JPS59222930A - Method for detection of marking position in electron beam exposure and device thereof - Google Patents

Abstract

PURPOSE:To contrive reduction in number of scanning as well as to enable to detect a marking position in a high speed and highly accurate manner by a method wherein a section where mark edge data will be taken in is provided, and a detection of marking position is performed by providing a plurality of threshold values between peak value levels. CONSTITUTION:A reflected electron detector 3 detects a reflected electron coming from a sample face whereon a mark 2 is formed, a peak signal detecting circuit 5 detects the counter value of the peak position of a detection signal, the peak value of each peak and offset level value L0 are detected, and a control circuit 6 converts the two peak values, which are located most closely to the offset level value L0, to an analogue voltage. Based on said analogue voltage, a suitable threshold voltage is given to n-pieces of comparators 11. A mark edge data take-in section Tw and the threshold voltage for each comparator 11 are set by the first electron beam scanning. Subsequently, the scanning of an electron beam is repeated in the number of times prescribed in advance. The mark position is determined by performing an arithmetic operation based on the data detected using an arithmetic operational circuit 15.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and apparatus for detecting a reference mark position at high speed and with high precision in electron beam direct writing.

In direct writing using an electron beam, in each step of manufacturing a microcircuit, it is necessary to draw at a predetermined position on the sample by slightly overlapping the taps. Therefore, in advance, the positioning mark 2
As shown in FIG. 1(a), a step is formed on the sample surface, the mark 2 is scanned with an electron beam to detect the mark position, and drawing is performed using this position as a reference. In conventional mark position detection, reflected electrons from the mark 2 generated by electron beam scanning are detected by a plurality of reflected electron detectors. The output waveform when the outputs of all the detectors are added is shown in FIG. 1(b). An appropriate threshold level LL is set for this detection signal to generate a binary signal. Count the beam scanning clock pulse at the same time as the beam scanning starts! -1, and read the count value C+f+C+r when the above-mentioned binarized error signal occurs. C.I.J.
, +C1r is mark edge position data, (
C,,+C1r)/2 as the mark position (・ru.

However, the detection signal is superimposed with the noise of the beam itself, noise due to backscattering of the beam, noise of the detector, etc., and these random noises reduce mark position detection accuracy. Therefore, the mark position is determined by performing beam scanning a plurality of times and averaging the mark position data obtained for each scan to improve mark detection accuracy. It is known that if scanning is performed m times, the detection accuracy will be improved by a factor of two.

In actual writing, the beam scans about 20 to 40 times to detect one mark (・).Because the beam is scanned multiple times like this, there is a problem that related marks cannot be detected. There was. 1 Impulse outpatient H
Sound may generate erroneous signals, resulting in erroneous mark position detection.

In conventional mark position detection, in order to prevent the influence of this erroneous signal, a value Δ-1c+rCJ corresponding to the mark width is calculated, and if Δ is not within the specified tolerance value, the detected mark/edge position data is c, , , C1r was rejected and the temperature was changed to ℃. However, with this method, even if the section of the detection signal corresponding to the mark edge is normal, the mark edge data is discarded due to an erroneous signal, resulting in a problem in that the mark position cannot be detected.

In order to eliminate these drawbacks, the present invention detects the positions and peak levels of a plurality of maximum and minimum peaks of the detection signal, and determines the interval for capturing mark/enon data based on the peak positions. Mark position detection is performed by setting multiple threshold values between the maximum peak closest to the offset level value and the user's peak value level among the detected peak values, to prevent erroneous signals. By removing marks and obtaining a large number of marks, positions, and data in a single electron beam scan, we reduced the number of set meals and achieved high-speed, high-intensity mark position detection. The present invention provides a mark position detection method and apparatus (I).

FIG. 2 is an embodiment of the present invention. This embodiment will be explained below.The control circuit 6 generates a scan start signal and sets the scan start point and scan distance 141 specified in advance.
1, the electron beam I scans over the mark 2. Simultaneously with the start of the scan, the counter 7 starts counting the number of scan clock pulses.

The backscattered electron detector 3 detects backscattered electrons from the sample surface on which the mark 2 is formed, and the detection signal is amplified to a constant level by the AGC circuit 4. The peak signal detection circuit 5 has a third
Detects the counter values P+, P2, P3, P4 of the peak positions of the detection signal and the respective peak values and offset level values, and sends this information to the control circuit 6.Control circuit 6 is
This peak position II:N information is temporarily stored in Renostar 8, and the mefuseno 1-level value is also stored.

The two peak values Vh, V, p, (h large peak f+fr Vh, and the minimum peak value V) closest to , are manually input to the DA circuit 10, IOa, and converted into an analog voltage. Based on this voltage, n comparators I] are assigned appropriate threshold voltages V1°v2 . ,, -vn(Vh>V, ,
v2. −Vn>V2) is given.

1, mark/edge/category capture section setting circuit 9
generates a mark/edge/category capture section setting signal Eg as shown in FIG. The mark edge data of only the section Tw of P is imported. As described above, in the first electron beam scan, the mark, edge, and take-in sections Tw are set, and the threshold voltages are set for each comparator H.

Thereafter, the scanning of the electron beam is repeated a predetermined number of times.Comparator 1 compares the detection signal generated by each scanning with each of the threshold voltages, and only when they match, the monostable multi・Generate a binarized signal pulse from the pie flake 12.ANDK1-13 ANDs this pulse signal and the Mark Enon Cater capture section setting signal Eg, and generates it in a period other than the above section TW. Exclude the pulsed signal.

The output signal of the AND gate is used to input the color to the latch circuit 14.
Ranochinog the 0 count value, and this data is expressed as 3'/-
The signal is sent to circuit 15. The arithmetic circuit 15 calculates the data CI J! detected for each n threshold level V, , V, , . . . + C2J, “””CnJ! +
M+ based on C1,r, C2r, """Cnr
” (C+1 +C2J', '"""CnJ 10C+r +Cwr +"・"Cnr)/2 n's 6
The mark position is determined by calculating it. This calculation is performed for each scan, and M=(MI+M2+...10Mm)/
m (here, m is the number of scans) is maintained as the final mark position.

As explained above, by setting the mark/edge/category acquisition section, it is possible to eliminate the negative influence from erroneous signals caused by impulse-like external noise N2, etc., as shown in Figure 3 (at) and the dotted line in Figure 4. In addition to being able to eliminate the mark detection accuracy, it is also possible to set a large number of threshold values (vl, v, , . . . -Vn) in the section where the detection signal waveform becomes steep. For example, if you do not set the caterer capture section T, v, the offset level L. Even if you set the value V1, it is difficult to detect the mark position accurately due to the noise. Even if the number of electron beam scans on the mark is set to /k, the same number of mark position data will be obtained and the detection accuracy will not change. Therefore, the threshold value v
If the number 1 can be set to a large number, the number of set meals can be reduced by an amount commensurate with the increase in the number of threshold values, which has the advantage of enabling high-speed detection of mark positions. In addition, due to the characteristics of the backscattered electron detector, the shape of the mark, etc., the actual detection signal is not symmetrical in the sections corresponding to the left and right mark edges, as shown in Figure 4, and the peaks of Pl and P4 and P2 and P3 There are very few words with matching values. If multiple thresholds are set between the maximum peak value and the minimum peak value in this state, the threshold signal and the detection signal may not always intersect normally, making it impossible to detect the mark position normally. -・Level value.

The two peak values closest to (maximum peak and minimum peak)
By setting a plurality of threshold values v1 between them, there is an advantage that this problem can be solved.

In the above example, the case where a step mark was used was explained (. , 11 mark).The detection signal obtained by scanning this mark with a short beam is as shown in Fig. 5 (bl).
. It is sufficient to detect the signal level L and set a plurality of threshold values Vi between these levels.

[Brief explanation of the drawing]

FIG. 1 (al is a diagram showing a reference mark, FIG. 1 is a diagram showing a base 1 (bl is obtained by scanning an Ij mark with an electron beam) detection signal and an embodiment of a conventional mark position detection method. 2 is a block diagram showing one embodiment of the present invention, and FIG. 3 is a block diagram showing an embodiment of the present invention.
Figures (a) and [b] are waveform diagrams for explaining an embodiment of the mark position detection method according to the present invention, Figure 4 is a time chart for explaining the present invention in detail, and Figure 5 is a waveform diagram for explaining an embodiment of the mark position detection method according to the present invention. FIG. 2 is a time chart for explaining another example of FIG. 1...'1L daughter beam, 2... Reference mark, 3...
・Reflection 11L detector, 4・AGCu path, 5・
Beak other number detection circuit r6, 6...control circuit, C
・ 7... Color/reduction, 8... Rensk, 9... Data acquisition section setting circuit, 10, 10a-DA conversion circuit, 11- Comparator, 12...° monostable multi-hiro router, 13...・AND circuit, 14...Latch circuit, 15
...Arithmetic circuit. Patent applicant Nippon Telegraph and Telephone Public Corporation agent Tsune Haku Mizu O and one other person exposed 4 illustrations 5 illustrations 3 illustrations

Claims (2)

[Claims] (1) An electron beam mark formed on a substrate that serves as a reference for position information is scanned with an electron beam, and changes in the intensity of reflected electrons from the mark and the substrate are detected as a detection signal of the mark by the scanning. In a beam exposure mark position detection method, the mark detection signal has multiple peaks indicating the maximum signal intensity, multiple peaks indicating the minimum signal intensity, and signal intensity from the substrate other than the mark portion. During the first lunar travel of the electron beam, the peak signal level I1m of the maximum peak and the minimum peak of the mark detection signal is reported and the position information of the peak is obtained. Based on the offset level 111 report, a part of the m high signal is specified and stored, and the peak value level is closest to the offset level value among the plurality of maximum peaks and minimum peaks. Select one very thick peak and one very small peak, set a plurality of threshold levels between the selected maximum peak value level and the selected minimum peak value level, and set the plurality of threshold levels. A pulse signal corresponding to the intersection point is generated only at the intersection point of the mark detection signal waveform and the mark detection signal waveform, and is within the specified section. The intersection point between the mark detection waveform and the plurality of threshold values is repeatedly detected in the designated section, and the pulse signal corresponding to the intersection point is repeatedly generated. A method for detecting mark position in electron beam exposure, characterized in that the position of the mark is detected based on the pulse signal. (2) means for scanning and reflecting the base mark formed on the substrate with an electron beam; circuit means for setting an appropriate threshold level for the detection signal; circuit means for generating a pulse signal corresponding to the intersection of the detection signal waveform and the detection signal waveform; and circuit means for determining the scanning amount of the electron beam using the pulse signal and obtaining mark position information from the scanning amount. In a mark position detection device for exposure, a partial section of the detection signal is specified and stored based on information on the peak signal level values of the maximum peak and minimum peak of the detection signal waveform, the position information of the peaks, and the information on the offset level value. have the means to
Means for selecting one by one the maximum peak and minimum peak whose peak value level is closest to the offset level value among the plurality of maximum peaks and minimum peaks;
means for setting a plurality of threshold levels between the selected maximum peak value level and the selected minimum peak value level, and a means for setting a plurality of threshold levels between the selected maximum peak value level and the selected minimum peak value level; Moreover, the electron beam is characterized by comprising means for generating a pulse signal corresponding to the intersection point only within the specified section, and means for detecting the position of the mark based on the pulse signal. Mark position detection device for exposure.