JPS6115325A - Mark detection method - Google Patents

Abstract

PURPOSE:To prevent an error in recognition of a mark position by inhibiting the process of extracting a position of charged particles from a wave-form signal when an amplification factor is beyond a specified value. CONSTITUTION:When a beam 20 scans a target 22 having a mark by a beam scan controller 21, a mark signal generated from the target 22 appear in an output line 29 as a change in voltage through a detector 23 and an offset level modulation amplifier 24. Output of an A/D converter 25 is added to a maximum value detecting circuit 26 comprising a maximum value holding register 27. In an amplification factor preset circuit 28, the output of the register 27 is converted into a coefficiency K which is satisfy an equation Eo=K*Ei=constant wherein Ei is a voltage level of the line 29 and Eo is a voltage level of an output line 31. And an amplification factor of an amplifier 30 is thus controlled. The circuit 28 outputs the signal 33 corresponding to the coefficiency Kalpha to a detection circuit 32 which compares the level Eo with a slice level of Eo/2 and detects a mark position. This function is controlled by the signal 33.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention is directed to highly reliable and automatic detection of mark positions on wafers or masks in charged particle beam devices such as electron beam lithography devices. The present invention relates to a mark detection method suitable for.

[Background of the invention]

In the conventional system, as described in Japanese Patent Laid-Open No. 55-85028, the amplification degree of the amplifier is adjusted so that the maximum amplitude of the mark waveform signal is constant. In this method, in order to know the strength of the mark waveform signal, a means was required to scan the charged beam prior to the mark detection operation and determine the optimum amplification factor of the amplifier from the maximum value of the obtained mark waveform signal. The processing time required for this is long, accounting for nearly 30% of the entire mark detection process.

・For this reason, the amplification factor of the amplifier is set in real time, and the above
A method was devised to omit this method, but in this case, it is not possible to know the maximum value of the waveform signal in mark detection processing in advance, and the amplification factor of the amplifier must be changed each time according to the strength of the input signal. It won't happen. Therefore, when the input signal is small, the amplification factor of the amplifier becomes large, and noise portion signals other than the mark waveform signal sometimes cross the slide level, resulting in incorrect mark position data being detected.

FIG. 1 is a diagram explaining this. FIG. 5A is a plan view of the mark, and in order to detect the mark 10, the charged particle vinyl is scanned as 11, 12, 13, and 14.

The signal obtained when the beam scans the mark 10 is
This results in 15 waveforms.

Figures (b) and (c) show the amplification factor of the amplifier of the detection circuit and the detection signal, with the number of scans plotted on the horizontal axis.

Since the scan 11 scans the area other than the mark, the amount of signal obtained is small, and therefore the amplification factor is set large so that the maximum value of the detection signal matches the target output 16.

When the beam falls on the mark, the signal amount increases and the amplification factor decreases, so as shown in FIG.
．． At 14, the mark signal matches the target value 16. This signal is compared with the slice level 17 to determine the mark position.

By the way, part A shown in Figure 1(c) is the period until the beam hits the mark, and even though the amplification factor is constant during this period, its value is large, so there is a slight change in the signal amount. The detection signal also changes so as to oscillate greatly. Therefore, the slice level 17 may be crossed as if there were a mark, which caused the mark position to be misidentified. - [Object of the Invention] The object of the present invention has been achieved by paying attention to this point, and even if an automatic amplification degree determination circuit for keeping the signal amplitude constant is used, only the mark waveform signal can be correctly recognized. The object of the present invention is to provide a mark detection method for detecting marks.

[Summary of the invention]

The reason why signals other than mark waveform signals are mistaken as mark signals is that the amplification factor is amplified to maintain a constant signal amplitude even for minute signals when the charged particle beam is irradiating areas other than the mark area. , noise, is at the point where the sound exceeds the slice level. In the present invention, when the amplification factor exceeds a certain value, the process of extracting the charged particle position from the waveform signal is prohibited to prevent misidentification.

Another method according to the invention is that in any signal,
The function of the automatic amplification factor setting circuit that keeps the signal amplitude constant has been revised, and the signal from the mark part is kept at an appropriate amplitude, but other noise signals are amplified at amplification i so as not to be unnecessarily amplified. This method uses the upper value for the rate. By doing this, the signal of the noise part no longer crosses the slice level, making it possible to prevent misidentification of the mark position.

[Embodiments of the invention]

An embodiment of the present invention will be described below with reference to FIG. When the beam 2o scans the target 22 having a mark by the beam scanning control unit 21, the mark signal generated from the target 22 passes through the detector 23 and the offset level adjustment amplifier 24, and appears as a voltage change on the output line 29. .

The output of the A/D converter 25 is stored in the maximum value holding register 27.
is added to a maximum value detection circuit 26 having a maximum value detection circuit 26. register 2
7 is obtained by setting the voltage level of the output line 29 to Ei in the amplification factor setting circuit 28, and setting the voltage level of the output line 31 to Ei.
o, it is converted to a coefficient K such that E o =K*Ei=constant, and controls the amplification factor of the multiplier 30.

Furthermore, the amplification factor setting circuit 28 outputs a signal 3'3 corresponding to the α value as a coefficient to the mark position detection circuit 32. The mark position detection circuit 32 compares the voltage level EO of the output line 31 with the slice level of EO/2,
This is a circuit that detects the mark position, but its function is based on signal 3.
It is controlled by 3.

That is, when the value of the coefficient exceeds a certain value, the detection operation in the mark position detection circuit 32 is prohibited by the signal 33, so that an erroneous mark position is not determined.

FIG. 3 is a diagram showing another embodiment according to the present invention.

Similarly to FIG. 2, when the beam 20 scans the target 22 having a mark by the beam scanning controller 21, the mark signal generated from the target 22 is transmitted to the detector 23. Output line 29 via output level adjustment amplifier 24
appears as a voltage change.

The output of the A/D converter 25 is stored in the maximum value holding register 27.
is added to a maximum value detection circuit 26 having a maximum value detection circuit 26. register 2
The output of No. 7 is sent to the amplification factor setting circuit 28, where the voltage level of the output line 29 is set to Ei, and the voltage level of the output line 31 is set to Eo.
When Eo=*Ei=constant, the coefficient K is converted to control the amplification factor of the amplifier 30.

However, the value of K has an upper limit value Kmax, which cannot be exceeded. Kmax is a sufficient value when the voltage level of the output line 31 during mark scanning is set to Eo, but is a small value for other noise signals.

A mark detection signal according to this embodiment is shown in FIG. 3(b).

Since the amplification factor in the noise section A is suppressed by the upper limit of the amplifier circuit, the noise is at slice level 1 as shown in Figure 1(c).
7 is not cut off, so there is no possibility of misunderstanding the mark position.

〔Effect of the invention〕

According to the present invention, it is possible to prevent erroneous recognition of mark positions in mark detection processing, and highly accurate mark detection is possible.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing changes in amplification factor and detection signal when an automatic amplification factor setting circuit is used, and FIGS. 2 and 3 are diagrams showing embodiments of the present invention. 24...Offset level-adjusting amplifier, 25...A
/D converter, 26... Maximum value detection circuit, 27...
・Maximum value holding register, 28 peak amplification factor setting circuit, 3°・
...Amplifier, 32...Mark position detection circuit, 33...
・MaY l mouth/ρ 2nd mouth 3rd ward

Claims (1)

[Claims] 1. Means for detecting reflected electrons or secondary electron signals generated from a mark by scanning a charged particle beam over the mark, and a mark detected by the means. means for automatically controlling the gain of an amplifier for amplifying the mark waveform signal so that the amplitude of the waveform signal is equal to a target value; and comparing the mark waveform signal with a slice level;
processing means for determining the charged particle beam position when the mark waveform signal crosses the slice level;
In the mark detection circuit for determining the mark position, a mark detection method is characterized in that a detection processing method is not changed between signal detection from a portion where a mark is present and signal detection from a portion where a mark is not present. . 2. The detection processing method in the mark detection circuit described above is characterized in that the processing means for determining the charged particle beam position is prohibited from operating while the amplifier gain exceeds a desired range. A mark detection method according to claim 1. 3. As a detection processing method in the mark detection circuit described above, an upper limit value is set for the gain of the amplifier of the detection circuit so that the signal waveform in the part where no mark exists in the mark waveform signal does not exceed the slice level. A mark detection method according to claim 1, characterized in that: