 |
Description  |
|
|
STAGE POSITIONING APPARATUS
1. Technical Field
This invention relates to a stage positioning apparatus, and relates in
particular to a stage positioning apparatus suitable for placing a
specimen in semiconductor device manufacturing and inspection apparatus.
2. Background Art
Various processing and observations in semiconductor manufacturing and
inspection apparatuses are usually carried out by placing the specimen on
an x-y stage.
In recent years, integrated circuit density has increased significantly,
and consequently, line width and interline spacing of circuits has become
smaller and micro-sized. Particularly, for sub-micron lithography, in
order to secure high precision positioning in the layering of circuit
patterns, it is essential to precisely position the patterns. High speed
handling of the patterns is important for productivity.
However, conventional x-y stages are operated by feedback control of some
actuators, for example, servo motors, provided on the x-y stage to move
the stage by means of a ball screw and other arrangements. With such
devices involving mechanical friction, positioning has not always been
performed satisfactorily at high speed and high precision. Further, even
for devices based on air bearings and linear motors to avoid mechanical
friction effects, there is a problem that the positioning apparatus can be
affected by acceleration and deceleration effects of a moving stage. The
stage can be affected by excitation of fundamental frequencies, which
adversely affects the accuracy of positioning.
For example, for a scan type stepper device, it is necessary to move the
x-y stage smoothly at high speed and high precision. Thus, there has been
an increasing demand for a high precision and high productivity x-y stage
for positioning specimens for measuring or fabricating purposes in a
semiconductor manufacturing apparatus. Similar needs exist for electron
microscopy, such that positioning must be carried out at sub-micron
accuracy precisely and quickly. Such precision positioning devices are
sensitive to vibration, such that even though precision positioning has
been originally performed, that position can be lost later due to
vibration.
For this reason, vibration elimination devices are used to isolate or
attenuate the vibrations that can be transmitted from an installation
floor or external disturbances transmitted through air, such as air
conditioning. However, vibrations that can be controlled by such vibration
elimination devices are limited to those produced by the x-y table. For
example, even if a semiconductor fabrication apparatus is placed on an
anti-vibration table, it is not possible to control the vibration of an
optical beam used to fabricate a specimen inside an apparatus for
semiconductor manufacturing. For this reason, when positioning is required
at sub-micron level precision in a specimen, the optical beam itself may
be displaced by vibration such that the most important aspect in the
fabrication, i.e., the beam position, cannot be precisely aligned.
DISCLOSURE OF INVENTION
This invention was made in light of the background described above, and it
is an object of the present invention to provide a micro-positioning
apparatus to enable stable and rapid positioning of a stage on which a
specimen is placed. Also, another object is to provide a stage positioning
apparatus that has a compact design and a low amount of leakage of
magnetic flux and can be operated in a vacuum environment.
An apparatus for positioning a stage comprises a stage for placing a
specimen to be radiated with a beam, actuators for levitating the stage
and controlling a movement of the stage, a first position sensor for
measuring a relative displacement between the stage and the actuators, a
second position sensor for measuring a relative displacement between an
actual radiation position of the beam on the specimen and a target
radiation position, and a controller for positioning the stage so as to
decrease the relative displacement detected by the second sensor.
According to this invention, the stage is directly positioned by measuring
the relative displacement between the actual radiation position of the
beam used for fabrication or measuring of the specimen and the target
radiation position, and the stage is moved so as to decrease the relative
displacement. By this process, the beam is accurately positioned on the
target radiation position. This process enables micro-positioning of the
beam to be carried out even when the beam itself or the stage itself is
vibrating, for example, in a semiconductor production facility.
A magnetically levitated stage comprises a levitation body having a table
section for placing a specimen and side plates extending from outer
peripheries of the table section, and actuators for levitating and
positioning the levitation body by controlling magnetic force generated by
electromagnets therein, the actuators being surrounded by the table
section and the side plates, wherein the actuators have permanent magnets
disposed near a center section thereof for supporting the weight of the
levitation body, electromagnets for controlling horizontal positioning
which are disposed in four corners in an outer periphery thereof, and
electromagnets for controlling vertical positioning which are disposed in
a middle section between the electromagnets for horizontal positioning.
In accordance with the present invention, a compact stage positioning
apparatus has been provided by disposing electromagnets and permanent
magnets serving as actuators inside a boxed space having bottom opening of
a levitation body.
BRIEF DESCRIPTION OF DRAWINGS
FIG. 1A is a schematic side view of a first embodiment of a stage
positioning apparatus, FIG. 1B is a plan view of the apparatus, and FIG.
1C is a partial enlarged view of the apparatus.
FIG. 2 is a block diagram of a control system for the apparatus.
FIG. 3 is another block diagram of the control system for the apparatus.
FIG. 4 is a flowchart of the control system for the apparatus.
FIG. 5 is a cross sectional side view of an embodiment of the stage
positioning apparatus.
FIG. 6 is a diagram of a variation of the apparatus shown in FIG. 5.
FIG. 7 is a block diagram of a cooling system control for the apparatus
shown in FIG. 6.
FIG. 8 is a diagram for explaining the stage positioning apparatus disposed
on a high performance vibration elimination device.
FIG. 9 is a diagram to show an arrangement of another embodiment of the
stage positioning apparatus.
FIG. 10 is a through side view of the apparatus shown in FIG. 9.
FIG. 11 is a cross sectional side view of the apparatus shown in FIG. 9.
FIGS. 12A.about.12C are diagrams to explain passive magnetic bearings based
on permanent magnets, which are employed in the apparatus shown in FIG. 9.
FIG. 13 is a block diagram of a control system for the apparatus shown in
FIG. 9.
FIG. 14 is a diagram to explain a magnetic shield having a labyrinth
structure.
FIG. 15 is another diagram to explain the magnetic shield having the
labyrinth structure.
FIG. 16A is a plan view of an electromagnet, FIG. 16B is a cross sectional
plan view of the electromagnet, and FIG. 16C is a cross sectional plan
view of the electromagnet.
BEST MODE FOR CARRYING OUT THE INVENTION
Preferred embodiments of these inventions will be explained with reference
to the drawings.
FIGS. 1A-1C show an embodiment of the stage positioning apparatus. In this
apparatus, a stage 11 for placing a specimen is supported at four corners
by actuators 12a, 12b, 12c and 12d. The actuators 12a, 12b, 12c, 12d are
provided with six-degrees of freedom (x, y, z, .alpha., .beta., .gamma.
directions) in the translational and rotational movements to generate
actuator control forces fx, fy, fz, f.sub..alpha., f.sub..beta.,
f.sub..gamma. ;. The stage 11 is for placing a specimen (wafer) for
fabrication or measurement purposes using an electron beam or a light beam
B. A displacement sensor (first sensor) 13 is provided for detecting the
stage position and generating relative displacement signals to the
actuators (stationary sections). Further, a controller 15 controls the
actuator control forces fx, fy, fz, f.sub..alpha., f.sub..beta.,
f.sub..gamma., according to the relative position signals from the
displacement sensor 13, actual radiation position and target radiation
position sensors which will be described later.
Actuators 12a, 12b, 12c, 12d are made of electromagnets or a combination of
permanent magnets and electromagnets. The pole surfaces of the actuators
are oppositely disposed with surfaces of magnetic material or permanent
magnets affixed on the stage. Therefore, by adjusting the coil current of
the electromagnets, magnetic forces on the stage can be affected, so that
the stage II can be levitated or moved or translated for positioning.
Electromagnets are controlled in the levitating(vertical) and horizontal
directions, and generate control forces in six-degrees of freedom in the
x, y, z, .alpha., .beta., .gamma., directions. Any of the actuators 12a,
12b, 12c, 12d is capable of being controlled precisely at high speed by
adjusting the current or voltage supplied to the actuators to generate the
necessary forces.
A specimen W such as a wafer is placed on the stage 11. The specimen W is
placed with a photo-resist for electron-beam exposure, for example, and
micro-patterns are formed by radiating the coated specimen W with an
electron beam B. The beam B is emitted from a beam source 35 and a pattern
is formed by radiating the beam B on a target position of the wafer W. For
positioning to the target position, the movement of the stage 11 is
controlled by the actuators 12a, 12b, 12c, 12d. It is obvious that for
making a line width of sub-micron precision, it is necessary to position
the wafer W at less than sub-micron precision.
Therefore, the positioning is required to directly align the target
radiation position with the actual radiation position. A sensor (second
sensor) 36 is provided to measure the relative positions of the actual
radiation position and the target radiation position. Measurement of
relative positions is performed as illustrated in FIG. 1C. A target
pattern T to show the target radiation position is provided on the wafer
W, where the pattern T is made of a material that can reflect the electron
beam B. When the beam B is maximally reflected with the pattern T, it
shows that the center Bc of the beam B and the center Tc of the pattern T
are coincided. When this matching occurs, the reflection is at a maximum,
which is detected by the sensor 36 so that it is possible to determine
that the target radiation position and the actual radiation position have
been aligned. When the beam center Bc and the pattern center Tc are
displaced, the reflected intensity decreases, and the amount of reflection
is measured to determine the relative positions of the target radiation
position and the actual beam radiation position.
Relative position signals of the actual radiation position of the beam B
and the target radiation position, and the relative position signals with
respect to the fixed side of the stage are input into the controller 15,
and the stage 11 is moved by using the actuators 12 so as to decrease the
relative displacement of the actual radiation position and the target
radiation position. That is, the actuators 12 drive the stage 11 so as to
directly position the target pattern T of the wafer W at the actual
radiation position of the beam B.
FIGS. 2 and 3 show block diagrams of positioning control systems. The
control system shown in FIG. 2 inputs a relative displacement Xr given by
the actual radiation position and the target radiation position according
to a sensor signal from the second sensor as the reference signal, and
controls the controller 15 to operate the actuators so that the sensor
signal X from the first sensor follows the sensor signal Xr from the
second sensor. This is carried out by inputting the displacement signal Xr
for the actual radiation position in the comparator 16, computing the
difference between the actual displacement signal Xr and the stage
position signal X, and generating compensation signals (operational
signals) so as to decrease the difference to zero, which are supplied to
the actuators 12a, 12b, 12c, 12d.
A plant 17 is used to show the relation between the input signal to the
actuators and the resulting stage position signal X, and the stage
position signal after the stage has been driven by the actuators is fed
back to the comparator 16. Thus, the controller 15 treats the relative
displacement of the stage and the actuator as the first control value, and
treats the relative displacement of the actual radiation position and the
target radiation position as the second control value, and the stage is
controlled to move in reverse phase in relation with these control values.
Therefore, when the relative displacement signal Xr of the beam radiation
position is not input, the stage is always positioned in the reference
position indicated by the relative displacement signal X relative to the
fixed section of the stage. When the relative displacement signal Xr for
the beam radiation position is input, actuators are moved according to the
signal, and the stage is positioned so that the relative displacement is
decreased by reducing the difference to zero between the actual radiation
position and the target radiation position. Thus, the stage 11 is
controlled to move so as to follow and align the target radiation position
with the actual beam position.
The control system shown in FIG. 3 performs feed-forward control to move
the stage so that the actuators are operated to follow the position
compensation signal by inputting the relative displacement signal Xr for
the beam radiation position through a transmission function Q. In the
control system shown in FIG. 3, the relative displacement signal Xr for
the second sensor, that is, the actual radiation position of the beam B
and the target radiation position, is used as reference signal, and is
input in the comparator 16, and the relative displacement signal X for the
first sensor will follow the signal Xr through the controller 15, and such
actions are the same as the control system shown in FIG. 2. The relative
displacement signal Xr for the beam position is input in the comparator
16, and the difference from the stage position signal X is computed, and
the compensation signals (operational signals) are generated, so as to
make the difference zero, and are supplied to the actuators 12a, 12b, 12c,
12d.
In this control system, the relative displacement signal Xr between the
actual radiation position of the beam B and the target radiation position
is added to the output signal from the controller 15 through the
transmission function Q. The transmission function Q is given, for
example, by the following relation.
Q=-(1+PH)/P
where P is a transmission function of the plant 17, and H is a transmission
function of the controller 15. Such a feed-forward control system enables
the expansion of the controllable frequency bandwidth significantly to
increase the stability thereof.
FIG. 4 shows a flowchart for converting a moving position compensation
signal based on the relative displacement signal between the actual
radiation position and the target radiation position to the compensation
signal for each actuator position. A command value of the movement of the
stage on the plane is given by coordinates X, Y. However, the actuators
12a, 12b, 12c, 12d for moving the stage to the commanded position are
provided on the four corner sections of the stage. Therefore, the
compensation signal at each of the actuators 12a, 12b, 12c, 12d is
required to be converted from the command signal value of the movement.
For this reason, a coordinate conversion array is used to convert the
command values of the coordinates to the position compensation signals for
each actuator.
Therefore, the controller H has a computation section to convert the
relative displacement between the actual radiation position and the target
radiation position to the relative displacement of the center of gravity
of the stage in the coordinates, a computation section to generate the
operational values corresponding to the displacement of the gravity
positions converted to the coordinates, and a computation section to
distribute the operational values at each of the acting points of the
electromagnets.
The stage 11 may be provided with a sensor for detecting vibrations. An
acceleration value of the specimen W obtained from the sensor may be input
in the controller H, which controls so as to decrease the vibration. Thus,
when the stage itself is vibrating, stage vibration can be attenuated so
as to improve positioning of the actual radiation position and the target
radiation position even more reliably. In this case, the controller H has
a computation section to convert the acceleration in the coordinates to
the position of the center of gravity of the stage 11, and a computation
section to generate the operational values based on the acceleration of
the converted coordinates, and a computation section to distribute the
operational values at the acting points of the electromagnets.
FIG. 5 shows an embodiment of the electromagnetic actuator section of the
positioning apparatus. The stage 11 with a specimen W is supported at its
four comers by the electromagnetic actuators 12 having electromagnets 21,
22. Electromagnet 21 performs positioning in the horizontal direction by
attracting a magnetic body fixed on the stage 11 by magnetic attractive
force which is controlled by the current supplied from the controller.
Electromagnet 22 similarly supports the stage 11 non-contactingly by
attracting a magnetic body 11v fixed on the stage 11 by magnetic
attractive force. Vibrations from the floor are prevented from reaching
the stage levitated by the actuators. The floating support may be produced
by using permanent magnets in addition to the electromagnets. This
arrangement reduces the current load of the electromagnets.
As shown in FIG. 5, when the actuators use electromagnets, leaking magnetic
flux generated by the actuators are prevented from affecting the stage by
a magnetic coating and by a plate 11a on the bottom surface of the stage
11. Further, a cover 11b of magnetic material having a labyrinth structure
is provided in such a way to surround the actuators 12 to prevent flux
leakage from the actuator. To prevent leaking magnetic flux from the space
in the labyrinth structure, a magnetic coating or plate 19a is provided on
the fixed surface 19 which attaches to the actuators 12. Further, the
electromagnets are protected by a can 20 so as to enable the stage
positioning apparatus to be used in a vacuum without the problem of
degassing. Also, the entire actuators may be covered by a can.
FIG. 6 shows an example of using a cooling system on the actuators. When
precision positioning is required, there is a problem of thermal
dissipation produced by the electromagnets. Heat from the electromagnets
deforms various parts of the actuators to cause a problem of preventing
the attainment of the precision required. A system of cooling based on a
quick acting Peltier element may be considered. The actuators shown in the
drawing are outer type actuators, so that the center part is the fixed
part, and outer parts are provided with a levitation body. Heat is
generated mainly from the fixed side of electromagnets so that Peltier
elements (cylindrical) 25 are bonded to the inside of the parts
(cylindrical) and water is flowed through an inside passage 27. In this
case, the outside of the Peltier element is the heat absorption side, the
inside is the heat generation side. By using this cooling system, cooling
water flowing from an inlet port 271 absorbs the heat transferred by the
Pettier element 25, and is discharged outside from an outlet port 270 to
maintain the temperature of the actuators at normal temperatures.
Also, to keep the temperature constant, the temperature of the fixed parts
of the electromagnets are detected by a temperature sensor 26 and the
current through the Peltier element 25 is controlled to maintain the
cooling system in a stable condition. FIG. 7 shows such a temperature
control system, and the temperature detected by the temperature sensor 26
is compared with a temperature command, and a temperature controller 28 is
operated in PID so as to make the difference zero. The output of the
temperature controller 28 is amplified in an electrical amplifier 29 and
is supplied to the Peltier element 25, thereby controlling the heat
transfer from the heat absorption side to the heat generation side. Thus,
the temperatures of the various parts of the actuators are kept within a
constant range.
FIG. 8 shows an example of installing the stage positioning system on an
vibration elimination device. A beam source 10, an electron-beam
generator, for example, is placed on a table 31 of a high performance
active vibration elimination device 32. The stage 11 having a specimen W
to be fabricated by the electron beam from the beam source 10 and the
actuators 12a, 12b, 12c, 12d supporting the stage are also placed on the
table 31. This configuration essentially eliminates the transmission of
vibrations from external regions to stage 11, actuators 12a, 12b, 12c,
12d, beam source 10, and enables even higher precision positioning. This
arrangement also prevents generation of vibration caused by the movement
of the positioning apparatus including the stage 11. The vibration
elimination device is most suitably produced by a non-contact magnetic
floating arrangement using the electromagnetic actuators, and a
combination device of air springs and electromagnetic actuators.
Also, stage positioning for actual radiation position and the target
radiation position is used mainly for position compensation in production
of micro-patterns using the electron beam. For example, for moving the
stage at a large distance for producing another pattern, the controller H
is controlled in PID mode according to feedback of the output signal from
the first sensor for measuring the relative displacements of the actuators
and the stage. In this case, the relative displacement signal Xr for the
actual radiation position of the beam B and the target radiation position
are stopped, and the input to the comparator 16, for example, would be
zero.
The second sensor for detecting the relative displacement for the' actual
radiation position of the beam B' and the target radiation position may
use the beam B for performing actual fabrication, or another beam B' which
is in parallel. In this case, the relative position is obtained by
radiating the specimen with the beam B and radiating the target pattern T
with the parallel beam B'. When using the beam B for fabrication and
relative displacement measurements, time-sharing may be used to detect the
two quantities.
As explained above, this invention enables the controlling of positioning
precisely by directly acting on the actual radiation position of the beam
and the target radiation position on the specimen. Therefore, the
positioning apparatus is most suitable for a specimen pedestal for
producing micro-patterns using a beam of the order of sub-microns.
FIGS. 9 through 11 show another embodiment of the stage positioning
apparatus. A levitation body F has a plate 52 for a center table to place
a specimen and four vertical plates 53 extending from the outskirts of the
plate 52, and the bottom of the plate 52 is open to form a box type plate.
Actuators A, as a fixed part, are placed inside the box type table. The
actuators comprise a permanent magnet 47 comprising one side of a passive
magnetic bearing for supporting the levitation body F inside the space
surrounded by the boxed plate 52 and the side plates 53, electromagnets
41, 42 to move the levitation body F in the horizontal direction, an
electromagnet 43 for moving the levitation body F in the vertical
direction, and a sensor 54 for detecting the relative displacement between
the pole surfaces of the electromagnet and the target fixed to the
levitation body F.
On the inside surface of the box type side plate 53 having a bottom opening
is provided with a target 44 for responding to the magnetic force in a
horizontal direction by the electromagnet 41 for the x-direction, and a
target 45 for responding to the magnetic force in the horizontal direction
by the electromagnet 42 for the y-direction. A target plate 46 for
responding to the magnetic force in a vertical direction by the
electromagnet 43 for the z-direction is fixed to the inside o f the side
plate 53.
The actuators A have magnetic bearings 47, 48 placed centrally inside a
space formed by the plate 52 serving as the table and side plates 53 at
outskirts of the plate 52 constituting the levitation body F, for
supporting the weight of the table, the electromagnets 41, 42 on the outer
four corners in the space, for moving the levitation body F in the x- and
y-directions, and electromagnets 43 for moving the levitation body F in
the vertical direction, the z-direction, in the middle between the
electromagnets 41, 42 for moving in the x- and y-directions. This
arrangement includes all the parts, including the coils, of the
eledromagnets installed in the space formed by the plate 52 and the side
plates 53 as floating body F.
By placing the electromagnets 41, 42 on the four corners of the box type
table, controls for the x-, y- and rotation about the z-direction can be
easily carried out. By placing the electromagnet 43 for controlling
positioning in the vertical direction between the electromagnets 41, 42,
the z-direction movement, rotational control about the x- and y-directions
can be easily carried out. Such controls can be carried out at high
precision using a static capacitor sensor and others for position control
sensors, and comparing the output signal with a target value to feedback
the signal to control the excitation current for the coil of the
electromagnets. This type of control can accomplish six-axes control at a
precision of the order of nanometers.
The electromagnet 43 for vertical direction control, as shown in FIG. 11,
comprises a pair of electromagnets, which are disposed in such a way that
each pole surface in the vertical direction is opposed to the other, and a
magnetic body 46, fixed to the levitation body F, is inserted between the
pole surfaces of electromagnets in the horizontal direction. But, this
vertical direction electromagnet may be disposed in such a way that each
has a pole surface on the top and bottom surface, so that both pole
surfaces are facing in the opposing vertical direction, and the pole
surfaces of the electromagnets are facing the magnetic body fixed in the
horizontal direction to the top and bottom side of the side plates of the
outer surfaces of the levitation body. Actuators A have magnetic bearings
47, 48 supporting the weight of the table F in about the center of the
space formed by the flat plate and side plates of the box type table. As
shown in the cross sectional view in FIG. 11, the flat plate 52 has a
vertical second side plate 53a in about the center of the table. A pair of
permanent magnet arrays 47, 48 is disposed between one surface of the side
wall 53a and one side of the fixed side 51 to serve as passive
electromagnets. In this embodiment, the side plate 53a of a rectangular
cross sectional shape has the permanent magnet array 48 at four outskirts,
and the surface stage opposing fixed side 51 also has the four permanent
magnets array 47. As shown in FIG. 12, the permanent magnet arrays 47, 48
are arranged in several layers so that the magnetizing direction of the N,
S poles are facing opposite to each other.
By arranging the permanent magnetic arrays 47, 48, as shown in FIG. 12A, in
opposing directions, and one side is fixed to the fixed side of the
actuators A, and the other side is fixed to the side surface of the side
plate 53a on the inside surface of the levitation body F so that each
faces the other. Then, as shown in FIG. 12B, the magnetic flux lines pass
through and between the magnetic plates of opposite polarities of the
respective permanent magnets on the floating side and the fixed side.
Therefore, as shown in FIG. 12C, the floating side permanent magnet arrays
fixed to the table generate vertical restoring force F against the fixed
side permanent magnets by the weight at the location positioned downward,
and balance is obtained to support the weight of the levitation body F.
Although the passive type electromagnetic bearings based on permanent
magnets are disposed at four locations between the side plates disposed on
the inside surface of the table and the fixed side, it is desirable that
the center location in the vertical direction is coincident with the
center of gravity of the table. This arrangement enables the flat plate 52
with the specimen to be levitated horizontally by the levitation body F in
a more stable manner.
A control system for the stage positioning apparatus is shown in FIG. 13. A
displacement sensor 53 for detecting displacement of the stage comprises a
total of six sensors, with three sensors for detecting horizontal
displacement of the stage and three sensors for detecting vertical
displacement of the stage. Output signals from the six sensors are
detected through sensor amplifiers, and converted to displacement values
or values corresponding to displacements by passing them through a
coordinate conversion device 1. Converted signals are processed by a
computation section such as PID or vibrational mode computers, to obtain
control values, which are converted to command values for each
electromagnet in the coordinate conversion device 2, and coil currents
corresponding to each of the command values are obtained through power
amplifiers to operate each electromagnet so as to position the stage. The
controller may be a digital or analog type.
A shield material 55, such as Permalloy, is bonded to the levitation body F
and on the outer surface of the fixed section 51 of the actuators A so as
to prevent leaking of magnetic flux from the electromagnet and permanent
magnets. However, the space between the table and the fixed sections
cannot be fixed with the shield material, and to prevent leaking from this
space, it is preferable to adopt the structure shown in FIG. 14. The outer
surface of the fixed side comprises side plates 51b, 51c made of a double
walled magnetic material, and the bottom section of the side plates 53
extending from the four peripheries of the table section (flat plate 52)
are inserted in the space between the double-walled side plates 51b, 51c.
This arrangement produces a shield with a labyrinth structure, and leaking
of magnetic flux from the electromagnets and permanent magnets can be
prevented. Also, the labyrinth structure of the shield may be made from a
triple-walled structure, as shown in FIG. 15.
Also, in this embodiment, the horizontal direction control electromagnets
41, 42 and the vertical direction control electromagnet 43 have the same
size and shape. Therefore, standardized components can be adopted so as to
provide the stage positioning apparatus economically. Also, as shown in
FIGS. 16A-16C, each coil C of the electromagnets is housed inside the pole
| | |
