JPS627126A - Charged beam exposure device - Google Patents

Abstract

PURPOSE:To allow the control of the exposure position of the charged beam with a simple configuration and higher accuracy by a method wherein an alignment mark is detected with a laser beam, and the shifting information for the stage and the positioning information for the charged beam are obtained on the basis of the detected results. CONSTITUTION:With laser beams LAX and LAY for detecting an alignment mark, a stage 1 is shifted by a stage position control device 32 to a place where it is possible to scan X-axis and Y-axis alignment marks 4Xb and 4Yb accompanied with a chip 3b adjacent to a chip 3a to be drawn first. Alignment detection signals S2X and S2Y generated with scanning are received by the exposure control device 31, the stage position control device 32 stops the stage 1, receiving the output signal, and the position is accurately detected by a stage position reading device 33. When based upon the stored information, the exposure control device 31 calculates the position of the chip 3a for starting the exposure and feeds the result to the charged beam position control device 7, the control device 7 deflectively controls the charged beam 5 through an electron lens 6.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Field of Application] The present invention relates to a charged beam exposure apparatus, and is particularly suitable for application to semiconductor manufacturing equipment.

[Conventional technology]

Conventionally, in semiconductor manufacturing equipment, a charged beam exposure device that uses a charged beam (e.g., an electron beam) and an exposure bag (e.g., a stepper) that uses light have been put into practical use as means for forming fine patterns on the surface of a wafer. There is.

Charged beam exposure equipment has the advantage of being able to relatively easily control the deflection of the charged beam using electrical signals, and is therefore suitable for drawing fine patterns in accordance with pattern signals, but it is difficult to mass produce it. is considered inappropriate.

On the other hand, a stepper has the advantage of being able to transfer a large number of patterns at once onto a relatively large area, and is therefore frequently used as a semiconductor manufacturing device that can actually achieve mass production.

[Problem that the invention seeks to solve]

However, in the field of semiconductor applications, there is a demand for more highly integrated semiconductor devices, and in order to achieve this, it is necessary to further miniaturize the patterns to be formed on the wafer. Therefore, in semiconductor manufacturing equipment, in the process of printing relatively large patterns, mass productivity is maintained by exposing using a stepper, while in the process of printing fine patterns, exposure is performed using a charged beam exposure device. It is possible that

In this case, conventional charged beam exposure equipment requires
The following problem occurs when aligning a charged beam with respect to an alignment mark formed on a wafer in a previous process.

First, when detecting an alignment mark, a conventional charged beam exposure apparatus uses a method in which a charged beam used for exposure is deflected toward the alignment mark, and charged particles reflected from the alignment mark are detected by a detection element. This method is considered to be able to improve alignment accuracy because the amount of deflection from the alignment mark can be determined by the charged beam for exposure itself. However, the signal level of the detection signal obtained in this way is weak, and therefore the S/N ratio of the detection signal inevitably deteriorates. Therefore, in order to reduce this effect as much as possible, it is necessary to take complicated measures such as repeatedly sweeping a charged beam over one alignment mark and taking the average value of the detection results.

Moreover, the signal level of this detection signal is very low when the photoresist layer is multilayered (this is used as a method to eliminate the proximity effect as the pattern becomes finer). As it becomes thicker, it becomes even weaker, so it is unavoidable that it becomes more difficult to achieve alignment with high precision.

The second problem is that since a charged beam for processing with relatively high energy is used for detecting alignment marks, the photoresist layer covering the alignment marks is processed, resulting in highly accurate alignment detection. In order to obtain data, a plurality of 7-alignment marks must be provided for each chip, resulting in a problem that the semiconductor manufacturing process becomes complicated.

The third problem is that the alignment marks used in charged beam exposure equipment need to have a pattern suitable for exposure with a charged beam, so the processing process using a charged beam exposure equipment is inserted between the processing steps using a stepper. In some cases, it is complicated to print separate patterns as alignment marks used in successive processing steps. Incidentally, in each processing step, steppers and charged beam exposure devices must print different alignment marks depending on whether the exposure device used in the next processing step is a stepper or a charged beam exposure device. There is complexity.

The present invention has been made in consideration of the above points, and uses a common alignment mark to perform wafer positioning in a stepper and wafer positioning in a charged beam exposure apparatus with a simple configuration that is sufficient for practical use. The purpose of this paper is to propose a charged beam exposure apparatus that can perform high-precision exposure.

[Means for solving problems]

In order to solve this problem, the present invention provides an alignment mark detection laser beam LAX that detects the alignment marks 4X and 4Y attached to each chip 3 sequentially arranged on the wafer 2 held on the stage l; L
AY generating means ILXS IIY and a means (6 , 8) and each chip 3 based on the positions of the alignment marks 4X and 4Y detected by the alignment mark detection laser beam LAXSLAY.
means (
31, 7), and a means 31 for calibrating the distance between the alignment mark detection laser beams LAX, LAY and the charged beam 5 using a reference position mark 35 attached to the stage 1.

[Effect]

When the charged beam 5 processes one chip, it detects the alignment marks 4X and 4Y attached to the chip at a predetermined distance (i.e., l pitch or multiple pitches) from the chip using an alignment mark detection laser. Beams LAX and LAY are detected.

This detection result indicates that each chip 3 on the wafer 2 is exposed to the charged beam 5.
It is used as a reference signal for the means 31, 32 for aligning the exposure position of the charged beam 5, and at the same time, it is also used as a reference signal for correcting the exposure position of the charged beam 5 on the chip 3.

This eliminates the need to detect the alignment marks 4X, 4Y with the charged beam 5, and also allows the pattern of the alignment mark to be burned by the charged beam exposure device 4 to match the alignment mark to be burned by the stepper. As a result, it is possible to realize a charged beam exposure apparatus with a simple configuration that is practically compatible with a stepper.

〔Example〕

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

In FIG. 1, a stage 1 holds a wafer 2 to be exposed. On the wafer 2, a large number of square chips 3 are sequentially arranged in the X and Y directions, and X-axis alignment marks 4X and YI! are placed on the outside of two adjacent sides of each chip 3. h alignment marks 4Y are formed.

The surface of the wafer 2 is exposed through an electron lens 6 by a charged beam 5 supplied from a charged beam source 8 .

The electron lens 6 is controlled by a charged beam position control device 7, which controls the deflection of both electric beams 5 according to a pattern signal S3 supplied to the charged beam position control device 7, thereby creating a minute pattern on the surface of the chip 3. Draw. The photoresist thus formed to cover the chip 3 is exposed in a predetermined pattern.

The X-axis and Y-axis alignment marks 4X and 4Y are provided with alignment mark detection laser beams AX and LAY.
are supplied from two laser beam systems 11X and IIY.

In laser beam systems 11x and IIY, laser beams generated from laser light sources 12X and 12Y are incident on field lenses 14X and 14Y via collimator lenses 13X and 13Y, respectively. In this way, laser light that connects a spot on the wafer 2 can be output to the output ends of the field lenses 14X and 14Y, and can be input to the beam shifters 15X and 15Y.

The beam shifters 15X and 15Y are vibration-controlled by, for example, drive signals SIX and SIY supplied from the alignment mark detection laser beam generator 16, thereby controlling the alignment detection laser beams LAX and LAY.
X-axis and Y-axis alignment marks by 4X and 4Y
can be swept according to drive signals SIX and SLY.

The beams sent out from the beam shifters 15X and 15Y pass through the beam splitters 16X and 16Y, and then are reflected by the mirrors 17X and 17Y and irradiated onto the surface of the wafer 2.

In this embodiment, the X-axis and Y-axis alignment marks 4
As shown in FIG. 2, X and 4Y are composed of a plurality of diffraction grating marks 21 arranged in a line at predetermined intervals, whereas alignment mark detection laser beams LAX and LAY It has a flat cross-sectional shape stretched in the direction of the laser beam L.
AX and LAY reciprocate a predetermined number of times in the directions of arrows a and a2 so as to cross the row of diffraction grating marks 21, so that each time the laser beams LAX and LAY cross alignment marks 4X and 4Y, the diffracted light is mirrored. 1
It is arranged to occur in the directions of 7X and 17Y.

This diffracted light is reflected by mirrors 17X and 17Y, guided to lenses 21X and 21Y by beam splitters 16X and 16Y, and incident on light receiving elements 23X and 23Y through spatial filters 22X and 22Y. In this embodiment, the spatial filters 22X and 22Y pass only the diffracted light generated from the X-axis and Y-axis alignment marks 4X and 4Y, respectively.

Thus, each time the alignment mark detection laser beams LAX and LAY cross the X-axis and Y-axis alignment marks 4X and 4Y, the light-receiving elements 23X and 23Y receive
Alignment detection signals S2X and S2Y are obtained and supplied to the exposure control device f31.

Next, the operation of this embodiment will be explained.

(i) The alignment mark detection laser beams LAX and LAY detect the X-axis and Y-axis alignment marks 4 attached to the chip 3b adjacent to the chip 3a to be drawn first.
The stage 1 is moved by the stage position control device 32 until it can sweep Xb and 4Yb.

After that, the laser beam LAX for alignment mark detection
and LAY are swept, and the alignment detection signals S2X and S2Y generated along with this are sent to the exposure control device! 31. At this time, upon receiving the signal output from the exposure control device 31, the stage position control device 32 stops the stage 1 (and therefore the wafer 2).

The position of the stage 1 in the XY directions at this time is precisely detected by the stage position reading device 33.

This detection signal and alignment detection signals S2X and S2
Based on the drive signals SIX and SIY at the time when Y occurs, the exposure control device 31 determines the arabino mark 4Xb, 4Y.
Store the coordinate position of b.

(ii) Based on this stored information, the exposure control device 31:
The coordinate position at which the charged beam 5 should first be irradiated, that is, the position of the exposure start point of the chip 7''3a, is calculated and the result is sent to the charged beam position control device 7. Upon receiving this, the control device 7 The deflection of the charged beam 5 is controlled through the lens G. As a result, the charged beam 5 is irradiated onto the exposure starting point of the chip 3a, and exposure of one field is started from here.

Here, the term "field" refers to a drawing area that is exposed only by controlling the deflection of the charged beam 5 without moving the stage 1. In this example, one chip is entirely exposed by four fields of exposure. shall be.

This one field exposure is performed by supplying drawing pattern information from the exposure control device f31 to the charged beam position control device 7, and controlling the deflection of the charged beam 5 accordingly.

It should be noted that within this field, naturally, there are portions where no pattern is to be drawn, that is, portions where the beam should not be irradiated, but this can be dealt with by deflecting the charged beam 5 to non-exposed portions.

-(iii) When the exposure of one field is completed, the exposure control device 31 sends a signal to the stage position control device 32 for positioning the stage 1 to the next field. As a result, the stage 1 is moved, the exposure start point of the next field is aligned, and one field is exposed in the same manner as described above. At this time, it is not required to accurately align the stage 1, and the alignment error of the stage 1 caused thereby is corrected by controlling the deflection of the irradiation position of the charged beam 5. Therefore, the time required for alignment can be shortened.

This correction control is performed by the following operation.

That is, the exposure control device 31 calculates the correction amount based on the accurate position information of the stage 1 obtained from the stage position reading device 33 and the position information of the exposure start point of the field to be drawn from now on. This is done by driving the charged beam position control device 7 according to the calculation result.

(iv) By repeating the operations in (iii), 4
When the exposure of a field, ie, one chip, is completed, the exposure control device 31 sends a signal to the stage position control device 32. In response to this, the control device 32 adjusts the stage position so that the alignment mark detection laser beams LAX and LAY detect the X-axis and Y-axis alignment marks 4Xcs 4Y attached to the chip 3C adjacent to the chip 3b to be written next.
Stage l is driven until it can sweep c. Then, the above-mentioned operations (ii) to iv) are performed, and the operating chip 3a is exposed.

The above operations are repeated to expose all the chips 3 on the wafer 2.

In addition to the above configuration, the relative positional relationship of the alignment mark detection laser beams LAX and LAY with respect to the charged beam 5 is calibrated by the reference position mark 35 provided on the stage 1.

In this embodiment, the reference position mark 35 is constituted by a diffraction grating mark 35A having a pattern similar to the alignment marks 4X and 4Y described above with respect to FIG.

That is, every time the exposure operation for one field using the charged beam 5 is repeated a predetermined number of times (for example, 50 times), the stage 1 is moved so that the reference position mark 35 is exposed to the 7 alignment mark detection laser beams LAX, , LAY, or the charged beam 5.
The stage l is moved by the stage position control device 32 until it is irradiated by the laser beam LAXSLAY, and the stage l is moved from the position of the stage l where the reference position mark 35 is detected by the laser beam LAXSLAY until the charged beam 5 detects the reference position mark 35. 1's moving distance in the XY directions is detected.

This moving distance can be measured with high precision by the stage position reading device 33. Therefore, correction of the drift of the charged beam 5 can be performed with high precision.

The exposure control device 31 calculates the relative position of the alignment mark detection laser beams LAX and LAY and the charged beam 5 based on this detection result, and sends a pattern signal to the charged beam position control device 7 so that this becomes the reference value. Perform calibration operation on S3.

In this way, the charged beam 5 is practically a laser beam LA.
It is calibrated to maintain a predetermined relationship with respect to X and LAY, and therefore the pattern exposure work is performed with high precision based on the alignment marks 4X and 4Y.

This calibration operation is automatically performed at a predetermined cycle, and this cycle is detected based on the number of exposure operations described above.
It may be detected each time the operating time of the charged beam exposure apparatus reaches a predetermined time (for example, one hour), or it may be detected either before or after the replacement operation of the wafer 2 is detected.

With the above configuration, since the alignment mark is detected using laser light, S/
A detection signal with a good N ratio can be obtained, and thus the photoresist can be prevented from being processed unnecessarily.

In addition, by being able to select the frequency of the laser beam for alignment mark detection to any value as necessary, the processing process using a stepper that uses a light beam, and the processing process using a charged beam that uses a charged beam, When manufacturing a semiconductor including a processing step using a beam exposure device, the pattern of the alignment mark in the charged beam exposure device can be matched to the specification of the alignment mark in the stepper. Therefore, it is possible to obtain a semiconductor manufacturing apparatus with good alignment accuracy throughout all processing steps.

In addition, in order to obtain a detection signal with a good S/N ratio from this reference position mark 35, the pattern may be formed using a heavy metal such as Au. 35 by the charged beam 5, the level of the detection signal cannot be very high, so as shown by the symbol a in FIG.
By scanning one by one and taking the average value of the plurality of detection results, more accurate position detection can be achieved.

Incidentally, when processing a wafer with 50 chips of lO [10] size using a charged beam exposure system with a throughput of 50 pieces/hour for 4-inch wafers,
Assuming that the processing time is approximately 1 [second] per l chip, the stepping distance is 0.1 (second) for 1 [field] 5 (mm), while it takes 0.15C to detect the alignment mark. seconds] can be assigned.

In the above embodiment, the X-axis and Y-axis alignment marks 4X and 4Y are detected by detecting the alignment marks 4X and 4Y attached to the chip next to the chip 3 exposed by the charged beam 5. However, the position where the alignment mark is detected is not limited to the adjacent chip, but may be the chip to be exposed, or one or more adjacent chips.

Furthermore, in the above description, an embodiment in which the present invention is applied is described in which stage 1 is sent one pitch at a time in step-and-repeat mode. The present invention can be similarly applied to a charged beam exposure apparatus that draws a pattern using the charged beam 5 while moving the charged beam 5 .

Further, in the above description, a case has been described in which the X-axis and Y-axis alignment marks 4X and 4Y and the reference position mark 35 are of a diffraction grating configuration. The present invention can also be applied to marks. For example, a method that detects scattered light from the edge of an alignment mark may be applied.

Furthermore, in the above, the pattern 4X, 4Y or 35A
Although the laser beams LAX and LAY are swept a plurality of times, the same effect can be obtained even if the laser beams LAX and LAY are swept only once.

〔Effect of the invention〕

As described above, according to the present invention, the alignment mark attached to each chip is detected using a laser beam, and based on the detection result, stage movement information and charged beam positioning information are obtained. As a result, with a simple configuration, it is possible to easily realize a charged beam exposure apparatus that can control the exposure position of a charged beam with high precision and can easily match the alignment operation of a stepper.

[Brief explanation of the drawing]

FIG. 1 is a schematic block diagram showing an embodiment of a charged beam exposure apparatus according to the present invention, and FIGS. 2 and 3 show its X-axis and Y-axis alignment marks 4X. 4 is a route map showing the configuration of a 4Y and a reference position mark 35. FIG. 1... stage, 2... wafer, 3...
...Chip, 4X, 4Y...X-axis, Y-axis alignment mark, 5...Charged beam, 6.
...electronic lens, 7...charged beam position control device, IIX, IIY...lens beam system, 23X, 23Y old...light receiving element, 31...・
Exposure II control device, 32... Stage position control device.

Claims (1)

[Scope of Claims] Means for generating an alignment mark detection laser beam for detecting alignment marks attached to each chip sequentially arranged on a wafer held on a stage; means for generating a charged beam that exposes each position in the chip corresponding to the position of the alignment mark detected by the alignment mark; and the position of the alignment mark detected by the alignment mark detection laser beam. means for aligning each chip to the exposure position of the charged beam based on the above, and calibrating the distance between the alignment mark detection laser beam and the charged beam at a predetermined period using a reference position mark attached to the stage. 1. A charged beam exposure apparatus characterized by comprising means for performing each exposure.