 7.

It can be said that this is a new method, unlike the concept of positioning using an orthogonal coordinate system, which is generally used for arranging circuit patterns (chips) and marks on a wafer. Explaining the principle, the following equation may be used to convert the orthogonal coordinate system 0-αβ to the polar coordinate system 0-rθ.

(i) When 0<α, β θ−Tan−' ・・・・・・・・・・・・
(2) α (11) When α〈0, β θ−Tan + π ...... (3) α (iii) When α=0, θ=0 ...... (4
) To determine the polar coordinate system 0-rθ, it is necessary to determine the origin of the θ coordinate. FIG. 6 is an explanatory diagram showing how to determine the coordinate system on the wafer W. The α direction indicating θ=0 corresponds to each exposure area S printed in regular rows by the stepper.
Match the horizontal coordinate system of the arrangement of ij. Specifically, wafer global alignment marks GWAt,,G
W A t z 1G W A t z, G W A
It is determined by the direction in which t a are lined up.

FIG. 7 shows one exposure area (shot area) formed by a stepper and the arrangement of various alignment marks, and here the shot area S+i in FIG. 6 is shown. Each mark is provided in a street line area (width of about 100 μm) between shot areas, and the mark X M 1x in FIG. shall be taken as a thing.

Also, shot area S1 in FIG. In addition, marks YM of shot areas S8, sz4, Sl, and S43.

Assuming that XM is also detected by the TTL method, as shown in FIG. , the resist layer in that area is removed (
or exposure).

Furthermore, the α axis and β axis of the orthogonal coordinate system 0-2β shown in Figure 6
The axis is set parallel to the arrangement of shot areas, and the origin ◯ is the center of rotation of the rotary table 7. Therefore, the origin ◯ does not necessarily coincide with the six points on the wafer W, and there remains an eccentricity of about 1 m at the maximum. However, this depends on the centering accuracy when sucking the wafer W onto the table 7, and it is possible to keep it to about ±10 μm.

The above global alignment marks GWA,,,GW
B, some of them (at least two marks GWA,
, and one mark GWBl□j) are detected by the alignment system shown in FIG. 3, the detection circuits YAC, θAC, mark position detection circuit 80, etc. in FIG. 5(B). .

Next, the operation of this embodiment will be explained. In this embodiment, the alignment system shown in FIG. It is necessary to accurately know in advance the relative positional relationship with the light irradiation optical system 1 with reference to the rotation center O. So first,
The method will be explained below.

FIG. 9 is a flowchart showing the position setting procedure of the wafer alignment system, and shows the steps performed using the reference wafer RW of FIG. 8. A circle RC with a known radius and a straight line RL passing through the center of the wafer are marked on the reference wafer RW.

Furthermore, FIGS. 8(A) and 8(B) are shown assuming that the reference wafer RW is eccentric when it is attracted onto the rotary table 7, and the center of the wafer RW (circle RC) is set as Wc. In FIG. 8(A), when the point where the straight line connecting the center Wc and the rotation center O intersects the circle RC is P+, Pz,
The distance #R-ax from the rotation center ○ to the point P1 is the outermost eccentricity, and the distance Rwin from the center O to the point P2 is the innermost eccentricity.

Incidentally, FIG. 8(B) shows the state after the reference wafer RW is rotated 180' around 0 from the state of FIG. 8(A). Therefore, the eccentricity (eccentricity) of the circle RC on the reference wafer RW is Rma -Rwin.

In addition, the straight line RL on the reference wafer RW has a diffraction grating shape equivalent to the marks X M I 3, GWA, etc. in FIG. 7 so that it can be detected by the Y sensor optical system 2 or the θ sensor optical system 3. It is a good idea to leave it as

Now, in FIG. 9, the following setting operation is performed using such a reference wafer RW.

(Step 100) The reference wafer RW is placed on the wafer holder 7 by the action of the loader/unloader 54, and the rotation axis control section 52 vacuum-chucks the wafer RW onto the wafer holder 7.

(Step 101) The rotation axis block 52 rotates the wafer RW, and the outer shape detection block (70.
71) The position of the outer shape of the reference wafer RW is measured by the function of 53, and the pre-alignment of the wafer RW with respect to the rotation axis 8, that is, the flat OF is
almost parallel to the axis.

(Step 102) 1iRL is detected by the functions of Y sensor WY and θ sensor Wθ. At the time of detection, the rotation axis 8 is stopped at a position rotated strictly by 180 degrees, and the position of the line RL is determined by scanning the Y sensor WY and the θ sensor Wθ in the y direction. In the state of FIG. 8(A), the Y sensor WY,
The Y mark detection circuit YAC1 and the position detection circuit 80 operate, and the y-coordinate II value Ya when the beam spot of the Y sensor WY coincides with the straight line RL is sent to the CPU 51.

Next, the rotary shaft 8 is precisely rotated by 180 degrees via the motor drive circuit 72 according to a command from the CPU 51 to set it to the state shown in FIG. 8(B), and then the θ sensor Wθ,
The θ mark detection circuit θAC and the position detection circuit 80 are operated to determine the y-coordinate I1EYb when the beam spot of the θ sensor Wθ coincides with the straight line RL.

(Step 103) From the previously obtained y-coordinate values Ya and yb, the CPU 51 calculates (Ya+Yb)/2=Yset, and adjusts the Y sensor WY and θ so that the position measurement value of the position sensor 84 in the y direction becomes Yset. The detection center (laser spot) of the sensor Wθ is simultaneously moved in the X direction.

When this movement is completed, the line passing through the rain detection center of Y sensor WY and θ sensor Wθ will be
It will match the axis.

(Step 104) Next, the rotation axis 8 is precisely rotated by 90', and the image sensor 46 of the light irradiation optical system l is used to directly &*PL.
be observed.

(Step 105) The CPU 51 calculates the amount of deviation between the center of the light irradiation unit EM (the center of the image 34° of the variable aperture 34) and the straight line RL in the X direction (the tangential direction of the circle RC) using the image sensor 46 and the image processing circuit 88. The light irradiation unit EM (or the entire irradiation optical system I) is moved and positioned in the X direction so that this amount of deviation becomes zero.

(Step 106) Rotate the reference wafer RW at a constant speed.

(Step 107) Measure the circle RC with the Y sensor WY1θ sensor Wθ. In this case, it is detected that the intersection of the circle RC with the X axis (FIG. 8) shifts in the X direction due to eccentricity. The Y sensor WY detects the circle RC on the left side of the rotation center O on the X axis in FIG.

At this time, the Y sensor WY is the measured value X of the position sensor 84.
2, the point P1 on the circle RC is detected as the outermost position (distance from the center O) Rmax, and the point P2 is detected as the innermost position (
Distance from center O) Detected as Ra1n.

Since the θ sensor Wθ detects the circle RC on the right side of the center O on the X-axis, the outermost position and the innermost position are similarly detected based on the measured value X of the position sensor 84.

However, when the Y sensor WY detects the outermost position (point P+), the θ sensor Wθ detects the innermost position (point P2). Note that during this operation, the Y sensor WY and the θ sensor Wθ may be moved in the X-axis direction to search for the circle RC.

(Step 108) Next, the CPU 51 calculates the Y that should be set when the circle RC is not eccentric by calculating (Rwax + Rs + in) / 2 = Rref.
The positions of sensor WY and θ sensor Wθ in the X direction are determined. The calculated value Rref corresponds to the radius of the circle RC from the wafer center Wc.

Then, the CPU 51 calculates the measurement value X□ of the position sensor 84,
Y sensor W so that x is equal to the calculated value Rref
Position Y and the θ sensor Wθ in the X-axis direction.

At this time, the light irradiation part E set in the previous step 105
The center position of M is corrected by (Rmax-Rain) in the X direction, so that the center of the image 34'' projected by the light irradiation unit EM is located on the y-ring in FIG. After this setting is completed, movement of the light irradiation unit EM in the X direction is prohibited until the next device calibration.

(Step 109) Next, the CPU 51 controls the drive motor 86.
, the light irradiation unit EM (or the entire light irradiation optical system 1) is moved in the X direction by the position sensor 85, and the rotating circle RC is set to be observed by the image sensor 46, and step 107. Similarly to 108, the distance 11I of point P of circle RC
The position sensor 8 calculates Rmax and the distance @Rmin of point Pg.
5. Image processing circuit Measure based on the detection results on the 8th day.

(Step 110) Then, the CPU 51 similarly performs (step 110).
Rmax + Rmtn)/2 = Rrer is calculated, and the drive motor 8 is adjusted so that the measured value of the position sensor 85 (distance measured value from the center 0) matches the calculated value Rref.
6 to l1111N to preset the center position of the image 34' of the light irradiation part EM.

(Step 111) Stop the rotation of the reference wafer RW.

(Step E12) The reference wafer RW is unloaded by the loader/unloader 54.

By the above procedure, the alignment system (Y sensor WY, θ
Both the sensor Wθ) and the light irradiation part EM (image 34') are set.

Thereafter, when processing the actual wafer W, the reference position in the X direction of the alignment system [Y
set and the reference position (distance) in the X direction Rref, and the reference position (distance 1 [) R in the X direction of the processing center of the light irradiation part EM]
ref is used as the reference for everything.

In addition, in the previous step 107, Y sensor WY and θ sensor Wθ
When detecting a circle RC, the detectors 24Y, 24θ and mark detection circuit YAC shown in FIGS. 3 and 5(B) are used.
1θAC may be used, but in this case, the measurement direction by the beam spot and the direction in which the circle RC should be measured (X-axis direction)
must match. Therefore, if this is difficult due to the configuration of the alignment sensor, the circle RC may be detected using the imaging elements 29Y, 29θ and the image processing circuit 8.

Next, the procedure for actually peeling off the resist on the wafer W is shown in step 10.
This will be explained using the flowchart shown in the figure.

As shown in FIG. 6, the wafer W has global alignment marks GWA11 and GWBiJ in each shot area S.
Each shot area S ij
Every)[? -RiXM! J and y marks YM, , are provided. The positional relationship of these marks is determined in advance at the time of exposure with the stepper, and the CPU 51 receives and holds the positional information from the design data section 81. Also,
Global alignment marks GWAIJ and GWBi,
Based on the positional relationship, find the X positions of the Y sensor WY and θ sensor Wθ in advance so that G W A = j can be measured at the two locations at the same time, and set the beam bores of the sensors yw and wθ at that position. , after such preparation the 10th
Steps 150 to 162 in the flowchart shown in the figure are entered.

(Step 150) Loader/unloader 54 wafer W
The wafer is placed in a centered state on the wafer holder (rotary table) 7 and vacuum suctioned.

(Step 151) The wafer W is rotated by the rotating shaft block 52, and the wafer W is pre-aligned with respect to the rotating shaft 8 based on the wafer outer shape by the action of the outer shape detection block 53. That is, the rotating shaft 8 is positioned so that the flat OF (α axis of the shot arrangement) substantially coincides with the X axis.

(Step 152) While scanning the Y sensor WY1θ sensor Wθ in the y direction from the reference position Yset, each sensor makes a global alignment mark G W A
z + and each y coordinate value of G W A t 4 Yg0, Y
Detect gza.

(Step 153) The CPU 51 determines the rotation error Δθg of the wafer W from the difference between the respective y-coordinate values yg, ..., Ygt4.
is determined, and the drive motor 9 of the wafer holder 7 is slightly moved so that the error Δθg becomes zero.

As a result, the α-axis of the shot array coordinate system on the wafer W becomes parallel to the X-axis, so the CPU 51 resets the angle information DSθ of the rotation angle detection counter 73 to zero.

After correcting the rotation of the wafer holder 7, the Y sensor WY1θ sensor Wθ is scanned in the y direction again, and the mark detection circuit YAC
, θAC, etc., the global mark G W A !
Check the identity of the y-direction positions of I and GWA, , and determine the y-direction position when the mark GWAa is detected by the Y sensor WY and at the same time the mark GWA t a is detected by the θ sensor Wθ. Ywa is determined from the position sensor 84. The CPU 51 selects this position Ywa and the reference position Y.
The difference from set is calculated and stored as the eccentricity ΔβW in the β direction between the wafer center Wc and the rotation center 0 shown in FIG.

Incidentally, when detecting the global mark in the above steps, if necessary, the wafer holder 7 is moved up and down to perform focusing.

(Step 154) Next, the CPU 51 rotates the wafer W at 90°.
Rotate exactly and stop.

(Step 155) Change the X position (distance till from the center O) of the Y sensor WY so that the global alignment mark cwBzz can be detected.

(Step 156) Scan the Y sensor WY in the y direction to measure the position Yi+b of the global alignment mark GWB2□ in the y direction, and find the difference between it and the reference position Yset. α
Calculate the eccentricity ΔαW in the direction.

Through the above steps, the pattern position on the wafer W can be specified using the polar coordinate system 0-rθ.

(Step 157) The wafer W is rotated at a constant speed.

At this time, the counter 73 is set to 360 from the zero reset position.
°It shall return to zero each time it rotates.

(Step 158) The position of the light irradiation part EM is set to the position r where light should be irradiated, and the size (a, b) of the variable aperture 34 is set.
is set, and the rotation angle of the variable aperture 34 is set corresponding to the value of θ in the O-rθ coordinate system. These settings are made by the design data section 81, coordinate transformation calculation section 82, CPU 51,
This is performed by a drive circuit 83, a motor 86, etc.

(Step 159) Alignment mark XM.

and YM, , are irradiated with a laser beam via a synchronization circuit 87 at a position O-rθ below the light irradiation part EM at a predetermined amount of light or an amount of light suitable for peeling off the resist.

(Step 160) If there are any other irradiation points remaining, the process returns to step 158. (Step 161) If no irradiation points remain, the rotation of the wafer W is stopped, and (Step 162) the wafer W is unloaded.

Among the above operations, regarding steps 158 and 159,
This will be further explained in detail with reference to FIGS. 10(B) and 10(C).

FIG. 10(B) shows the arrangement of each coordinate system immediately after alignment is completed in step 156, and the orthogonal coordinate system α゛β″ whose origin is the wafer center Wc is the coordinate system αβ in FIG.
is set parallel to.

Furthermore, on the α゛ axis is the global mark G W A t +
, G W A t z, G W A z a, etc., and the global mark GW is placed on the β' axis.
It is set so that the GW 80, etc., are located. Fifth
The design data section 81 in Figure (B) shows this wafer center Wc.
Marks to be removed from the resist based on M, Y
The position coordinates of M ij are stored. For example, the mark XM, 3 is stored as coordinate values (>'H1+3, YXIII3), as shown in FIG. 10(B).

Upon completion of step 156, the coordinate system αβ is
It matches the coordinate system xy in Figure (B), and in that state, the eccentricity between the rotation center O and the wafer center Wc (ΔαW, ΔβW)
is required.

Further, the image 34° of the variable aperture stop 34 in the light irradiation optical system 1 is located at a distance HRref from the rotation center 0 on the y-axis.

Here, the coordinate exchange calculation unit 82 in FIG. 5(B) calculates the design coordinates of the mark
Input the amount of eccentricity (ΔαW, ΔβW), and use the previous formula (
Using 1), calculate the distance 111 r x M + s from the center of rotation O to the center point of marks Since it is located on the positive side of the α′-axis direction on “β”, the above equation (2) is applied, and the angle θ between the α′-axis (or α ring) and the line segment with radius rXMIs is
X M Is is calculated using the following formula.

(6) Next, the CPU 51 calculates the angle θXM,s from the wafer holder 7.
In this example, Fig. 10 (
As shown in B), when the processing point (image 34°) is located on the y-axis and the α'-axis of the array coordinate system is parallel to the X-axis, the measured value DSθ of the counter 73 is set to zero. Therefore, the rotation angle θXM, 3' of the wafer holder 7 until the mark XM+3 is located on the y-axis is determined, for example, by the following equation, depending on the image limit of the mark XM+s in the coordinate system αβ.

θXM+i'=π/2−θXM11 (7) In Figure 1 (C), the wafer holder 7 is at an angle of θXM1. ′
By moving the position of the image 34° by the light irradiation unit EM to the position of the distance NrX M + s, the center of the image 34° and the center of the marks matches.

However, if left as is, the shape of the image 34' is rotated with respect to the shape of the resist removal area, so the CPU 51 uses a drive circuit to rotate the variable aperture 34 counterclockwise by an angle θX M + β'. The command is output to 83.

As described above, the radial position of the image 34' is set, and the image 3
4" rotation direction and calculation of the angle θXM', the CPU 51 sends the trigger synchronization circuit 87 to the angle θXM'.
Output the value of +x'. The synchronous circuit 87 is the counter 7
The measured value DSθ of 3 is θXM.

Each time (every rotation), the processing laser light source 31
Sends a 1-pulse trigger signal to.

Normally, when removing a resist using an excimer laser or the like, it is difficult to cleanly remove the resist with a single pulse of high beam power, and multiple pulses are required. Therefore, each time the laser light source 31 oscillates, the CPU 51 detects the photoelectric detector 4 in FIG.
The output of step 1 is monitored, and the removal of the resist is confirmed in real time by the presence or absence of fluorescence generated by absorption of ultraviolet light by the resist.

Incidentally, the CPU 51 selects the mark X M to be removed from the resist,
The polar coordinate values (r, θ) of , YM, , are calculated in advance, and the machining order is determined, for example, from the mark with the radius r located at the outermost position to the mark located at the innermost position. In this way, the movement of the light irradiation part EM (image 34°) on the y-axis is
Since it is only necessary to move in one direction from the outside toward the center of rotation (or in the opposite direction), the throughput is higher than when the processing points are moved randomly.

In the above embodiment, Y sensor WY and θ sensor W
Although it is assumed that θ and the light irradiation part EM are different from each other, by changing the configuration of the optical system, the above three parts can be combined into one or two parts. Although only one light irradiation section EM is provided, a plurality of independently operating irradiation sections EM can be provided to improve throughput.

Further, the outer shape detection block 53 is replaced with a mechanical positioning section when positioning is performed based on the outer shape by mechanical contact.

In the above embodiment, the resist is removed while rotating, so that even if droplets are generated, they fly to the outside of the wafer W due to centrifugal force and do not remain on the surface of the wafer W.

As an auxiliary means for maintaining this cleanliness, it is also possible to provide means for flowing clean air, nitrogen, oxygen, ozone, or other gas from above the center of rotation O, and to flow the gas from the center toward the outer diameter.

To further improve the degree of cleaning, it may be possible to flush a fluid such as pure water from above the center of rotation.

In the case of positive resist, a developer can sometimes be used in place of pure water.

In the above embodiments, the resist was removed using a thermal or photochemical reaction using a pulsed laser beam.
In the case of a positive resist, the purpose can also be achieved using CW light (continuous light emission) that is sensitive to the resist, such as a mercury lamp light source or a He-Cd laser. In this case, it is better to use the wafer holder 7 in the embodiment as the rotary table of the coater/developer and stop the rotation of the wafer W at a calculated angular position during light irradiation. - is disclosed in, for example, Japanese Unexamined Patent Publication No. 1-179317. Resist and developer may scatter around the rotary table of the coater/developer and may contaminate the optical system. Therefore, a transparent flat plate for protection is provided between the alignment sensor and the light irradiation part, and if it becomes dirty, this plate should be installed. Only the flat plate needs to be cleaned. Cleaning a flat plate can also be automatically controlled remotely by providing a nozzle that dispenses cleaning liquid.

When the resist on the alignment mark is peeled off continuously with the coater operation, alignment measurement accuracy can be easily improved by performing the alignment operation using the wafer alignment mark before resist coating.

When removing resist or measuring mark positions while rotating the wafer W, in order to be able to observe the image, use a pulsed light source as the light source for image observation and synchronize with the rotation of the wafer W. All you have to do is make it emit light.

When a configuration is adopted in which a cleaning liquid or a developer can be flowed over the wafer W, the resist can be peeled off while the liquid is flowing, as long as the thickness of the layer of the liquid on the resist is within an allowable variation value.

When irradiating light while rotating the wafer W, the position of the wafer W may shift with respect to the holder 7 due to high-speed rotation. Therefore, after rotating the wafer W at high speed, reduce the speed to low speed or stop the rotation, and then adjust the alignment mark position again. It is better to acknowledge this at once, and for a substrate where the wafer W is stretched due to centrifugal force, it is better to be able to perform scaling correction for the amount of stretching.

After peeling off the resist, if necessary, it is possible to use an image observation system to confirm whether the resist has been peeled off at the correct position.

In the unlikely event that the resist is peeled off in areas other than the designated areas, with a negative resist, the resist can be removed from the entire surface of the wafer using a developer, and then the resist can be reapplied and the process can be repeated. If a coater/developer is used in common, these steps can be completed while being placed on the same spinner.

In the case of a positive resist, it is sufficient to irradiate the entire surface with photosensitive light and then peel it off using a developer.

〔Effect of the invention〕

As described above, according to the present invention, while the substrate such as a wafer is placed on the rotary table, pre-alignment and precision alignment are performed based on the external shape, and light is irradiated to the necessary areas on the same table. This has the effect of saving travel time, allowing alignment to be done in a short time, and shortening processing time. There is also the advantage that the device is smaller and the area required is smaller. Furthermore, when a local portion of the resist is removed while the rotary table is being rotated, droplets are blown away from the wafer by the centrifugal force of the rotation, so that the cleaning performance of the wafer is not impaired.

If the rotary table is used also as the rotary table of the coater/developer, there is an effect that the processing time for the series of steps of resist application and local resist removal can be shortened, and the surface can be easily cleaned with liquid. When a positive resist is used in this embodiment, the resist can be removed without the risk of splashing by local irradiation with photosensitive light and treatment with a developer. Furthermore, the present invention can be similarly utilized for local removal of thin films other than resist.

[Brief explanation of drawings]

FIG. 1 is a plan view showing the overall configuration of a thin film removal device according to an embodiment of the present invention, FIG. 2 is a front view of the device shown in FIG. 1, FIG. 3 is a diagram showing the configuration of an alignment sensor optical system, and FIG. The figure shows the configuration of the light irradiation optical system for processing, Figure 5 (A) shows the schematic configuration of the control system block of the thin film removal device, and Figure 5 (B
) is a circuit block diagram showing the detailed configuration of the control system, Fig. 6 is a plan view showing the pattern arrangement on the wafer to be processed, and Fig. 7 is a plan view showing the arrangement and relationship between the alignment mark and the resist removal section. 8(A) and 8(B) are plan views showing the pattern arrangement of the reference wafer, FIG. 9 is a flowchart of the initial setting operation using the reference wafer, and FIG.
) is a flowchart of the alignment operation of the wafer to be processed and the light irradiation operation for removing the resist, FIG. 10 (B)
, (C) are diagrams specifically showing the conversion from an orthogonal coordinate system to a polar coordinate system, and Figures 11 (A) and (B) are diagrams showing the configuration of a stepper that uses the conventional resist removal method. be. [Explanation of symbols of main parts] 1...Light irradiation optical system, 2...Y sensor optical system, 3...θ sensor optical system, 7...・Wafer holder (rotary table), 8...
... Rotating axis, 9 ... Motor, 10 ...
... Rotation angle detection section, 31 ... Laser light source for processing, 34 ... Variable aperture diaphragm, 51 ... CPU, 71 ... Wafer outer shape measurement section , 80... Mark position detection circuit, 81.
... Design data section, 82 ... Coordinate transformation calculation section, 83 ... Aperture stop F' 4 circuit, 84.85
...Position sensor 87...Trigger synchronization circuit, W...Wafer, RW...Reference wafer, Sij...Shot area, X M! j SY M i i...Resist removal mark, GWAij, GWB, □...Global alignment mark.

Claims (8)

[Claims] (1) In an apparatus that receives a sample having a thin film formed substantially uniformly on the surface of a substrate on which a predetermined pattern is formed, and removes a local portion of the thin film at a predetermined position with respect to the pattern, a rotary table that holds a sample and rotates it around the approximate center of the sample; a detection means that detects a specific portion of a pattern on the sample; a detection result of the detection means and a designed coordinate position of the pattern; measuring means for determining the position of the local portion to be removed on the thin film in a polar coordinate system based on the rotation center of the rotary table; based on the polar coordinate values determined by the measuring means; A thin film removal apparatus comprising: beam irradiation means for irradiating a removal energy beam onto a localized portion of the thin film. (2) The patterns on the substrate are arranged based on an orthogonal coordinate system, and the measuring means includes means for converting a position on the orthogonal coordinate system into a value on the polar coordinate system. The thin film removal device according to item 1. (3) the thin film removing device is provided integrally within a resist coating device for forming a photosensitive layer on the surface of the substrate;
2. The thin film removing apparatus according to claim 1, wherein the rotary table is also used as a spinner table of the resist coating apparatus. (4) The thin film removing device is provided integrally within a developing device for developing the photosensitive layer formed on the surface of the substrate, and the rotating table is also used as a spinner table of the developing device. The thin film removing apparatus according to claim 1. (5) The beam irradiation means is a synchronization means for irradiating the energy beam onto the substrate when the angular position of the rotary table during rotation reaches a predetermined relationship with respect to the polar coordinate value determined by the measurement means. The thin film removing device according to any one of claims 1 to 4, characterized by comprising: (6) Thin film removal according to any one of claims 1 to 4, wherein the beam irradiation means includes a laser light source that emits pulsed laser light as the energy beam for removal. Device. (7) The beam irradiation means includes a light source that emits a continuous beam as the removal energy beam, and a switch means for switching between irradiation and interruption of the beam to the substrate. The thin film removal device according to any one of Items 1 to 4. (8) Receive a substrate on which a predetermined circuit pattern area and alignment marks are formed on the surface and coat the surface with a photosensitive thin film having a substantially uniform thickness, and place a substrate corresponding to the position of the mark. The apparatus for removing a localized portion of the thin film includes: a rotary table that holds the substrate and rotates it approximately at the center of the substrate; a rotation angle detection means that detects information regarding the rotation angle of the rotary table; a mark detection means movably provided at an arbitrary position for detecting a mark on the substrate; irradiating an energy beam toward the substrate for removing a local portion of the thin film; a beam irradiation means whose irradiation position on the substrate is variable; measuring means for determining the position of a local portion to be removed on the thin film in a polar coordinate system based on the rotation center of the rotary table; and a beam irradiation position of the beam irradiation means based on the polar coordinates determined by the measuring means; in the radial direction of the rotation, and irradiates the beam when the determined angle value of the polar coordinates and the angle information from the rotation angle detection means reach a predetermined relationship. A thin film removing device comprising: a control means for controlling;