JPH04207013A - Position detector - Google Patents

Abstract

PURPOSE:To eliminate the adverse influence of noise light so as to detect the relative positional deviation between a mask and wafer with high accuracy by oscillating an alignment luminous flux on the mask and wafer in a direction in which the oscillation is insensible from the detecting direction. CONSTITUTION:Alignment marks 5 on a mask 1 are scanned with a laser beam from a light source 10 by oscillating a mirror GM to which the laser loam is made incident through a collimator lens 12 and the transmit and diffracted luminous flux of the laser beam is made incident to an alignment mark 3 on the surface of a waver 2. The luminous flux made incident to the surface of the wafer 2 is condensed by means of a light receiving lens 13 and made incident to the surface of a detecting section 11. The luminous flux made incident to the section 11 is photoelectrically converted and an A/D-converting section 19 and computing means 20 find the relative in-plane positional deviation between the mask 1 and wafer 2 from the incident position of the alignment luminous flux to the detecting section 11. When the light noise caused by speckles is averaged by slightly oscillating the alignment luminous flux in a direction (y) which is insensible from the position detection, highly accurate positioning can be realized.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to a position detection device, and for example, in an exposure apparatus for manufacturing semiconductor elements, the present invention relates to a position detection device that detects a position on a first object surface such as a mask or a reticle (hereinafter referred to as "mask"). This invention relates to a position detection device suitable for performing relative in-plane positioning (alignment) between a mask and a wafer when exposing and transferring a fine electronic circuit pattern formed on a second object surface such as a wafer. It is.

(Prior Art) Conventionally, in exposure apparatuses for manufacturing semiconductor devices, relative alignment between a mask and a wafer has been an important element for improving performance. Particularly in alignment in recent exposure apparatuses, precision of, for example, submicron or less is required due to the high integration of semiconductor devices.

At that time, the distance between the mask and the wafer is measured with a surface distance measuring device, etc., and after controlling the distance to a predetermined distance, position information is obtained from a so-called alignment pattern for positioning provided on the mask and the eight surfaces of the sea urchin. is used to align both sides. As an alignment method at this time, for example, the amount of deviation between both alignment patterns is detected by performing image processing, or alignment method as proposed in U.S. Pat. This is carried out by, for example, using a zone plate as a pattern, irradiating the zone plate with a light beam, and detecting the focal point position of the light beam emitted from the zone plate on a predetermined surface.

In general, alignment methods using zone plates are
Compared to the method A using a mere alignment mark, this method has the advantage of being able to perform alignment with relatively high precision without being affected by defects in the alignment mark.

FIG. 7 is a schematic diagram of a conventional position detection device using a zone plate.

In the same figure, a parallel light beam emitted from a light source 72 passes through a half mirror 74 and is condensed at a condensing point 78 by a condensing lens 76, after which it is placed on a mask alignment pattern 68a on a mask 68 surface and a support base 62. The wafer alignment pattern 60a on the surface of the wafer 60 is irradiated. These alignment patterns 68a and 60a are composed of reflective zone plates, and each form a focal point on a plane perpendicular to the optical axis including the focal point 78. At this time, the amount of deviation of the focal point position on the plane is detected by guiding the light onto the detection surface 82 using the condenser lens 76 and the lens 80.

Based on the output signal from the detector 82, the control circuit 8
4 drives the drive circuit 64 to position the mask 68 relative to the wafer 60.

Figure 8 shows the mask alignment pattern 6 shown in Figure 7.
8a and an explanatory diagram showing the imaging relationship between the light beams from the wafer alignment pattern 60a and the wafer alignment pattern 60a.

In the figure, a part of the light beam diverging from the condensing point 78 is diffracted by the mask alignment pattern 68a,
A focal point 78a indicating the mask position is formed near the focal point 78. Further, the other part of the light beam passes through the mask 68 as zero-order transmitted light and enters the wafer alignment pattern 60a on the surface of the wafer 60 without changing its wavefront. At this time, the light beam is diffracted by the wafer alignment pattern 60a, and then passes through the mask 68 again as zero-order transmitted light, condensing near the converging point 78 and focusing on the wafer position.
form. In the figure, when the light beam diffracted by the wafer 60 forms a condensing point, the mask 6ε simply acts as a transparent state.

The position of the focal point 78b due to the wafer alignment pattern 60a formed in this manner is determined according to the amount of deviation Δσ of the wafer 60 from the mask 68 in the direction (lateral direction) of the mask urchin. The amount of deviation Δσ corresponding to the amount of deviation Δσ along the plane orthogonal to the optical axis including
’.

The amount of deviation Δσ' at this time is determined using the absolute coordinate system provided in the detector 82 as a reference, and the amount of deviation Δσ is thereby detected.

(Problems to be Solved by the Invention) In the above-described position detection device, a laser light source with a large optical output and good directivity is used as the light source means.

On the other hand, since the laser light source has good coherence, noise light such as interference fringes due to multiple reflections between the mask and wafer and interference fringes due to speckle scattered light from minute unevenness of the mask and wafer may appear on the detection surface. There is a problem in that the light intensity distribution or cross-sectional shape of the laser beam incident on the detection surface changes depending on the position of incidence on the detection surface.

If the incident position of the noise light on the detection surface is constant, it is often spatially fixed, so the detection signal does not change over time, and the light intensity distribution of the incident light flux is averaged over time. However, this noise light cannot be canceled out. Therefore, the focal point of the laser beam on the detection surface (for example, the center of the circular beam) and the reference point based on the light intensity (for example, the point at which the light intensity is
If you try to detect the movement of the laser beam, the cross-sectional shape of the beam and the light intensity will change, so the amount of movement of the incident point of the laser beam will not match the amount of movement of the actual focal point of the laser beam, and as a result, the position There is a problem in that a deviation detection error occurs.

The present invention reduces detection errors caused by speckle scattered light when a laser beam with good coherence is used as a light source, and detects with high precision the relative positional deviation between a first object and a second object arranged facing each other. The purpose is to provide a position detection device that can perform

(Means for Solving the Problems) The position detection device of the present invention provides alignment for detecting relative in-plane positional deviations on a first object plane and a second object plane, which are arranged opposite to each other. A mark is provided, and out of the light flux from the light projecting means, the light flux that has been affected by the alignment mark on the first object surface and the alignment mark on the second object surface is guided onto a predetermined surface, and the light flux on the predetermined surface is When detecting the relative positional deviation between the first object and the second object by detecting the incident position of the light beam with the detection means, the light beam from the light projecting means is driven by the driving means in the direction in which the relative positional deviation is detected. It is characterized in that minute vibrations are made on the first and second objects in an insensible direction.

Here, the non-sensing direction is, for example, a direction orthogonal to the position detection direction (alignment direction). By vibrating the luminous flux in this direction, speckle noise in the light intensity distribution within the focal point on the detection means surface is averaged over time and converted into an electronic signal to detect the position of the focal point of the luminous flux. are doing.

(Example) Figures 1 (A) and (C) are schematic diagrams of the main parts of the first embodiment of the present invention, and Figure 1 (B) is a schematic diagram of the optical path in which the main parts of Figure 1 (A) are expanded. It is a diagram. Note that FIG. 1(B) shows a case where positional deviation within the XY plane is detected.

This embodiment shows a case in which the present invention is applied to an exposure apparatus for manufacturing proximity type semiconductor devices.

1 is a mask as a first object, and an electronic circuit pattern is formed on its surface. 2 is a wafer as a second object. 5 is an alignment mark provided on the first surface of the mask, and 3 is an alignment mark provided on the second surface of the wafer. The alignment marks 3 and 5 are each one-dimensional or two-dimensional.
It consists of physical optical elements such as grete lenses and Fresnel zone plates that act as dimensional lenses, or diffraction gratings that do not act as lenses.

In this embodiment, the relative positional deviation between the mask 1 and the wafer 2 in the XY plane is detected using the light flux subjected to the wavefront conversion effect by the alignment marks 5 and 3. Alignment mark 3. An example of a pattern when (5) is constructed from a Fresnel zone plate is shown in FIG.

A laser light source 10 emits a highly directional and coherent light beam. 12 is a collimator lens;
A laser beam from a laser light source 10 is made into a parallel beam and is made incident on a mirror GM capable of minute vibrations. The mirror GM vibrates and scans the alignment mark 5 surface of the mask 1 in a direction (X direction) perpendicular to the alignment direction (X direction) with the incident light beam. 11 is a detection unit, for example, C
It consists of line sensors such as CDs. Reference numeral 13 denotes a light-receiving lens that receives the light beam from one surface of the mask and guides it to the detection section 11. 17 is a wafer stage, and wafer 2
is listed. Reference numeral 18 denotes stage driving means, which drives the wafer stage 17.

The mirror GM in this embodiment is made of, for example, a galvanometer mirror, includes a point where the light beam after passing through the collimator lens 12 is reflected on the mirror 0M surface, and has a drive means with respect to an axis parallel to the alignment direction (X direction). 21 makes it vibrate slightly.

Then, the irradiation position on the mask surface is vibrated and scanned with a predetermined amplitude and period using the light beam from the mirror GM.

In this embodiment, the laser beam from the laser light source 1o is made to enter the mirror GM as a parallel beam by the collimator lens 11. By vibrating the mirror GM, the light beam passing through the mirror GM scans the alignment mark 5 surface on the mask 1 surface. A light beam that has been subjected to a wavefront conversion effect, such as a transmission diffraction effect, at the alignment mark 5 is made incident on the alignment mark 3 on the surface of the wafer 2. A light beam that has been subjected to an optical action, such as a reflection and diffraction action, at the alignment mark 3 is transmitted through the alignment mark of the mask 1 without being subjected to a diffraction effect, and is condensed by the light-receiving lens 13, where it is delivered as an alignment light beam onto the surface of the detection unit 11. incident.

FIG. 3 is an explanatory diagram of the intensity distribution of the alignment light incident on the surface of the detection unit 11. The alignment light flux incident on the detection unit 11 is charged and converted, and a signal having a waveform similar to that shown in FIG. 3, for example, is obtained.

The output signal from the detection section 11 is A/D converted by the AD conversion section 19 and sent to the calculation means 20. and calculation means 2
0, the position of incidence of the alignment light beam on the detection unit 11 (
The relative in-plane (
The positional shift (in the XY plane) is being calculated.

As shown in FIG. 1(C), the wafer stage 17 is driven by the stage drive means 18 based on the signal from the calculation means 20.
is moved via an actuator 22 to align the mask 1 and wafer 2. Note that DH is a detection head, which includes a laser beam filO, a mirror GM, and a detector 11.
is stored.

Now, as shown in FIG. 1(B), the mask 1 and the wafer 2 are shifted by Δσ in the parallel direction, and the distance from the sea urchin A 2 to the convergence point of the light beam reflected by the alignment mark 3 of the wafer 2 is a. , if the distance to the converging point of the light beam that has passed through the alignment mark 5 of the mask 1 is b, then the amount of gravity shift Δδ of the condensing point of the alignment light beam on the detection unit 11 surface is Δδ=Δa x (−+ 1 )・・・・・・・・・
...(a). In other words, the center of gravity shift amount Δδ is (b / a
+1) will be enlarged twice.

For example, if a=0.5 mm and b=50 mm, the center of gravity shift amount Δδ is expanded by a factor of 101 from equation (a).

In addition, the center of gravity shift amount Δδ and position shift amount Δσ at this time are (a
) It is clear from the equation that there is a proportional relationship.

Assuming that the resolution of the detection unit 11 is 0.1 μm, the positional deviation amount Δσ has a position resolution of 0.001 μm.

In this example, the center of gravity of the light flux is the point within the light receiving surface of the detection unit where the integral value becomes 0 vector when the product of the position vector of each point on the light receiving surface multiplied by the light intensity at that point is integrated over the entire cross section. However, the point where the light intensity is the highest may be used as the representative point.

The principle of this positional deviation detection method in this embodiment has been proposed by the present applicant in, for example, Japanese Patent Laid-Open Nos. 1-233305 and 2-1506.

Next, the features of this embodiment will be explained.

In general, when a laser beam enters a fine unevenness on a mask surface or a fine unevenness on a wafer surface coated with A1 (~1000 people), speckle scattered light is generated by these unevenness and noise light is generated. As a result, it is superimposed on the alignment beam on the detection surface. The intensity of noise light caused by this speckle scattered light increases as the coherence of the laser light source increases. The incident position and size of this noise light vary depending on the relative position and spacing of the mask and wafer.

Since the noise light does not change over time unless the laser light source, mask, wafer, etc. change relative to each other, the noise light cannot be removed even if electrically averaged over time.

Furthermore, since the spot diameter of the noise light is approximately equal to the diameter of the alignment light beam, when it overlaps with the alignment light beam on the detection surface, it cannot be separated spatially or on the spatial frequency axis.

As a result, as shown in FIG. 4, for example, the waveform of the output signal from the detection section is disturbed and the center of gravity of the incident light beam on the surface of the detection section fluctuates, resulting in a positional deviation detection error. Since this detection error occurs at the position of the light spot on the detection surface, the positional deviation detection error cannot be reduced even if it is averaged after being converted into an electrical signal.

Therefore, in this embodiment, in order to optically and spatially average the noise light due to the speckle scattered light superimposed on the locally concentrated alignment light flux, the alignment light flux 91 is shifted from the position on the alignment mark surface as shown in FIG. During detection, a slight vibration is made in an insensitive direction (X direction).

For example, alignment direction (X direction) and orthogonal direction (X direction)
It is made to vibrate slightly within the range of 1,000 to 2,000 people.

Generally, the detection unit 11 includes a plurality of elements (lla) in the positional deviation detection direction (X direction) as shown in FIG.

flb, tic,...).

For this reason, in this embodiment, as shown in FIG. 9, when detecting positional deviation, the alignment light beam 91 is
is vibrating.

The amplitude of the minute vibration of the alignment light beam in this embodiment is determined by the system in the direction perpendicular to the alignment direction of the light beam, and the period of this vibration depends on the accumulation time of the detection unit.

The size of the alignment mark for positioning is usually within a width of several tens to hundreds of micrometers because it is provided on the scribe line, which is an area that is made easy to separate chips that are printed in a grid pattern on a single wafer. is normal.

Therefore, the diameter of the alignment light beam (in the alignment vertical direction) must not only cover the entire alignment mark, but also have a light intensity that does not deviate from the alignment mark even if the irradiation position is vibrated, and is sufficient to be detected on the detection unit. .

Therefore, it is desirable that the amplitude of the minute vibration in this embodiment be several hundred microns, and the vibration period be an integral multiple of the accumulation period of the detection section.

In this example, as shown in FIG. 9, the alignment light beam is slightly vibrated relative to the mask and wafer in the X direction, which is insensitive to position detection, and the minute vibration of the alignment light beam at this time is used to detect speckles. Highly accurate positioning is achieved without reducing detection accuracy by averaging optical noise and generating minute vibrations in a direction that is insensible when substantially orthogonal to the position detection direction.

FIG. 5 shows the voltage applied to the piezoelectric element and the CC as the detection unit 11 when a piezoelectric element is used as the driving means 21 of the mirror GM to slightly vibrate the alignment light beam.
FIG. 4 is an explanatory diagram of a timing chart showing the relationship with the output signal from the detection section when D is used.

In the above embodiment, an example was shown in which minute vibrations of the alignment light beam were performed using a galvanometer mirror, but it is also possible to use an acousto-optic element AO, as shown in FIG. 10, for example.

In FIG. 10, a laser beam from a laser light source 101 whose drive is controlled by a driving means 105 is converted into a parallel beam by a collimator lens 102 and sent to an acousto-optic element 103.
It is input to. The acousto-optic element and driving means 104 controls the vibration frequency applied to the acousto-optic element 103 and controls the light flux (diffraction light) emitted from the acousto-optic element 103 by forming a phase grating.

In the figure, only the primary light L1 is used, so
Other zero-order light LO and secondary light L2 are blocked by a light-shielding plate 106. By periodically changing the vibration frequency applied to the acousto-optic element 103 by the acousto-optic element driving means 104, the primary light L1 emitted therefrom is slightly vibrated.

(Effects of the Invention) According to the present invention, as described above, by vibrating the alignment light beam in a direction insensitive to the detection direction on the mask and wafer, the mask can be used as a light source when a laser light source with good coherence is used as the light source means. Achieved a position detection device that can eliminate the adverse effects of noise light based on speckle scattered light generated by minute irregularities of the mask and wafer, and can detect relative positional deviation between the mask and the wafer with high accuracy. can do. Moreover, speckle noise can also be reduced without increasing the size of the device.

[BRIEF DESCRIPTION OF THE DRAWINGS] FIGS. 1(A) and 1(B) are schematic diagrams of main parts and a partially expanded explanatory diagram of the first embodiment of the present invention, and FIG. 1(C) is a diagram of FIG. 1(A).
Figure 2 is a schematic diagram of a part of the alignment mark in Figure 1 (A), Figures 3 and 4 are in Figure 1 (A).
FIG. 5 is an explanatory diagram of the intensity distribution of the alignment light beam on the detection section surface; FIG. Figure 6 is a schematic diagram of the detection unit in Figure 1 (A), Figures 7 and 8 are schematic diagrams of a position detection device using a conventional zone plate, and Figure 9 is an alignment mark and alignment light beam in the present invention. FIG. 10 is an explanatory diagram of another embodiment in which the alignment light beam is slightly vibrated according to the present invention. In the figure, 1,72 is the first object (mask), 2°73 is the second object
Object (wafer), 3.5 is an alignment mark, 10 is a laser light source, 11 is a detection unit, 12 is a collimator lens, 13 is a light receiving lens, 17 is a wafer stage, 18 is a wafer drive means, 19 is an A/D conversion 20 is a calculation means,
GM is a mirror, and 21 is a driving means. Fig. 1 (A), Fig. 1 (B) Fig. 1 (C) Fig. 2 Fig. 6-1→X direction Fig. 3 -14 N! -3n! -2;';? -1i
+1 nj;: ? Dou (Type 113), 4g, Fig. 4 Sensor position Fig. 7 Fig. 8 Fig. 9 Fig. 10

Claims (1)

[Claims] (1) Alignment marks are provided on the first object plane and the second object plane, which are disposed facing each other, to detect relative in-plane positional deviation between the two, and the By guiding the light beams acted upon by the alignment marks on the object surface and the second object surface onto a predetermined surface, and detecting the incident position of the light beams on the predetermined surface with a detection means,
When detecting a relative positional deviation between the first object and the second object,
A position detecting device characterized in that a light beam from the light projecting means is slightly vibrated on the first and second objects by a driving means in a direction insensitive to a direction in which relative positional deviation is detected.