JPS5821581A - Measurement of electron beam diameter - Google Patents

Abstract

PURPOSE:To enable a highly accurate measurement of the electron beam diameter by forming a sharp edge of a metal marker in such a manner that the thickness thereof is less than one-tenth as large as the beam diameter to be measured. CONSTITUTION:A sharp edge 1a is scanned with an electron beam Ei and reflected electrons Er thus obtained are detected. Based on the detection signals, the diameter of the electron beam Ei is measured. In this case, said sharp edge 1a is made up of a metal marker 1 in such a manner that the thickness thereof is less than one-tenth of the beam diameter to be measured. This enables the evaluation of the brightness of an electron gun at a high accuracy thereby avoiding uneven exposure, variations and the like in the describing by electron beam.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in an electron beam diameter measuring method for measuring the electron beam diameter by the Sharp Edge method.

Recently, mold IIA/
An electron beam lithography system is used to form the pattern, and the brightness of the electron gun and resist sensitivity are important factors that affect the line writing time in this system. The brightness β is expressed by the following equation, where the beam current at the drawing surface is 11% and the beam diameter is -.

β@CII/纏 --(1) As can be seen from the first equation, in order to accurately evaluate 1 luminance β, it is necessary to accurately measure the beam diameter d.

Conventionally, a wire method, a sharp edge method, etc. have been used to measure the diameter of an electron beam.

These scan a thin wire or sharp edge with an electron beam, and calculate C based on the reflected electron signal obtained at this time.
- It measures the diameter of the lumen (To), and the above measurement can be performed relatively easily.

However, each of these types of methods has the following problems. In other words, since there are limits to the thinness of the wire that can be obtained and the smoothness of one mating surface in the wire method, when V-^II is less than 10 (7111k), especially less than 1 [male], the measurement accuracy is at a disadvantage. Invite a decline. Further, in the sharp edge method, a sharp edge is formed mainly by utilizing the opening of a silicon single crystal. This takes advantage of the fact that the bone opening plane of a silicon single crystal is the (1113 plane).For example, if a silicon single crystal with a (100) plane is expanded, the (111) plane will be in an undercut state. Because it appears, the cut is clogged (1001mt(1111m)
m to clt angle #64 degrees and sharp and 1k), the mackerel sharp A ySp will be formed. However, it is difficult to say that forming shears/edges by the above-mentioned expansion and opening is easy, and it is not always possible to obtain a sharp edge with just one OgI opening. In addition, there are problems such as the sharp edges formed due to expansion deteriorate during use and the edges become rounded, resulting in a decrease in the accuracy of the V-me field determination described above. Ta.

Therefore, as a modification of the sharp edge method, a round shape is being considered in which the sharp edge is formed with a gold marker. The gold marker has a high electron reflectance and does not deteriorate, making it extremely effective for measuring beam diameter. However, when using this gold marker, the following problems arise. That is, there are the gold i-force method and the vapor deposition method, but no matter which method is used, the edge portion is rounded with a radius of curvature corresponding to the thickness of the gold marker. This edge O roundness becomes a factor that reduces the measurement accuracy of the beam aia+, especially the measurement accuracy of a narrow beam diameter.

The present invention has been made in consideration of the above-mentioned circumstances, and its purpose is to develop and develop a method for measuring the diameter of an electron beam using the Sharp Edge method. It is in.

First, the present invention will be explained in detail. The decrease in the accuracy of the beam radius measurement is due to the roundness that occurs in the edges.

Eliminating roundness and reducing the radius of curvature to zero! -It is impossible due to the manufacturing method of power. Therefore, in order to reduce the roundness of the edge, K may be reduced by reducing the thickness of the marker and reducing the radius of curvature of the edge, which is approximately equal to the thickness. but,! - If the force is applied thinly, the difference in reflected electron rate between the marker and the sample substrate on which the acid marker is formed becomes small, and the sensitivity of the -me measurement decreases considerably. That is,
! - If the force is made thin, the beam measurement sensitivity will be reduced, and conversely, if the marker is made thick, the beam measurement accuracy will be reduced. Taking this point into consideration, the inventors of the present invention have conducted extensive research, and the present invention focuses on this point in that it is possible to reduce the thickness of the marker to 1/10 of the beam diameter, which should be constant, without making the thickness of the marker significantly thinner. In this method, the sharp edge used for measuring the electron beam diameter is formed from a metal marker, and the thickness thereof is 1/10 or less of the beam diameter to be measured. Furthermore, according to the present invention, the diameter of an electron beam can be measured with high precision using the Sharp Edge method, which is extremely effective for evaluating the electron gun brightness of an electron beam exposure apparatus.

Hereinafter, details of the present invention will be explained with reference to the drawings.

FIG. 1 is a schematic diagram showing mark portions and beam strong motion distribution according to the present invention. @Naka 1 Dukin i-force% xa The underlying conductive substrate is, for example, silicon single crystal. Money! - The edge portion 1a of the force 1 is formed with a rounded radius of curvature equal to the thickness RK#A of the marker 1. J is the marker 1 as a result of the incident electron group being applied to the marker 1 or the Hiro substrate 2.
This is a backscattered electron detector that detects backscattered electrons 12r reflected from the substrate 2.

The incident electrons gi are usually circular, and their intensity distribution is a Gaussian distribution (standard deviation σ). Now, if the distance from the origin 0 to the beam center is x1 and the angle between the specular electron reflection direction and the backscattered electron detection direction seen from the incident point is F←), then the signal strength II detected by the backscattered electron detector 3 is a ma F
Proportional to (b).

When the incident electron It is scanned one cycle from left to right on the paper, it is calculated by the following equation as a function of the change x of the electron beam entering the backscattered electron detector JK.

I(d-/N(d・awF6c)・strike=”(2) However, This is the intensity distribution.The calculation result of the above-mentioned formula 2 is shown in Figure 2], and in this case!-K1
Thickness R-0,3 (ratio of reflectivity between deflection force 1 and substrate 2: 81 m3:1, '9, standard deviation C of beam intensity distribution / mother) As a meter], song il
l P * is 1 mO, 4 (Peng), -sP Mayuzumi is -0, 2
(trend), curve PsB y -0,1CM&), - line P4
is the case of −0,04 (trend). Since the true beam diameter d is defined as L8@#, the above curve PI ePl
The true beam diameter of eP@ml'a is 1 (#s
), 0.5 [Island], 0, 26 (Elegance), 0.1 (#l
), the nine-point calculation curve shown in Figure 2 agrees well with the experimental results.

The beam diameter can be determined from a measurement curve similar to that shown in FIG. 2 as follows. That is, let H be the difference between the maximum level and the minimum level of the backscattered electron detector IO detection signal, and the difference between the maximum level and Q, IH smaller than the minimum level is 0.1.
The width in the scanning direction between the large point H is measured as the beam diameter.

The beam diameter obtained in this way and the true C-mu diameter d-'2
．． ! Figure 3 shows the relationship with 5ell. −
The 0 curve Qt where the thickness 8 of force 1 is /#2 meters is l-0
, 1C island), the curve Qs is R-0,3 [edition], - line Q error is 凰-0,5 [style], song 1! Q a hiro R-1,0 (trend)
This is the case. Further, this is a case where the linear measurement beam diameter shown by a broken line is equal to the true beam diameter, that is, the R-00 case. As can be seen from FIG. 3, the smaller the thickness R of the marker 1, the more accurately the beam can be evaluated. When the beam diameter is 1, it is estimated almost accurately that Fi curve is 1 and shimmer force is 1.
When the thickness 8 of the marker 1 is -0.1 (am) or less, the measurement error becomes large when the thickness R of the marker 1 is 0.1 (JIIIL) or more. Furthermore, although not shown in the figure, when the beam diameter was 10 (MK), the expansion curve Q4 approached a straight line, and the intensity of the beam was negligible. According to the calculation results shown in Figure 3, when the C-mu diameter is large, the beam diameter is measured to be thinner than the true beam diameter, and when the beam diameter is below a certain level, it is evaluated to be thicker than the true beam diameter. K becomes. However, the boundary ore marker 1 of the beam diameter
It differs depending on the thickness K.

Fig. 4 shows the results of an experiment with 1000 (X) and 5000 (1), which was measured using an #10 electronic C-me drawing device.

The vertical axis is the measured beam diameter, and the horizontal axis is the diameter of the second condenser lens in the electron optical lens barrel of the above device. The larger the mathematics in 1 scale, the smaller the beam diameter * 1llill
llt e 8 hiro! -Force 1 (thickness “gR cut tL
-t'tL1000 (1-) Th! Hi 5000 (D
The case is shown below. As can be seen from Figure 4, when the measured beam diameter reaches 0.3C, for beams thicker than this! - Force thickness 5000 (t) o side * t 000
CL') O%t) Yo11IA<na], 0, 3 (71
11&) and above are Hirogyaku. Therefore, the results of the theoretical calculations shown in FIG. 2 are in good agreement with the experimental results, and are thus found to be correct.

And ζrode, money! - There are various manufacturing methods for KIII that forms force 1, but ideally the radius of curvature is 8 in Figure 1.
It is very difficult to make the thickness RK zero. As mentioned above, the gold marker 1 usually has a radius of curvature that is comparable to the thickness RK.

Therefore, from a practical standpoint and from the theoretical calculations and experimental results mentioned above, for example, 1 [μ
To accurately determine the beam diameter of the beam, the thickness is 0.1.
It turns out that it is sufficient to use the gold marker (ss). Furthermore, these results show that accurate measurements can be made as long as the thickness of marker 1 is 1/10 or less of the beam diameter to be measured. Thus, according to the present invention, it is possible to evaluate the electron gun brightness with high precision, and it is possible to avoid wobbling during exposure unevenness in electron beam drawing.

Although gold is used as the material for the marker, any metal with high electron reflectance such as platinum, aluminum, silver, or nickel may be used instead of gold. Moreover, it is of course applicable not only to the electron beam drawing apparatus but also to the electron beam drawing apparatus of each building. In addition, various modifications can be made without departing from the gist of the present invention.

[Brief explanation of drawings]

Figures 1 to 4 are for explaining the present invention in detail, respectively. Figure 1 is a schematic diagram showing the marker part and beam intensity distribution, and Figure 2 shows the calculation results of the backscattered electron intensity. Figure 3 is a diagram showing the measurement beam diameter with respect to the true beam liK by adjusting the marker thickness by 1/7, and Figure 4 is a diagram showing the experimental result. 1... Money! -Force, J I@@Substrate, S... Backscattered electron detector, 1a/-2y section. Applicant's agent Patent attorney Takehiko Suzue Figure 1 Figure 2 Figure 2 9R4 Figure 23456'

Claims (1)

[Claims] The reflected electrons obtained by scanning the electron beam are detected, and based on this detection signal, the electron C-
1. A method for measuring the diameter of an electron beam, characterized in that the sharp edge is formed of a metal marker, and its thickness is 1710 mm or less of the diameter of the beam to be measured.