JPH04262351A - Electron beam device - Google Patents

Abstract

PURPOSE:To adjust astigmatism of an electron beam simply and with always stable accuracy by correcting the astigmatism on the basis of image signals. CONSTITUTION:An electron beam device 10 is broadly grouped into an optical system 20 and a circuitry system 40, wherein the optical system 20 is equipped within a mirror barrel 21 with an electron gun 22, optical lenses 23, 24, deflector 25, secondary electron sensor 26, objective lens 27, astigmatism correcting coil 28, and X-Y table 29 in the sequence as named. An electron beam 30 emitted by the gun 22 is converged and deflected and is led to the surface of a specimen (for example, semiconductor IC chip) 31 placed on the X-Y table 29. In the adjustment mode, the drive voltage for the coil 28 changes in sweeping with progress from the left to right end of a two-dimensional image (and from the top to bottom), and an image with the least astigmatism (the sharpest image) is observed in an appropriate region of the two-dimensional image. That is, the astigmatism of electron beam is corrected on the basis of image signals, so that adjustment of astigmatism can be made simply and with always stable accuracy.

Description

[Detailed description of the invention]

0001

[Industrial Field of Application] The present invention relates to an electron beam device, and more particularly, to a device for observing a two-dimensional image of a sample surface using an electron beam. Regarding.

In recent years, devices using electron beams have been put into practical use as devices for observing layout patterns of semiconductor integrated circuits.

[0003] This device irradiates the sample surface with an electron beam whose cross section is narrowed down to a minute circle, and the two
It captures secondary electrons (or reflected electrons) and reproduces a two-dimensional image, making it possible to obtain extremely precise images and is particularly effective in pattern analysis of large-scale integrated circuits with high integration density.

By the way, in such a device, an electron optical lens is used to focus the electron beam. Electron optical lenses, like normal optical lenses, have a so-called "astigmatism characteristic" in which the beam cross section becomes elliptical. To obtain an accurate circular cross section, adjustment for aberration correction (
Below, astigmatism adjustment) is essential.

[0005] Such adjustments are generally troublesome, and automation has been desired.

[0006]

2. Description of the Related Art Conventional astigmatism adjustment using, for example, an astigmatism correction coil is known.

FIG. 7 is a conceptual diagram thereof. The astigmatism correction coil is configured by arranging, for example, eight coils at every angle of 45 degrees around the optical axis of the electron beam. 0 degrees, 90 degrees, 18
Form the first group with four coils at 0 degrees and 270 degrees, and the remaining coils at 45 degrees, 135 degrees, 225 degrees and 31 degrees.
Form a second group of four coils of 5 degrees.

The first group includes the astigmatism control voltage Vst
x, and the second group has an astigmatic control voltage Vst
y is given. By adjusting Vstx, 45
An ellipse extending in a degree (or 225 degrees) or 315 degrees (or 135 degrees) direction can be corrected into a circular shape. Further, by adjusting Vsty, an ellipse extending in the 270 degree (or 90 degree) or 0 degree (or 180 degree) direction can be corrected into a circular shape.

That is, by using the astigmatism correction coils consisting of the first group and the second group, it is possible to correct ellipses in all directions and obtain accurate perfect circles.

0010

However, in such a conventional electron beam device, Vstx and Vs
Visually observe the image while changing ty appropriately, and check Vstx and Vs when the sharpest image is obtained.
Since the adjustment involved finding the value of ty, it was highly dependent on the skill of the individual, and there were problems such as the accuracy of the adjustment differed depending on the person, and furthermore, the adjustment was not easy and was extremely time-consuming. .

[0011] Accordingly, an object of the present invention is to perform astigmatism adjustment simply and always with stable accuracy.

0012

[Means for Solving the Problems] In order to achieve the above object, the present invention irradiates the surface of a sample with an electron beam whose cross-sectional shape is corrected by a magnetic field from a predetermined correction coil, and reflects the electron beam from the surface of the sample. In an electron beam device that forms an image signal of the sample surface based on electrons or secondary electrons,
It is characterized by performing astigmatism correction of the electron beam based on the image signal, preferably by analyzing the frequency spectrum included in the image signal, and preferably performing the astigmatism correction by analyzing the frequency spectrum included in the image signal. The method is characterized in that the image signal is analyzed while sweeping the voltage applied to the correction coil, and preferably, the time when the image signal containing the highest frequency spectrum is obtained is recognized as the minimum astigmatism adjustment point. shall be.

[0013]

[Operation] In the present invention, astigmatism of the electron beam is corrected in substantially the same procedure as in normal image observation of the surface of a sample. Therefore, astigmatism adjustment can be performed easily and always with stable accuracy.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be explained below based on the drawings. 1 to 6 are diagrams showing an embodiment of an electron beam device according to the present invention.

First, the configuration will be explained. In FIG. 1, the electron beam device 10 is roughly divided into an optical system 20 and a circuit system 40. The optical system 20 includes an electron gun 22, several optical lenses 23, 24, a deflector 25, , secondary electron detector 26, objective lens 27, astigmatism correction coil (hereinafter referred to as
A correction coil) 28, an XY table 29, etc. are sequentially arranged to focus and focus the electron beam 30 from the electron gun 22.
It is deflected and guided to the surface of a sample (for example, a semiconductor integrated circuit chip) 31 placed on an XY table 29.

On the other hand, the circuit system 40 has deflection voltages Vx, Vy
, a control circuit 41 that generates astigmatism control voltages Vstx, Vsty, and focus control voltage Vf, and a first drive circuit 42 that amplifies Vx and Vy to drive the deflector 25. The switch 43 is turned ON in the aberration adjustment mode, and the Vstx and Vsty switches are turned ON in the measurement mode.
is passed through as is, and in adjustment mode Vstx, Vst
The voltage obtained by adding Vx and Vy to y (Vx+Vstx, Vy
+Vsty);
A second drive circuit 45 that amplifies the output of Vf to drive the correction coil 28, a third drive circuit 46 that amplifies Vf and drives the objective lens 27, and converts the output signal of the secondary electron detector 26 into a digital signal. An analog/digital converter (ADC) 47 for converting, a frame memory 48 for developing this digital signal into a two-dimensional image, a display device 49 for visually displaying the two-dimensional image, and an image signal for each region of the two-dimensional image. and an evaluation circuit 50 that analyzes and outputs a parameter (details will be described later) representing the degree of astigmatism of the electron beam.

FIG. 2 is a diagram showing a preferred example of the evaluation circuit. In FIG. 2, 50a is a memory access circuit;
0b is FFT (Fast Fourier Trans
50c is a calculation circuit. The memory access circuit 50a receives addresses Ai,j (i,j are area numbers) that specify each area of the two-dimensional image according to the area size designation signal ΔP from the control circuit 41.
are sequentially generated, and image signals Di,j for each area are transferred to the FFT circuit 50b. The FFT circuit 50b performs n×n fast Fourier transform processing on the transferred image signal Di,j to decompose it into spatial frequency spectral components fspu, and the calculation circuit 50c calculates the highest frequency spectral component fspu for each region. Calculate. Note that the memory M of the control circuit 41 stores fspu for each area.
Among them, the maximum fspu (hereinafter referred to as fspumax) is retained.

Next, the operation will be explained. FIG. 3 is a diagram of various voltage waveforms in the adjustment mode in the above system. Vx
is a scanning voltage in the line direction, and Vy is a scanning voltage in the frame direction. That is, by sweeping Vx while increasing Vy stepwise, the surface of the sample 31 can be scanned two-dimensionally. These Vx and Vy have the same waveforms as in the normal measurement mode, but the following Vstx, Vsty and Vf are different from those in the measurement mode.

In other words, Vstx and Vs in the adjustment mode
ty is the size in which Vx and Vy are superimposed (Vs
tx=Vstx+Vx, Vsty=Vsty+Vy), from the minimum Vstx to the maximum Vstx,
Similarly, the Vsty varies from the minimum Vsty to the maximum Vsty. This means that Vstx
, Vsty has been swept. Further, Vf is increased stepwise by ΔVf for each frame.

[0020] Here, FBT represents a period between frames, and this FBT corresponds to a blanking period (flyback time) in raster scanning. In this embodiment, the evaluation process described below is executed using the FBT period.

FIG. 4 is a flowchart of adjustment mode processing including the evaluation processing. In this flow diagram, step 60
to 64 are evaluation processes, and the evaluation processes are repeated a predetermined number of times (specifically, the number of frames forming a two-dimensional image of the sample 31).

The evaluation process is first started by taking in one frame's worth of image signals from a two-dimensional image (hereinafter also referred to as SEM image) held in the frame memory 48. Here, FIG. 5 is an example of a SEM image, and i
A square frame whose two sides are the axis (column) and the j-axis (row) represents the scanning range of the electron beam. After scanning i1 to i6 in the first row (j1), scanning i1 to i6 in the second row (j2)
This is repeated until the fifth row (j5). Note that the maximum values of i and j are i=6 and j=6 in the figure, respectively.
Although the number is 5, this number is for the purpose of simplifying the explanation, and the actual number is much higher.

During the measurement mode, the driving voltages applied to the correction coil 28 are constant values Vstx and Vsty. On the other hand, during the adjustment mode, the drive voltages applied to the correction coil 28 are Vx+Vstx and Vy+Vsty, respectively.
becomes. This is because in the adjustment mode, switch 43 is turned on and Vx and Vy are superimposed by the adder circuit.

Therefore, in the adjustment mode, as the SEM image progresses from the left end to the right end (and from the top end to the bottom end),
The driving voltage to the correction coil 28 changes in a sweeping manner, and more specifically, changes greatly in accordance with the changes in Vx and Vy, resulting in an image with the least astigmatism (in other words) in an appropriate area in the SEM image. the sharpest image) is observed.

In this embodiment, the SEM image is divided into grid shapes (regions surrounded by broken lines in FIG. 5), and the amount of astigmatism is evaluated based on the image signal of each region. That is, in step 61 of the flow diagram, first, the values of i and j are set as i1,
j1 and one area (i1,
FFT processing the image signal of j1) (step 61a),
The upper limit value fspu of the spatial frequency is determined from the processing result (step 61b). Here, the spatial frequency is a spectral distribution of frequencies forming a regional image signal, and the sharper the focus of the image, the higher the frequency component.

[0026] Therefore, by extracting the maximum upper limit value fspu for several area images constituting one frame, an image area with less astigmatism can be specified. Step 62 is a step for performing such identification, in which f
The maximum value of spu (fspumax) is determined, and the number (imax, jmax) of the area where the maximum value is obtained is held.

[0027] Then, in step 63, (imax,j
The value of (Vstx, Vsty) corresponding to ).

That is, in FIG. 3, in the adjustment mode, (1) sweeps Vstx and Vsty in one frame to find the area with the least astigmatism from among several area images that make up that frame; area fsp
umax, and Vst while scanning the area
x, Vsty are held, then (2) Vf is increased by ΔVf, and the above process (2) is executed for the next frame. This process is repeated until the final frame.

The fspumax collected for each frame is stored in the memory M of the control device 41. During this storage, it is compared with the old value in the memory M, and if fs
Only pumax is stored in memory M. For example, as shown in FIG.
, 2), the area (5, 2)
is higher, so in this case, fspu of area (5,2)
is held in memory M as fspumax. Therefore, the maximum fspumax is always held in this memory M, and as a result, the fspumax of the entire ESM image is
The maximum value of x is determined, and Vf, Vstx, and Vsty corresponding to the maximum value fspumax are determined as appropriate astigmatism correction parameters (step 66).

As described above, in this embodiment, (1) the driving voltage applied to the correction coil 28 is swept, (2) the SEM image is divided into a plurality of regions, and (3) the most Since a region with little astigmatism is specified, (4) an appropriate adjustment voltage can be automatically determined from the drive voltage when scanning that region.

[0031] Therefore, the adjustment work can be facilitated and simplified;
In addition, it is possible to obtain the effect that astigmatism adjustment can always be performed with stable accuracy.

[0032]

According to the present invention, an electron beam whose cross-sectional shape is corrected by a magnetic field from a correction coil is irradiated onto a sample surface, and the shape of the sample surface is determined based on reflected electrons or secondary electrons from the sample surface. In the electron beam device that forms an image signal, the astigmatism of the electron beam is corrected based on the image signal, so astigmatism adjustment can be performed simply and always with stable accuracy.

[Brief explanation of the drawing]

FIG. 1 is a system configuration diagram of an embodiment.

FIG. 2 is a configuration diagram of an evaluation circuit according to an embodiment.

FIG. 3 is a voltage waveform diagram of one embodiment.

FIG. 4 is an operational flow diagram of one embodiment.

FIG. 5 is a diagram of a SEM image of one example.

FIG. 6 is a spatial frequency diagram of one embodiment.

FIG. 7 is a configuration diagram of a correction coil.

[Explanation of symbols]

30: Electron beam 31: Sample 43: Switch 44: Addition circuit 50: Evaluation circuit

Claims (4)

[Claims] 1. A sample surface is irradiated with an electron beam whose cross-sectional shape is corrected by a magnetic field from a predetermined correction coil, and an image signal of the sample surface is formed based on reflected electrons or secondary electrons from the sample surface. What is claimed is: 1. An electron beam device that performs astigmatism correction of the electron beam based on the image signal. 2. The electron beam apparatus according to claim 1, wherein astigmatism correction is performed by analyzing a frequency spectrum included in the image signal. 3. The electron beam apparatus according to claim 1, wherein the image signal is analyzed while sweeping the voltage applied to the correction coil. 4. The electron beam apparatus according to claim 2, wherein the time when the image signal containing the highest frequency spectrum is obtained is recognized as the minimum astigmatism adjustment point.