Radiant energy – Irradiation of objects or material – Irradiation of semiconductor devices
Reexamination Certificate
2000-02-07
2002-06-11
Lee, John R. (Department: 2881)
Radiant energy
Irradiation of objects or material
Irradiation of semiconductor devices
C250S492200
Reexamination Certificate
active
06403973
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to an electron beam exposure method for performing exposure while a sample table is being moved continuously and an apparatus and a device using the same, and particularly to an electron beam exposure method and apparatus capable of preventing degradation in the accuracy of exposure due to glitch noise of a DA converter and reducing the number of trace region pass-changeovers, and a device manufactured using the same.
As one of methods for implementing high throughput of an electron beam exposure apparatus may be mentioned continuous shifting exposure for performing exposure while a sample table is being moved continuously. Trace control for first detecting the amount of movement of the sample table and next feeding back the amount of movement thereof for the purposes of the deflection of an electron beam in real time is essential to perform the continuous shifting exposure. For example, Continuous writing method for high is speed electron-beam direct writing system HL-800D (J.Vac.Sic.Technol.B11(
6
)) is disclosed as an example of the trace control, which has been known to date.
FIG. 3
is a diagram for describing the trace control method employed in the above-described conventional example.
The operation of
FIG. 3
will be explained. Exposure is effected on a sample
11
placed on a sample table
12
by means of an electron beam
10
while the sample table
12
is being moved continuously. First, a laser interferometer
2
measures the coordinates of the sample table
12
with the exposure sample
11
placed thereon. A trace signal calculation unit
6
calculates the amount of deflection equivalent to displacements of the coordinates of the sample table
12
and coordinates to apply the electron beam, i.e. trace deflection data. A pattern generator
1
is capable of generating various device's pattern data inclusive of the sample
11
. Exposure deflection data outputted from the pattern generator
1
and trace deflection data outputted from the trace signal calculation unit
6
are respectively converted to analog values by DA converters
7
and
8
. These two analog values are added together by an analog adder
9
, followed by input to a deflector
5
. Thus, the deflector
5
deflects the electron beam
10
to a desired position on the exposure sample
11
.
A cycle for updating trace deflection data calculated by the trace signal calculation unit
6
is determined from the transfer speed of the sample table
12
and required exposure accuracy. Namely, the updating cycle becomes short as the transfer speed of the sample table
12
becomes fast and the exposure accuracy becomes high. In the conventional example, the trace signal or trace deflection data is updated in a cycle of 100ns (10MHz). Since the DA converter
7
used for exposure deflection and the DA converter
8
used for trace deflection are different in required property, DA converters
7
and
8
dedicated to them are provided. Namely, the DA converter
7
used for exposure deflection is activated in synchronism with applying (shot) timing of the electron beam and the electron beam is applied after the output thereof has been settled. Thus, since the settling time of the DA converter
7
results in exposure wasting time, its responsivity must be made fast. On the other hand, the DA converter
8
used for trace deflection needs to update data in a cycle shorter than the DA converter
7
, i.e., during the application of the electron beam. Therefore, DA converters small in glitch noise produced upon changes in the outputs of the DA converters are used.
When the amount of deflection based on the trace deflection exceeds a predetermined value, a trace region pass-changeover operation for adding the amount of the trace deflection to the amount of deflection of a high-level deflector
23
and restoring the amount of the trace deflection to the initial value is performed. Namely, when the amount of the deflection exceeds the predetermined value since the high-level deflector
23
shown in
FIG. 3
has a deflection range larger than that of the low-level deflector
5
, the trace deflection data is transferred from the trace signal calculation unit
6
to a DA converter
22
for the deflector
23
through the pattern generator
1
, and the value calculated by the trace signal calculation unit
6
is returned to 0.
Since the amount of movement of the sample table
12
can be fed back for the deflection of the electron beam
10
in real time according to the above-described exposure method, exposure can be implemented at high speed and with high accuracy while the sample table
12
is being moved continuously.
As described above, the DA converter is used for the deflection of the electron beam in the electron beam exposure apparatus. The DA converter is commonly comprised of a plurality of current sources different in weight. There are provided switches for turning on and off the respective current sources. Necessary values are selected from the plurality of current sources and the selected outputs are added together and thereafter outputted, whereby an arbitrary large output can be obtained. Since, however, the respective switches vary in operating time, a current having an unintentional magnitude might be outputted when data is switched over to another. This is called “glitch noise”. The magnitude of the glitch noise is determined depending on the number of the activated switches and the sizes of the current sources switched thereby. Thus, the maximum glitch noise is produced in the case of a
{fraction (1/2+L )} full scale at which all the bits of input data are inverted. If a full scale
15 is represented in 4 bits based on a binary number (four switches), for example, then the maximum value results in ‘1111’ (15 represented in a decimal number) and the minimum value results in ‘0000’ (0 similarly). Since the
{fraction (1/2+L )} full scale is set between
7 and 8, all the bits are interchanged at the time of a changeover from ‘0111’ to ‘1000’, so that switching timings of all the switches are mismatched with one another, thus leading to the generation of the maximum glitch noise.
Since the generation of the glitch noise is a problem about the structure of each DA converter in this way, it is very difficult to reduce the glitch noise. Although the glitch noise can be reduced by inserting a filter, a delay in response is produced so that it could not lead to practical applications.
The deflection of the electron beam by the electron beam exposure apparatus of the sample table continuous shifting system can be roughly divided into exposure deflection for exposure and trace deflection for trace control. As to the exposure deflection, deflection data is set to a DA converter and a beam is kept in an off state until the output thereof is settled. Thus, the glitch noise produced upon transition of the output does not influence the result of exposure. On the other hand, the trace deflection for feeding back the amount of movement of a sample table is deflection for correcting the amount of movement of the continuously-moved sample table according to the deflection of the electron beam. Therefore, the glitch noise is produced in the DA converter
8
in FIG.
3
. The execution of high-accuracy trace deflection needs to update trace deflection data in a short cycle. It is thus necessary to update the trace deflection data in a cycle shorter than that for graphics exposure, and the data is updated even during the application of the electron beam. When the trace deflection data is brought to data which causes large glitch noise in the DA converter, a displacement in exposure position due to the influence of the glitch noise occurs, so that the accuracy of exposure is degraded. Thus, it is necessary to limit the range of trace deflection so that the DA converter is used within a range represented in such a level that the glitch noise does not influence exposure.
However, a problem arises in that when the range of the trace deflection becomes narrow, the number of trace r
Ando Kimiaki
Nagata Koji
Okumura Masahide
Takahashi Hiroyuki
Hitachi , Ltd.
Lee John R.
Mattingly Stanger & Malur, P.C.
Quash Anthony
LandOfFree
Electron beam exposure method and apparatus and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Electron beam exposure method and apparatus and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Electron beam exposure method and apparatus and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2935508