JP3334341B2 - Electron beam exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electron beam exposure method, and more particularly to a method for detecting a mark for positioning.

In recent years, the degree of integration and functions of semiconductor devices have been improved, and they are widely used in fields such as information and communications. As a process technology, high integration by fine processing is important.
Although the limit of the pattern width is about 0.3 μm in the photolithography technology, fine processing of 0.1 μm or less can be realized with an alignment accuracy of 0.05 μm or less in electron beam exposure.

[0003] Therefore, if an electron beam exposure apparatus capable of exposing 1 cm ² in about 1 second can be provided, it is possible to use 1 gigabit or 4 gigabits without allowing other lithography techniques to follow in terms of integration degree, alignment accuracy, processing time, reliability, etc. This makes it possible to manufacture gigabit memories and 1-megagate LSIs.

[0004]

2. Description of the Related Art FIG. 3 is a block diagram showing an example of an electron beam exposure apparatus. In the figure, lanthanum hexabolite (L
aB ₆ ) The cathode, Wehnelt, and anode constitute an electron gun, which accelerates and emits electrons at an acceleration voltage of 10 to 50 KV. The emitted electron beam is aligned with the optical axis by an alignment coil AL1, and a rectangular first shaping aperture (first slit) is formed.
, And converged by the first lens, deflected by the slit deflector and the slit return deflector to form an image of the first shaping aperture on the second shaping aperture (second slit). The portion passing through the second shaping aperture becomes a rectangular electron beam.

[0005] By changing the voltage applied to the slit deflector, the position of the image of the first shaping aperture is moved,
The shape of the beam passing through the second shaping aperture can be shaped into an arbitrary rectangle.

Further, the rectangular electron beam having passed through the second shaping aperture is irradiated by an irradiation lens (second lens), and is reduced to 1/100 by a two-stage reduction lens (third lens, fourth lens). On the way, electrons which are largely shifted from the axis are removed by a round aperture.

[0007] Finally, a rectangular electron beam is projected onto the semiconductor substrate by the projection lens (fifth lens). FIG. 4 is a configuration diagram showing an example of a signal processing circuit for detecting a positioning mark.

In the drawing, 1 is a positioning mark, 2
Is a detector, 3 is an amplifier, 4 is an A / D converter, 5 is an adder, 6 is an original waveform memory, 7 is a digital arithmetic circuit (a),
8 is a peak detection circuit and a 0 level detection circuit, 9 is a differential waveform memory, 10 is a digital operation circuit (b), and 11 is a smoothed original waveform memory.

When electron beam deflection data is generated from the pattern generator of the exposure apparatus and the positioning mark 1 is scanned with an electron beam, a reflected electron signal corresponding to the step of the mark is generated. This signal is detected by a photodetector (PIN diode) 2 and amplified.

Next, this signal is sent to a digital memory unit, where the signal is converted from analog to digital and stored in a waveform memory. This waveform memory performs averaging in accordance with the number of electron beam scans to obtain a signal with a good S / N ratio.

Further, the stored data is digitally differentiated to determine the peak position of the differentiated data. Since the beam is reciprocated on the mark, four peaks are obtained. The peak position is used as the edge position of the mark, and the deviation between the beam deflection center and the mark groove is calculated as the positional deviation amount (differential peak method, see FIG. 5).

On the other hand, in the original waveform, the center position of a tangent drawn on the original waveform at a point corresponding to a point where a differential peak is given, and the horizontal lines A and B passing through the point of the original waveform corresponding to a point where the differential waveform becomes zero And the deviation between the beam deflection center and the mark groove is calculated as a positional deviation amount (slice level method, see FIG. 6). Here, the intersection between A and B and the original waveform is called a zero-cross point.

There are, for example, the following two types of beam deflection methods for detecting the mark position. The first method is shown in FIG.
This is a method in which an electron beam is scanned over both edges of a mark groove as shown in FIG. The original waveform and the differentiated waveform obtained by this method are as shown in FIG. This method is referred to herein as rough scanning.

In the second method, as shown in FIG. 8A, only one edge of the groove of the mark is scanned for a short time, then the deflection center of the beam is moved by a predetermined mark width, and the other edge is moved. This is a short scanning method. The original waveform and differential waveform obtained by this method are as shown in FIG. This method will be referred to herein as fine scan.

[0015]

When the mark is detected by the rough scan, the beam scanning distance is long. Therefore, if the scanning pitch is made small for improving the accuracy, the number of scanning points increases and the scanning time becomes long. Further, when the time is long, there is a disadvantage that the periphery of the mark is exposed.

On the other hand, when the mark is detected by the fine scan, the scanning time is short because the beam scanning distance is short, and the influence on the pattern around the mark is small. However, when the position of the center of the beam deflection and the center of the mark are largely displaced, or when the shape of the mark is different from the design size through the process, there is a high possibility that the beam will not hit the mark edge. There is.

SUMMARY OF THE INVENTION It is an object of the present invention to accurately and quickly detect the position of a mark for alignment.

[0018]

Means for solving the above problems are as follows: 1) Beam scanning for detecting the position of a mark for alignment is performed in two stages, and the first scanning is performed with a scanning width larger than the mark width to make the center of the mark. In the second scan, the position deviation between the beam deflection center position and the beam deflection center position is calculated, and the beam deflection center position is shifted by the position deviation amount obtained in the first scan so as to come to the mark center position. An electron beam exposure method in which scanning is performed, and the sum of the positional deviation amount obtained in the first scanning and the positional deviation amount obtained in the second scanning is used as the deviation amount between the mark position and the beam deflection center position; The two scans have a shorter scan width than the first scan,
(3) The electron beam exposure method according to (1), wherein the scanning pitch is fine and the number of scans is large; or 3) The mark center position and mark width are calculated from the mark edge positions obtained in the first scan, and the mark The electron beam exposure method according to the above 1, wherein the scanning interval is used when scanning both edges, or 4) the scanning width of the second scanning is set to be closest to the differential peak of the differential waveform obtained in the first scanning. Derivative value exceeds 0 2
It is achieved by the electron beam exposure method according to the above item 1, which includes at least one of the following points.

[0019]

According to the present invention, the beam scanning for detecting the mark position is performed in two stages, the first scanning is performed with a scanning width larger than the mark width, and the difference between the mark position and the beam deflection position (irradiation position) is determined by this scanning. The second scan is performed only on the edge portion of the mark by calculating and shifting the beam deflection position on the sample by this difference. Here, the sum of the difference obtained in the first scan and the positional shift amount obtained in the second scan is defined as the shift amount between the mark and the beam irradiation position.

That is, according to the present invention, the second scan (fine scan) is performed based on the information obtained in the first scan (rough scan), and the advantages of both are provided, namely, the high speed of the first scan and the speed of the second scan. Incorporating high precision, mark position detection is performed accurately and at high speed.

[0021]

1 (A) to 1 (C) and 2 (A) to 2 (C) are explanatory views of an embodiment of the present invention. First scan (rough scan): FIG. 1A is an explanatory diagram of the first scan (rough scan), FIG. 1B is an original waveform, and FIG. 1C is a differential waveform.

In the first scan (rough scan), the mark is scanned over both edges of the mark, and the differential peak method is used to calculate the difference δ between the beam deflection center position and the mark center position, and the mark size ( Mark width).

Now, assuming that the scanning width of the beam is 0 to 100 on the coordinates and the differential peak values are A and B, the beam position = (0 + 100) / 2 = 50 the mark position = (A + B) / 2 δ = 50− ( A + B) / 2 mark size = B−A. In the figure, a, b, c, and d are zero-cross points.

Second scan (fine scan): FIG. 2 (A)
FIG. 2B is an explanatory diagram of the first scan (rough scan), FIG. 2B shows an original waveform, and FIG. 2C shows a differential waveform.

Next, the deflection center position is corrected by the difference δ calculated in the first scan, and the mark size calculated in the first scan is given as the scanning interval of the second scan. The required deflection width for the second scan is a value obtained by adding a margin α to a larger value of b−a or d−c.

In the figure, a ', b', c ', d' are zero-cross points, and this information is necessary for detecting the mark position.
Next, features of the two-stage scanning will be described.

In the first scan, since the beam is scanned over both ends of the mark, it is sufficient that the beam hits the mark edge at the time of the second scan. Therefore, the pitch of this scan may be relatively coarse. Although the distance is large, the time required for mark detection is short.

In the second scan, only the mark edges are scanned in accordance with the edge positions obtained in the first scan. That is, since the difference δ between the mark center and the beam deflection center is obtained in the first scan, the shift amount is added to the deflection amount of the deflector, and the beam deflection center and the mark center are corrected to be substantially at the same position. I do.

In the second scan, after scanning one mark edge, a position shifted by the mark size (mark width) obtained in the first scan is set as the other mark position, and only the edge portion is scanned. Since this mark size is measured by actual scanning, it is not affected by a change in the mark shape due to the process, and shows an accurate mark width.

In the second scan, since accurate position measurement is required, the position of the mark is detected by reducing the scan pitch and applying, for example, the above-described slice level method. Even if the scanning pitch is made fine in the second scanning, the scanning distance is short, so that the time required for the scanning is shorter than when rough scanning is performed at the same fine pitch.

Further, when the mark edge is scanned in the second scan, if the slice level method is used, the mark position is detected by using the zero cross point and the peak position of the differential waveform. The scan width is determined in the first scan so as to include the scan width.

[0032]

According to the present invention, the position of a mark for positioning in electron beam exposure can be detected accurately and at high speed.

[Brief description of the drawings]

FIG. 1 is an explanatory view of an embodiment of the present invention (1).

FIG. 2 is an explanatory view of an embodiment of the present invention (2).

FIG. 3 is a configuration diagram of an electron beam exposure apparatus.

FIG. 4 is a configuration diagram of a mark position detection signal processing circuit.

FIG. 5 is an explanatory diagram of a differential peak method.

FIG. 6 is an explanatory diagram of a slice level method.

FIG. 7 is an explanatory diagram of a rough scan.

FIG. 8 is an explanatory diagram of a fine scan.

[Explanation of symbols]

 1 Positioning mark 2 Detector 3 Amplifier 4 A / D converter 5 Adder 6 Original waveform memory 7 Digital operation circuit (a) 8 Peak detection circuit and 0 level detection circuit 9 Differential waveform memory 10 Digital operation circuit (b) 11 Smoothed original waveform memory

────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-59-69924 (JP, A) JP-A-56-33830 (JP, A) JP-A-3-104109 (JP, A) (58) Field (Int.Cl. ⁷ , DB name) H01L 21/027 G03F 9/00

Claims (4)

(57) [Claims] 1. A beam scanning for detecting a position of a mark for alignment is performed in two stages, and a first scanning is performed with a scanning width larger than a mark width to calculate a positional deviation amount between a mark center position and a beam deflection center position. In the second scan, only the edge portion of the mark is scanned such that the center position of the beam deflection is shifted to the center position of the mark by the positional shift amount obtained in the first scan, and the position obtained in the first scan is changed. The shift amount and the second
An electron beam exposure method, wherein a sum of positional deviation amounts obtained by scanning is used as a deviation amount between a mark position and a beam deflection center position. 2. The electron beam exposure method according to claim 1, wherein the second scan has a shorter scanning width, a smaller scanning pitch, and a larger number of scans than the first scanning. 3. A mark center position and a mark width are calculated from a position of a mark edge obtained in the first scan, and the scan interval is used when scanning both edges of the mark in the second scan. The electron beam exposure method according to claim 1. 4. The scanning width of the second scan includes at least two points where the differential value exceeds 0, which is closest to the differential peak of the differential waveform obtained in the first scan. Item 7. An electron beam exposure method according to Item 1.