JP2908099B2 - Substrate alignment method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of aligning a substrate such as a wafer in a semiconductor exposure apparatus using pulsed light as illumination light.

[0002]

2. Description of the Related Art In the field of baking apparatuses for substrates such as semiconductor wafers, step-and-repeat type projection exposure apparatuses (hereinafter, referred to as "steppers") have become mainstream due to miniaturization of circuit patterns and enlargement of wafers. It is becoming. Various methods have been devised and implemented for positioning of a reticle and a substrate such as a wafer (hereinafter, referred to as “alignment”) in an exposure apparatus represented by a stepper.

Conventionally, in the case of the TTL system, an illumination light source for alignment (for example, a laser beam) is irradiated from above the reticle, and the reflected light from the reticle mark and the wafer mark is detected by a photoelectric detecting section. The relative position shift between the wafer and the reticle has been detected by obtaining the center of the mark pattern from the video signal.

[0004]

However, according to the above-mentioned conventional method, when the illumination light for alignment is a pulse light such as a pulse laser, speckles are generated in the above-mentioned video signal and accurate measurement is performed. Becomes impossible. The details are described below.

In the TTL system described above, generally, illumination light for alignment is obtained from the same light source as exposure light. A case will be described in which an illuminating light for alignment is obtained from an excimer laser illuminating system in this TTL type alignment in an excimer laser stepper using an excimer laser as an exposure light source.

[0006] If the laser beam has a high spatial coherence and is irradiated on a wafer or the like as it is, specifications and interference fringes will occur. For this purpose, the phase of the speckle / interference fringe is changed for each pulse by vibrating the beam or by a rotating diffuser, and the pulse is irradiated a plurality of times, thereby eliminating the influence of the speckle / interference fringe by an integration effect. .
In the case of wafer exposure, several tens to several hundreds of pulses are required to remove interference fringes and speckles.

However, in the case of alignment, the detection timing of the reflected light by the photoelectric detection unit and the time for capturing the reflected light pose a problem. For example, considering the case where an imaging means such as an NTSC-compliant CCD camera is used for the photoelectric detection unit, the exposure (accumulation) time on one screen is equal to the field accumulation time.
It is 1/60 second, and 1/30 second for frame accumulation.
Now, assuming that the pulse rate of the excimer laser is 2000 Hz, the number of pulses irradiated during this accumulation time is about 3 pulses in the case of field accumulation, and 6 to 6 in the case of frame accumulation.
The number of pulses is 7, and interference fringes and speckles cannot be sufficiently removed.

FIG. 5 shows an NTSC-compliant frame storage CCD.
9 shows a time chart of conventional detection when a camera is used. ST is a measurement start signal, and capture of a video signal is started from the next VD signal. VD is a vertical synchronizing signal, and LS is a laser output pulse which is irradiated so as to have a uniform light amount in each field. IM is C
It shows the contents of a memory that takes in a video signal output from a CD camera through a control unit. As can be seen from this figure, in the conventional method, a single image is illuminated with only a few pulses (six pulses in this time chart) of a laser output pulse. In this case, the high spatial coherency of the laser light is obtained. It is not possible to obtain an image without speckles and interference fringes caused by the above.

In order to solve this problem, one CCD camera is used.
There is a method in which alignment is performed by a long-time exposure operation that is performed with an exposure time of / 30 seconds or more.

FIG. 6 is a time chart when the CCD camera is operated in the long exposure mode in which the exposure time is 1/30 second or more. In ST, measurement start and stop are controlled. As a result, an image without speckles and interference fringes illuminated by several tens of laser output pulses can be obtained.

However, in this case, since long-time exposure is performed, the background noise increases due to the influence of the dark current of the CCD, and the resulting image has a poor S / N ratio. I can't get it.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems of the related art, and has been made in consideration of the problem of alignment marks having no speckles and uneven illuminance in positioning a substrate such as a wafer using pulsed light as irradiation light. It is an object of the present invention to provide a substrate positioning method capable of performing the positioning of the substrate with high accuracy by obtaining an image.

[0013]

In order to achieve the above object, the present invention provides a method of aligning a substrate, comprising the steps of: Detecting a displacement of the substrate by the method, and moving the substrate based on the detected displacement, wherein the image is displayed only when the imaging unit is opened.
It has an electronic shutter for generating a signal, an intermittent video signal generated by this <br/> and that opens and closes the electronic shutter in synchronization with the light emission of each pulse <br/> scan light by memory means By storing and adding a predetermined number of the video signals,
The position shift is detected by detecting a position of the alignment mark and comparing the position with a predetermined reference position.

An image signal of an addition reference mark which does not move relative to the image pickup means is added to the image signal,
Thus, the error at the time of addition may be corrected.

[0015]

According to the method of the present invention, only when each pulsed light is irradiated, the imaging means receives the reflected light of the alignment mark and generates a video signal, whereby the influence of noise such as dark current is generated. A predetermined number of video signals are stored in the memory circuit without receiving the signal, and an image of the alignment mark without speckle and uneven illuminance is obtained.

Therefore, the displacement of the substrate is detected with high accuracy.

[0017]

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

FIG. 1 is a block diagram showing one embodiment. Illumination light emitted from an illumination system 1 which is an illumination means using a pulse laser as a light source passes through a reticle 2 and a projection lens system 3 and a wafer stage. The wafer 5 which is the substrate on 4 is exposed.

The wafer stage 4 is reciprocated by a stage drive system 6 as a driving means in directions perpendicular to each other along two axes (X axis and Y axis) in a plane perpendicular to the plane of the drawing. One axis perpendicular to the plane (Z
Axis).

Reticle stage 2 for holding reticle 2
a is also reciprocally movable in the directions along each of the X-axis and the Y-axis by the stage drive system 6, and is rotatable around the Z-axis.

That is, the stage drive system 6 adjusts the relative positions of the reticle stage 2a and the wafer stage 4 by the main controller 12 described later.

An alignment detection system 7 for detecting a positional relationship between alignment marks, which are alignment marks provided on each of the reticle 2 and the wafer 4, functions to suppress the speckle by changing the wavefront of the illumination light with time. A part of the illumination light from the illumination system 1 provided as an illumination light for alignment is used, and the alignment mark of the reticle 2 and the alignment mark of the wafer 5 are detected from the reflected light of the mirror 8 by the CCD camera 9 as the image pickup means.

In an actual reduction projection exposure apparatus, a detection system similar to the alignment detection system 7 is provided at a position symmetrical with respect to the optical axis of the projection lens system 3, but is omitted in FIG.

The CCD camera 9 of the alignment detection system 7
Is input to the signal processing circuit 11 via the control unit 10. The electronic shutter of the CCD camera 9 opens only for a short time including the emission of each pulse light of the pulse laser of the illumination system 1 to generate a video signal.

The signal processing circuit 11 stores the obtained video signal in a memory circuit 11a, and the memory circuit 11a as a memory means adds a plurality of video signals to obtain an image of an alignment mark.

The signal processing circuit 11 detects the state of image formation on the light receiving surface of the CCD camera 9 of the superimposed image of the alignment mark of the wafer 5 and the alignment mark of the reticle 2 based on the image, and forms the image. A signal representing the amount of deviation between the wafer 5 and the reticle 2 according to the state is sent to the main controller 12.
Output to

The synchronizing signal generator 13 generates a synchronizing signal for the detection system and a synchronizing signal to the laser, and sends the synchronizing signal to the control unit 10, the signal processing circuit 11, the main controller 12, the laser controller 14 and the driving system 6. give. Note that the synchronization signal generator 13 is a signal processing circuit 11
Alternatively, it may be included in main controller 12.

The time during which the pulse laser emits light is several tens of nanoseconds, and the image obtained at other times becomes background noise such as dark current and stray light. Synchronize with the above, eliminate the accumulated charge except during the laser irradiation, and add the video signal obtained only during the laser irradiation on the memory, so that there is no speckle / illuminance unevenness and little background noise An image with a good S / N ratio can be obtained.

FIG. 2 is a time chart of the present embodiment.

SH is the ON timing of the electronic shutter, and LS is the laser output pulse. The laser is output while the electronic shutter is on. IM is from the CCD camera 9 to the control unit 1
This indicates the contents of the memory that has taken in the video signal output through 0, and takes in the memory circuit 11a while adding it to the memory itself at the time of taking in. At this time, normalization of data in the memory is performed at the same time so as not to overflow.

When the necessary number of pulses has been captured, measurement is performed from the obtained image, and based on the measurement result, the wafer stage 4 and / or the reticle stage 2a are moved to align the reticle 2 with the wafer 5.

The present invention can be implemented using a plurality of image memories.

In this case, the images of each field are sequentially taken into the memories 1, 2,... When the necessary number of pulses has been captured, the images in the memories are added, and measurement is performed from the image obtained by the addition. The reticle 2 and the wafer 5 are aligned based on the measurement result.

As a result, an image free from laser speckles and interference fringes can be obtained, and good measurement accuracy can be obtained without being affected by background noise due to dark current or the like.

In this embodiment, the imaging means is a CCD camera conforming to NTSC. However, cameras of other standards such as PAL may be used, and cameras operating at original timing instead of cameras conforming to these standards may be used. Absent. Further, a scanning position detecting element such as a one-dimensional line sensor may be used.

Next, a second embodiment will be described.

When image processing is performed using a CCD camera, a synchronization signal generated in the processing section is synchronized with the CCD camera side or the signal processing timing of the image processing section is synchronized with a synchronization signal of the CCD camera side. However, in this case, a pull-in means such as a PLL is used to achieve synchronization, and each video signal is shifted by a fraction of each pixel due to the jitter of the pull-in. Therefore, in the case of such a configuration, a high-accuracy image cannot be obtained due to the influence of the jitter simply by adding the video signals as described above.

For this purpose, an addition reference mark is put in the visual field of the video signal, and when adding, the measurement is performed not on the pixel basis but on this addition reference mark, and the measured value is used to reduce the jitter of each video signal. The addition is performed in consideration of the resulting shift.

FIG. 3 is a block diagram showing the present embodiment, in which the addition reference mark 15 is illuminated by the illumination light of the light source 16 and the reflected light is incident on the CCD camera 9. The light source 16 is a stable CW light source such as a DC-lit LED. It is desirable that the measuring section has a memory for each captured image. Except that the addition reference mark 15 and the light source 16 are used, the configuration is the same as that of the first embodiment.

FIG. 4 shows an image having an addition reference mark according to this embodiment. For a normal NTSC compliant camera,
Since the jitter in the horizontal direction is more important than the effect of the jitter in the vertical direction, the mark is provided in the vertical direction. Since the addition reference mark requires high-precision measurement, it is preferable that the addition reference mark is composed of a plurality of marks. Alternatively, the mark pitch may be made finer, and the position may be measured by moire with the CCD by approaching the CCD resolution limit.

[0041]

Since the present invention is configured as described above, the following effects can be obtained.

The displacement of the substrate can be detected with high accuracy from images without speckles and uneven illuminance, and the position can be aligned with the reference position with high accuracy. Accordingly, high-precision semiconductor exposure printing can be performed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an embodiment.

FIG. 2 is a time chart of the first embodiment.

FIG. 3 is a block diagram showing a second embodiment.

FIG. 4 is a diagram showing an image of an alignment mark according to a second embodiment.

FIG. 5 is a time chart of a conventional example.

FIG. 6 is a time chart of another conventional example.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Illumination system 2 Reticle 2a Reticle stage 3 Projection lens system 4 Wafer stage 5 Wafer 6 Stage drive system 7 Alignment detection system 8 Mirror 9 CCD camera 11 Signal processing circuit 11a Memory circuit 12 Main control device 13 Synchronous signal generator 14 Laser controller 15 Addition reference mark 16 Light source

Claims (2)

(57) [Claims] A step of detecting a position shift of the substrate based on a video signal of an imaging means for receiving reflected light from an alignment mark of the substrate irradiated by the pulse light; and detecting the position shift of the substrate based on the detected position shift. Moving the substrate, wherein the image signal is generated only when the imaging unit is opened.
It has an electronic shutter, before synchronously with light emission of each light pulse
It generates an intermittent video signal by opening and closing the serial electronic shutter, by adding to a predetermined number of storing the video signal by the memory means, to detect the position of said alignment mark is compared with a predetermined reference position Thereby detecting the positional deviation. 2. An image signal of an addition reference mark which does not move relative to an image pickup means is added to a video signal, and an error at the time of addition is corrected by the video signal of the addition reference mark. The method according to claim 1, wherein: