JPH0334646B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides a shape having a shape such as a trapezoid or a barrel shape in which the distance between two opposing sides changes continuously, or a shape having a constriction or a protrusion on at least one of the two opposing sides. , relates to an electron beam exposure method suitable for drawing figures in which the opposing distance between two sides changes locally.

When drawing a figure, for example a trapezoidal figure, on a sample such as a semiconductor substrate by electron beam exposure, the cross-sectional shape of the electron beam irradiated onto the sample is formed into a rectangular shape, and its width is set to a constant value of about 0.5 μm. On the other hand, a drawing method is adopted in which the electron beam is moved in a direction perpendicular to the longitudinal direction at a pitch smaller than the above-mentioned width while changing the longitudinal dimension. In addition,
To change the cross-sectional dimension of the electron beam, for example, two slit plates with polygonal holes are placed on the path of the electron beam, and a deflector is placed between the slit plates to change the deflection force of the deflector. A method is adopted in which the irradiation position of the electron beam passing through the hole in the slit plate located above is irradiated onto the slit plate located below by control.

In this way, with the drawing method in which the electron beam is moved in a direction perpendicular to the longitudinal direction at a pitch smaller than the width, most of the area to be drawn is exposed multiple times to the electron beam with a minute width. Therefore, it is considered that the dimensional accuracy in the width direction does not need to be very accurate. That is, in the case of such drawing, the current situation is that little consideration is given to the width of the electron beam.

However, the electron beam current density distribution of an electron beam is not strictly uniform; in the case of an electron beam having a rectangular cross-sectional shape, the electron beam becomes smaller at its edges, and has a deviation from the ideal electron beam current density distribution. Due to this sag, the electron beam current amount is reduced from the ideal electron beam current amount, but the rate of reduction is much larger in the width direction than in the longitudinal direction of the electron beam, which has a rectangular cross-sectional shape. Therefore, particularly high accuracy is required in setting the dimensions not only in the longitudinal direction but also in the width direction. If this setting accuracy is low, the irradiation electron dose will vary from the appropriate irradiation electron dose. On the other hand, unlike shapes such as trapezoids, rectangular shapes that can be drawn by moving electron beams with appropriate rectangular or square cross sections without overlapping each other, There is no need to change the longitudinal dimension of the electron beam cross section for each irradiation, and there is no need to set the electron beam width to a very small width and draw it by multiple irradiations. There are few factors that change the dose.

Therefore, it has been found that when drawing a figure in which a figure such as a trapezoid and a rectangular figure coexist, exposure unevenness occurs due to the difference in exposure amount when drawing both figures. By the way, although the cross-sectional dimension variable method described above allows the cross-sectional dimension (longitudinal dimension) of the electron beam to be changed, the positional accuracy of the two slit plates for making the cross-sectional shape of the electron beam into a rectangular shape is difficult. It has been difficult to accurately set the width of the cross section of the electron beam due to factors such as the voltage controllability of the deflector that determines the width of the electron beam, and the difficulty in making the slit at right angles.

For example, if the width that should be set to 0.5 μm is set to 0.6 μm, when drawing is performed by moving an electron beam with a minute width at a certain pitch in a direction perpendicular to the longitudinal direction, the electron beam exposure amount will be approximately This increases by as much as 20%, resulting in a deviation from the appropriate exposure amount and uneven exposure.

The present invention moves an electron beam having a rectangular cross-sectional shape in a direction perpendicular to its longitudinal direction, and
This was done in view of the above-mentioned inconveniences that occurred when drawing figures that could be drawn by changing the longitudinal dimension during the movement process. The present invention is characterized in that the dimension of the electron beam in the short side (width) direction is controlled.

The drawing accuracy of shapes such as trapezoids or triangles that require the electron beam to move in a fixed direction and change the longitudinal dimensions of the electron beam during the movement process is as follows: Although it is greatly influenced by the dimensional accuracy in the longitudinal direction of
It is hardly affected by the dimensional accuracy in the width direction. By the way, the dimensional accuracy in the longitudinal direction of an electron beam having a rectangular cross section can be accurately controlled by the cross-sectional dimension variable method described above. Therefore, as long as we focus only on drawing accuracy, The desired drawing accuracy can be ensured by the known cross-sectional dimension variable method.

On the other hand, as described above, the dimensional accuracy in the width direction, which has little effect on the drawing accuracy itself, will have a significant effect on exposure unevenness if it is not set accurately.

In the electron beam exposure method of the present invention, first, the dimension in the long side direction of the electron beam cross section and the dimension in the moving direction of the electron beam are set to the same first dimension by a cross-sectional dimension variable method. A first electron beam current value is determined by measuring the electron beam current. Next, the dimension of the rectangular cross section in the long side direction of the electron beam is fixed to a first dimension, and the dimension of the rectangular cross section in the electron beam movement direction is set to a second dimension shorter than the first dimension. The electron beam current is set to Find the reference value. After this, the dimension of the cross section of the rectangular shape in the electron beam movement direction is adjusted to a third value so that the second electron beam current value obtained by actually measuring the electron beam current is equal to the reference value of the electron beam current. This third dimension is set as the width in the direction of electron beam movement and exposure is performed to draw a target figure. According to the electron beam exposure method of the present invention, it becomes possible to draw figures with high drawing accuracy without exposure unevenness.

FIG. 1 is a schematic diagram of an electron beam exposure apparatus using the exposure method of the present invention. In the figure, reference numeral 1 denotes an electron gun, and an electron beam from the electron gun is irradiated downward through a lens 2. 3 is an electron beam cross-sectional dimension (and shape) variable means, which includes slit plates 4 and 5 having holes, a lens 6 disposed between these two slit plates, and a deflector 7.
Consists of. The deflector 7 actually consists of one for the X direction and one for the Y direction, and since an electron beam cross-sectional dimension designation signal is sent from the electronic computer 8 via the amplifier 9, the electron beam passes through the hole in the upper slit plate 4. The imaging position of the electron beam on the lower slit plate 5 is controlled by a deflection force according to the electron beam cross section designation signal,
From the hole in the lower slit plate 5, an electron beam having a cross section of a size corresponding to the electron beam cross section size designation signal is generated. This electron beam is projected by a projection lens 6 onto a sample (not shown) such as a semiconductor substrate or a mask, and at the same time is deflected by a pattern drawing position signal sent from the electronic computer 8 via an amplifier 10. The sample is scanned by the deflection force of the device 11, and a desired figure is drawn on the sample (not shown). Reference numeral 12 denotes a Faraday cup which is arranged to be replaceable with the sample in order to collect the electron beam from the hole in the lower slit plate 5, and is arranged on the optical axis in place of the sample during non-graphic drawing. The current of the electron beam collected by the Faraday cup 12 is transmitted to the amplifier 1.
3 to the electronic computer 8.

The case of drawing a trapezoid A as shown in FIG. 2 using such an apparatus will be described. First, the Faraday cup 12 is placed at the sample position instead of the sample. Next, the cross section of the electron beam from the electronic computer 8 is shown as B in FIG. 2, and the dimensions in the X direction and Y direction are x ₁ ,
Amplifier 9 sends a designated signal with the same dimensions as y ₁ (=x ₁ ).
is sent to the deflector 7 via the. By the way, when specifying the dimensions x and y of the electron beam cross section, the following considerations should be taken.

To draw a trapezoidal pattern, first set x 1 to be equal to the longitudinal direction (X direction) dimension of the first rectangular cross section for drawing the trapezoidal pattern, and then set x ₁ to the same dimension as this dimension x ₁ . The dimension _y1 in the width direction (Y direction) is set so that the electron beam has a square cross section. This dimension x ₁ and dimension
Let y ₁ be the first dimension.

Then, this electron beam is collected by a Faraday cup 12, and this electron beam current is sent to an electronic computer 8 via an amplifier 13. The electronic computer 8 stores this first electron beam current value _Is in its internal memory. next,
Dimensions of the first electron beam cross section C for drawing trapezoid A (the dimension in the X direction is the first dimension x ₁ , the dimension in the Y direction is the second dimension y _c , where y _c ≪ x ₁ ) is sent from the electronic computer 8 to the deflector 7 via the amplifier 9, and the electron beam current I ₁ (second
(electron beam current) is sent to the electronic calculator 8 via the far day cup 12 and amplifier 13. During this period, the electronic computer 8 has already multiplied the first electron beam current I _s by the ratio value of the second dimension y _c and the first dimension y ₁ (=x ₁ ) (I _s ×y _c /y ₁ ) and store this result in the memory as a reference value.Compare the sent second electron beam current _I1 with this reference value and make sure that the difference between the two values becomes zero. Then, the signal corresponding to the error is added to (or subtracted from) the Y-direction dimension designation signal to adjust the second Y-direction dimension y _c to obtain an appropriate Y-direction third dimension y _c Set ′ accurately. Next, the Faraday cup 12 and the sample are replaced, and trapezoid A is drawn. This method of drawing involves changing the dimension of the electron beam cross section in the X direction (the dimension in the Y direction is the adjusted third dimension y _c )
Move it little by little in the direction of arrow Q from the top.
The dimension in the X direction is changed by the deflector 7 of the electron beam cross-sectional dimension (and shape) varying means 3 according to the command from the computer 8.
The electron beam is moved by a deflector 11 according to instructions from an electronic computer 8.

When drawing a trapezoid inclined at 90 degrees with respect to the trapezoid A as shown in FIG. 2, drawing is performed as described above while keeping the dimension of the electron beam cross section in the X direction constant and varying the dimension in the Y direction.

The drawing accuracy of the trapezoid drawn as described above is extremely high because the dimensional change in the X direction of the electron beam cross section is accurately controlled, and also because of the dimensional change in the X direction due to movement in the Q direction. Also, since a control operation is added to adjust the amount of electron beam irradiation to an appropriate amount by controlling the dimension in the Y direction, uneven exposure can be reliably eliminated.

As is clear from the above explanation, the electron beam exposure method of the present invention uses an electron beam having a width having a rectangular cross section so that the electron beam current amount is the same as the electron beam current amount due to the ideal electron beam current density distribution. This method controls the dimension in the direction (Y direction) and draws figures such as trapezoids with an electron beam having the controlled width dimension. It is now possible to draw figures with high drawing precision without exposure unevenness, for which it has been difficult to prevent the occurrence of unevenness.

In the above explanation, a trapezoid is drawn as an example, but drawings with a rectangular cross section, such as barrel shapes, triangles, rhombuses, or drawings with a constriction or protrusion on at least one of two opposing sides, Of course, it is also possible to scan with an electron beam and to draw figures that require changes in dimensions during the scanning process.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the configuration of an exposure apparatus used in the electron beam exposure method of the present invention, and FIG. 2 is a diagram specifically explaining the trapezoid drawing method. 1... Electron gun, 2, 6... Lens, 3... Electron beam cross-sectional dimension variable means, 4, 5... Slit plate,
7,11...deflector, 8...electronic computer, 9,10,
13...Amplifier, 12...Fuaraday cup.

Claims (1)

[Claims] 1. An electron beam with a rectangular cross section is moved over the sample in a direction perpendicular to the long side of the cross section, and in the process of movement, the dimension of the long side of the cross section is changed to draw a predetermined figure. First, the dimension in the long side direction of the electron beam cross section and the dimension in the moving direction of the electron beam are set to the same first dimension, and the electron beam current at the same setting is actually measured to determine the first electron beam current. Then, the dimension of the cross section of the rectangular shape in the long side direction of the electron beam is fixed to the first dimension, and the dimension of the cross section of the rectangular shape in the direction of electron beam movement is set to the first dimension. Set to a shorter second dimension, and multiply the ratio value of the second dimension and the first dimension of the electron beam of the rectangular cross section by the actually measured first electron beam current value. After determining the reference value of the electron beam current, a second electron beam current value obtained by actually measuring the dimension of the cross section of the rectangular shape in the electron beam movement direction is the reference value of the electron beam current. An electron beam exposure method, characterized in that a third dimension is set to be equal to the width of the electron beam, and the exposure is performed using the third dimension as the width in the electron beam movement direction.