JPH0646554B2 - Automatic focusing method in electron beam equipment - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic focusing method, and more particularly to an automatic focusing method in an electron beam apparatus.

[Background of the Invention]

The waveform signal required for focus detection is obtained as follows. That is, the first method is a cross-shaped diameter of several tens of μm.
This is a method in which the metal wire is scanned with an electron beam in the X direction or the Y direction, and the transmitted electron signal obtained at that time is detected by the Faraday cup provided on the upper part of the wire. The second method is a width of several tens of μm and a height of 0.2 to 0.3 μm formed on Si.
The crossed metal mark with a length of several 100 μm is scanned with an electron beam, and the reflected electron signal obtained at that time is detected by the semiconductor detector SSD.
It is a method to detect with.

Then, using one of the above two signals, focus detection is performed as follows.

The first method of focus detection adjusts the focus so that the width of the waveform signal is minimized. The second method obtains a differential signal once from the above waveform signal and adjusts the focus so that the width of the differential signal is minimized. The third method obtains a differential signal twice from the above waveform signal and performs focus adjustment so that the peak value of the differential signal becomes maximum. The fourth method uses a mark having a size sufficiently smaller than the size of an electron beam used as a target when obtaining a waveform signal, and the waveform signal obtained at that time is equivalent to the differential signal already described in the first method. Since the signal has characteristics, the focus is adjusted so that the width of the signal is minimized. The fifth method differentiates the signal obtained when performing the third method once, and performs focus adjustment so that the peak value of the differentiated signal becomes maximum.

The first to third methods are often used as the conventional methods.
For example, the method described in JP-A-51-23065 is
The first method has drawbacks such that it is difficult and unstable to extract the width of the waveform signal in a defocused state, and the focus detectability is inferior to that of the differential method.

The second to fifth methods are excellent in focus detection ability because they use differential signals, but have a drawback that focus detection tends to be unstable because they use differential signals as they are.

[Object of the Invention]

The present invention has been made in view of this point, and an object thereof is to provide an automatic focusing method which is stable and has high speed while maintaining high detectability in an electron beam apparatus.

[Outline of Invention]

The present invention utilizes a differentiating method with excellent focus detectability to detect the state of focus, and with respect to the instability of focus detection that occurs with the adoption of the differentiating method, a large number of focus detection signals are detected. This is solved by adopting a method of performing curve approximation and performing final focus detection using the approximation curve.

If the in-focus point cannot be detected by one curve approximation, the result obtained from the curve is used again to repeat the series of operations of extracting the focus detection signal, curve approximation, and focus detection to finally achieve the in-focus point. The method of detection is also used together.

Example of Invention

 First, the principle of the present invention will be described.

As shown in FIG. 1 (a), a target or mark 2 made of Au, for example, is scanned with an electron beam 1 in the direction of the arrow, and reflected electrons (or secondary electrons or transmitted electrons) at that time are scanned.
Is detected by the semiconductor detector SSD and converted into an electric signal,
A mark waveform signal as shown in FIG. When this waveform signal is differentiated once, a one-time differential signal (first-order differential signal) corresponding to the size of the electron beam as shown in FIG. 7C is obtained, and the signal is further differentiated, that is, twice differentiated. A two-time differential signal (secondary differential signal) indicating the sharpness of the electron beam distribution as shown in FIG. That is, the maximum value of the twice differentiated signal is defined as the sharpness. The sharpness value is inversely proportional to the blur of the beam. That is, when the focus is achieved, the sharpness value is large, and conversely, when the focus is not achieved, the above value is small.

FIG. 2 shows an example of the focus adjustment of the electron optical system and the twice differentiated signal. In the figure, (a) is a once-differentiated waveform, and (b) is a twice-differentiated waveform. In this figure, the waveform in the middle is the two-time differential signal at the time of focusing. The left and right waveforms show the case where defocus of 100 μm occurs (in this example) downward and upward, respectively.

Fig. 3 shows that the focal plane of the electron optical system changes in 5 steps in the range of 120 µm above and below the object surface in 60 µm steps ((c) in the figure is the second derivative waveform diagram, and one cycle of each waveform corresponds to one step) )
An example of the result obtained by extracting the maximum value of the twice-differentiated signal obtained at each step, that is, the sharpness, and applying the quadratic curve approximation to the sharpness is shown. From this figure, the approximate curve
It can be seen that the maximum value (b) of (a) matches the in-focus point (when the focus error is 0).

Next, a specific example when the present invention is adopted in an electron beam drawing apparatus will be described. In the above device, there are usually two focus adjusting lenses, O and OS. First, the current value I _o of the O lens is set to a predetermined value. In this state, the current value I _os of the OS lens is changed in 9 steps, the mark waveform signal is taken in at each step, and the differentiation process is performed twice to detect the sharpness. After that, a quadratic or cubic curve is fitted to each I _os and the detected sharpness, and it is determined whether the maximum value of this approximate curve occurs near the middle of the I _os change width of 9 points. Find out. If it occurs near the center, it means that the in-focus point is detected.
_os is extracted and the OS lens is set to a predetermined constant value I _osl (for example, zero), and at the same time, the extracted current value I _os is converted into the current value of the O lens (the conversion formula of I _os and I _o is _{Predetermined} ), and the current value I _oj is O
Focus on the same position by pouring it into the lens.

On the other hand, if the maximum value of the approximate curve is not near the center, the current value of the OS lens corresponding to the maximum value found at that time is referred to, the current value I _o of the O lens is slightly changed, and focus detection is performed by the same method. . Since the number of times of detecting the in-focus point is predetermined, if the in-focus point cannot be detected within the number of times, the operator is notified that the in-focus point cannot be detected.

FIG. 4 shows an example of the state of focus detection and adjustment according to FIG. In the example of the figure, the focus detection in the X direction has been performed four times, and the current value I _o of the first O lens is 164.3 mA as shown in FIG. As shown in d), the current range of the OS lens is 162.9 mA, and the range of change in the current value of the OS lens is the same -100 mA to 100 mA for all four times.
Note that (b) and (c) in the figure show the second time and the third time in the X direction, respectively.
The change of _Io of the time is shown. Thus, once the in-focus point in the X direction is obtained, it is possible to adjust whether or not the in-focus point in the Y direction is also in focus at I _o at that time. This is to check the state of astigmatism, and if the current difference between the two is within the allowable value, the average value of the two is also used as the in-focus current value, and when it exceeds the allowable value, the operator is notified as astigmatism. In the example of FIG. 4, after the focus adjustment in the X direction, as shown in (e) of the figure, the focus detection in the Y direction is completed once, and it is known that the astigmatism is within the allowable value. It

FIG. 5 shows an embodiment of the present invention. In the figure, 1 is an electron beam, 2 is a mark such as gold (Au) provided on a sample stage, 3'and 3 "are detectors of backscattered electrons (or secondary electrons), and 4 and 5 scan the beam. X, Y deflector for
6 ', 6 "are current-voltage converters, 7 is an adder, 8 is an analog-digital converter (A / D converter), 9, 10,
Reference numerals 16 and 17 denote digital-analog converters (D / A converters), 11 and 12 counters for providing deflection data, and 13
Is a latch memory for establishing input signal data, 14 is a focus lens O, 15 is another focus lens OS, 18, 19
Is a latch memory for temporarily storing data that gives the current values I _o and I _os of the O lens and the OS lens, 20 is a memory for storing a mark waveform signal obtained when Z _os is changed, for example, in 9 steps, and 21 is a waveform signal. A differential circuit for performing a sum-of-products operation for performing a second-order differential processing on
A maximum value detection circuit for extracting the maximum value of the second derivative signal Q, that is, the sharpness Q _i (i = 1 to m, m = 9 in this example), 23
Is a memory for temporarily storing Q _{1 to} Q _m , and 24 is a control circuit composed of a microcomputer, etc. (In this example, n = 3) is determined, Q _max is determined, and the OS coil current I _osmax at Q _max is determined.

B _{nk in} equation (1) is calculated using the method of least squares. (Here, r = 0 to n). Here, r = 0 to n First, the control circuit 24 causes the latch memories 18 and 19 to operate.
Data is set to 0 and the current value of the first OS lens is set. Subsequently, deflection data is given to 11 and 12, and the mark 3 is scanned by the electron beam 1 to obtain a waveform signal. Then, the current value of the OS lens for the second time is set, a waveform signal is obtained by the same method, and at the same time, the sharpness Q ₁ is obtained from the first time waveform signal by using each circuit of 21, 22, and 23. . As described above, the waveform signal is captured and the sharpness is extracted at the same time.

After changing the I _os value in 9 steps, using the equation (2), b ₃₀ ~
calculated coefficients b _33, to complete the equation (1). Next, the maximum value Q _{max of} sharpness and I _osmax are obtained using the equation (1). The current value I _oj of the O lens corresponding to this I _osmax is obtained and the data is transferred to 18 latch memories. At this time I
_{If the osmax} is in the vicinity of the intermediate value I _{os5 of} the I _os values I _{os1 to} I _os9 changed in 9 steps, the current value of the OS lens is set to a predetermined constant value I _osl because the focus detection is completed. On the other hand, if I _osmax is not near I _os5 , the in-focus point is not detected, the current value of the O lens is set to I _oj , and
I _os is changed in 9 steps, and a series of operations from the acquisition of the mark waveform signal to the extraction of Q _max and I _osmax at each step are repeated. When focus detection and focus adjustment in the X direction are completed, focus detection and adjustment in the Y direction are performed.

As can be seen from the above, the determination of in-focus detection is performed by | I _osmax −I _os5 | ≦ α, and this determination parameter is given in advance, and α = 5 mA is used in this embodiment.

〔The invention's effect〕

As described in detail above, according to the present invention, the focal length of the electron lens is changed in steps of m points, the out-of-focus state is obtained from the focus information extracted from the detection signal waveform at each step, and The quantity n using the method of least squares
By approximating to the following polynomial, obtaining the optimum position from this approximation formula, and feeding back the result to the electron lens to adjust the electron lens, stable focus detection and focus adjustment, and high resolution Can be implemented in. In particular, in the above case, when m ≧ 5 and n ≦ 3, stable and high-speed focus adjustment becomes possible.

Further, it is not necessary to employ a large number of OS current sample points at the time of focus detection, and the sample points necessary for determining an approximate polynomial are sufficient, and an optimum value can be obtained with an approximate curve without sampling a large number of points. Therefore, the processing time is greatly shortened. That is, according to the present invention, focus detection and focus adjustment can be performed in a short time.

As described above, according to the present invention, the electron beam or the like is deflected and scanned on the target through the electron lens, the reflected electrons, the secondary electrons or the transmitted electrons are detected, and the focus state of the electron beam is recognized from the detection signal. Is very effective when applied to an electron beam apparatus configured in, for example, an electron beam exposure apparatus or a scanning electron microscope, but its basic idea is not limited to these electron beam apparatuses, and for example, laser light, etc. It is also effective when applied to a charged particle beam device that uses focus detection or ion beams.

BRIEF DESCRIPTION OF THE DRAWINGS FIGS. 1, 2, 3, and 4 are diagrams for explaining the principle of the present invention, and FIG. 5 is a block diagram showing an embodiment of the present invention. 1 ... Electron beam, 2 ... Mark, 3 ', 3 "... Detector, 4,5 ... X, Y deflector, 6', 6" ... Current-voltage converter, 7 ... Adder , 8 ... A / D converter, 9, 1
0 ... D / A converter, 11, 12 ... Counter, 13 ...
... Latch memory, 14 ... Focus lens O, 15 ... Other focus lens OS, 16, 17 ... D / A converter, 18,
19 ... Latch memory, 20 ... Memory, 21 ... Differentiation circuit, 22 ... Maximum value detection circuit, 23 ... Memory, 24
...... Control circuit.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Internal reference number FI Technical display location H01L 21/027 H04N 5/232 A (72) Inventor Genya Matsuoka 1-280, Higashikoigakubo, Kokubunji, Tokyo Address: Central Research Laboratory, Hitachi, Ltd. (72) Masahide Okumura, 1-280, Higashikoigakubo, Kokubunji, Tokyo Metropolitan Research Laboratory, Hitachi, Ltd. (72) Susumu Kosasa, 1-280, Higashikoigakubo, Kokubunji, Tokyo Hitachi, Ltd. Central Research Laboratory (72) Inventor Kiichi Takamoto 1831 Ono, Atsugi-shi, Kanagawa Nippon Telegraph and Telephone Public Corporation Atsugi Telecommunications Research Laboratories (72) Inventor Tsuneo Okubo 1831, Ono, Atsugi-shi, Kanagawa Nippon Telegraph and Telephone Atsugi Electric Communication Research Laboratory (56) Reference JP-A-56-138926 (JP, A) JP-A-56-416 63 (JP, A) JP 53-128975 (JP, A) JP 56-7341 (JP, A)

Claims (1)

[Claims] 1. A structure for detecting a reflected electron, a secondary electron or a transmitted electron by deflecting and scanning an electron beam on a target through an electron lens, and recognizing a focus state of the electron beam from a detection signal thereof. In the electron beam apparatus, the electron lens is composed of first and second lenses, and the lens current value of the second lens is set in a state where the lens current of the first lens is set to a certain constant value. At m points (m ≧
5) is changed stepwise, the sharpness of the electron beam at each time is obtained from the detection signal at each step, and the lens current value of the second lens at this m point and the obtained sharpness value are obtained. Is used to curve-fit the relationship of the change in the sharpness value with respect to the change in the lens current value of the second lens to a polynomial curve of the n-th order (n ≦ 3) using the least squares method, and the obtained approximation Until the curve reaches the maximum value within the change range of the lens current value of the second lens, the first value
When the set value of the lens current of the second lens is changed stepwise and the above operation is repeated, the approximate curve shows the maximum value within the change range of the lens current value of the second lens. To obtain the lens current value of the second lens corresponding to the maximum value, and to realize the optimum focus state with the obtained lens current value of the second lens and the lens current value of the first lens at that time An automatic focusing method in an electron beam apparatus, which is characterized in that the optimum lens current value of