JP2517638B2 - Position detection device - Google Patents

Description

The present invention relates to a highly accurate alignment apparatus, and more particularly to a relative alignment apparatus for a mask or reticle and a wafer in a semiconductor exposure apparatus.

[Conventional technology]

In recent years, the degree of integration of semiconductors has increased, and already 1Mbit to 4Mb
Mass production of it DRAM has begun, and it is moving toward higher integration.

With such higher integration, the minimum line width on the wafer surface has become 1 μm or less, and the alignment accuracy between the mask and the wafer required for the semiconductor exposure apparatus is high.
It is less than 0.2 μm.

Conventionally, regarding alignment, alignment marks have been provided on a mask and a wafer, and alignment has been performed based on the position information of these.

Some alignment methods perform evaluation by detecting a pattern shift on the mask and the wafer by image processing, or by using a zone plate as a mark to detect the position of the focal point of the irradiation beam. Among them, attention is paid to the method in which zone plates are respectively formed on the mask and the wafer and the positional deviation of the incident beam is detected, because they are not easily affected by mark defects (USP 4,037,969, JP-A-56-1570).
33). Figure 10 shows the configuration.

The parallel beam emitted from the light source 72 is condensed by the condenser lens 76, passes through the condensing point 78, and is irradiated on the alignment mark on the mask 68 and the alignment mark on the wafer 60. These alignment marks are composed of reflection type zone plates, and form converging points on a plane that includes 78 and is orthogonal to the optical axis. The lens 7
6 and 80 are used to guide and detect on the detector 82. At this time, FIG. 11 shows in detail the image formation relationship of the light from the mask mark and the wafer mark.

A part of the light beam emitted from the converging point 78 is diffracted by the mask mark on the mask 68, and a condensing point representing the mask position is formed in the vicinity of the converging point 78. A part of the light passes through the mask 68 as zero-order transmitted light and is irradiated on the alignment mark on the wafer 60 without changing the wavefront. The light diffracted from the wafer mark passes through the mask 68 again as 0th-order transmitted light,
The light is condensed in the vicinity of 78 to form a light condensing point that represents the wafer position. That is, the mask has only a function of passing through when forming the focal point by the wafer 60.

The position of the focal point of the wafer mark formed in this manner moves by Δσ ′ to the same extent as Δσ along the plane orthogonal to the optical axis including 78 in accordance with the deviation Δσ of the wafer with respect to the mask. By detecting this Δσ ′ by the photoelectric detector 82, the amount of lateral deviation between the mask and the wafer is detected, and the control circuit 84 which receives the output signal of the detector 82 drives the fine movement mechanism 64 based on this signal to mask the mask. 68 and wafer 60
Align and.

[Problems that the invention is trying to solve]

As described above, when the positions of the mask and the wafer are detected using a physical optical element such as a Fresnel zone plate, monochromatic light (or quasi-monochromatic light) is used as a light source, and usually laser light is used. The illuminance distribution on the alignment mark is generally close to the Gaussian distribution. In particular,
In the case of a semiconductor laser, as shown in FIG. 5, the divergence angle of the light flux has different divergence angles in the vertical direction and the parallel direction of the active layer, and the divergence angle in the vertical direction is generally large. FIG. 5 shows a state in which the light source image of the semiconductor laser 1 is formed by providing the condenser lens 2. The illuminance distribution near the pupil plane of the condenser lens 2 is generally a far field pattern, and the illuminance distribution near the light source image. Is called a near field pattern, and when the astigmatism of the semiconductor laser is sufficiently small, the wavefront at the position of the light source image (this is called the beam waist position) becomes a plane wave, and the radius of the e ^-2 intensity of the spot at that time is the beam. Called waist radius.

On the other hand, the alignment mark in the semiconductor exposure apparatus is preferably on the scribe line for the purpose of reducing the occupied area, and in that case, the alignment mark may be placed within the scribe line width of 20 to 100 μm or less (the length is 500 μm or less). It is customary. At this time, a method of forming a beam waist on the alignment mark by collecting light with the condenser lens 2 described above is generally used. However, the illuminance distribution at this time is the Gaussian distribution as described above, and has the following problems.
That is, as shown in FIG. 6, when the beam and the mark deviate from the normal positional relationship and the illuminance distribution on the alignment mark fluctuates significantly, a phenomenon accompanied by movement of the center of gravity of the diffracted focus of the focused light beam occurs. For example, in the example of FIG. 10 described above, the spot center of gravity of the sensor surface formed by each of the grating lens on the mask and the grating lens on the wafer moves, which causes a decrease in alignment accuracy. This causes a variation in the center of gravity of the diffracted light beam even in other physical optical elements. To solve this problem, it is possible to position the irradiation light with respect to the mark with high accuracy or to enlarge the irradiation spot diameter on the mark and use the light of the central portion. However, the former requires a complicated mechanism for measuring the light projecting and receiving optical systems, and the latter reduces the signal light reaching the sensor surface and illuminates the unmarked portion with illumination light to eliminate scattered noise components from it. It has the drawback of increasing the S / N ratio.

In view of the above-mentioned drawbacks of the conventional example, the present invention detects a deflected light beam by a physical optical element such as a Fresnel zone plate on an object to detect the position of the object, even if the positional relationship between the beam and the mark varies. Detection accuracy is less likely to deteriorate, and
It is an object of the present invention to provide a position detection device that is unlikely to cause a decrease in S / N.

[Means for solving problems]

In order to solve the above-mentioned problems, the present invention provides a plurality of light beams and superimposes a part of each light beam to form a light beam having a cross-sectional shape corresponding to the shape of a physical optical element. The distribution is made substantially uniform without lowering the energy of, and the emitted light flux does not fluctuate even if there is some fluctuation between the mark and the beam.

That is, compared with the case where a plurality of light fluxes are set at a predetermined interval and a combined light flux is applied to the alignment mark and a single light flux is applied, the irradiation energy distribution is closer to uniform within the alignment mark. At the same time, the alignment is arranged so that the energy is significantly reduced outside the alignment mark. Therefore, without lowering the S / N of the system that detects the center of gravity of the luminous flux,
It is possible to provide a positioning device that does not deteriorate the positioning accuracy even if the setting accuracy of the positioning mark of the pickup housing including the light receiving system is poor. This conceptual diagram is shown in FIG.

〔Example〕

FIGS. 1 (a), (b) and (c) are views of a mask aligner provided with a position detecting device of the first embodiment which best shows the features of the present invention. FIG. 1 (a) shows the present embodiment. FIG. 1B is a schematic diagram of the position detecting device of the example, FIG. 1B is the same principle diagram, and FIG. 1C is a detailed view of the position detecting mark portion in the present embodiment.

The mask 101 is attached to the membrane 97, and it is supported by the aligner body 95 via the mask chuck 96. A mask-wafer alignment head 94 is arranged above the main body 95. In order to align the mask 101 and the wafer 102, alignment marks 41M and 41W are printed on the mask 101 and the wafer 102, respectively.

1 and 1'are semiconductor lasers which are monochromatic light sources, 2 and 2'are condenser lenses for condensing the light flux of the semiconductor lasers,
Reference numeral 3 is a half mirror for combining the two light fluxes.

The light beams emitted from the light sources 1 and 1'are collimated by the light projecting lenses 2 and 2 ', respectively, and the beam splitter 92
And is incident on the exposure area 43 including the alignment mark 41M. The alignment mark 41M is a transmissive zone plate and has the function of a convex lens that focuses light on the point Q. Wafer alignment mark 41W is a reflection type zone plate and point Q
It has the function of a convex mirror that forms an image of the light condensed on the detection surface 90 of the photoelectric detector 38.

When the wafer 2 is laterally displaced from the mask 1 by Δσ _W under such an arrangement, the positional deviation Δδ _W of the light amount center of gravity of the light flux on the detection surface 90 is expressed as follows.

That is, The amount of displacement is doubled.

For example, if the mark 41W-Q interval a _W = 0.5 mm, and the mark 41W-detection surface interval b _W = 50 mm, then it becomes 99 times. At this time, Δδ
_W whereas .DELTA..sigma _W, is proportional As apparent from the equation, if degradation farming sensor is 0.1 [mu] m, position shift amount .DELTA..sigma _W to be detected it is position resolution of 0.001 [mu] m.

In this way, the amount of positional deviation Δσ _W is obtained by the photoelectric detector 38, and based on this amount, the wafer stage driving means (not shown) moves the wafer stage 102a in the direction for correcting the deviation by the amount of positional deviation Δσ _W. As a result, the mask and the wafer can be accurately aligned.

By combining the two light fluxes, the light flux radiated on the mask 101 has a substantially uniform light intensity in the longitudinal direction (x direction) of the mark as shown in FIG. 1 (b). Become. Therefore, even if the incident position of the light beam is slightly shifted in the vertical direction in the drawing, the symmetry of the light intensity of the portion of the light beam which is irradiated on the mark is practically maintained. Therefore, the fluctuation of the diffracted light due to the above-mentioned deviation of the light flux hardly occurs as shown in FIG. Also, the area that illuminates the periphery of the mark is
This is sufficiently smaller than the case where the spot diameter of the irradiation light beam is enlarged and the mark is irradiated at the central portion where the fluctuation of the diffracted light is small. Therefore, the amount of ambient light including scattered light from the peripheral portion can be reduced. The marks 41M and 41W have a lens function only in the x direction as a deflection function, and have a function in the y direction to deflect the traveling direction of the light beam at a constant angle in accordance with the diffraction order.

 FIGS. 2 and 3 show another embodiment 2 and 3, in which the polarization
Combined using beam splitter 7 and half-wave plate 8
It is an example. These are half mirrors 3 in the first embodiment.
To remedy the drawbacks of losing energy
You. That is, the polarization plane of a semiconductor laser is generally far
And the polarized beam
Due to the characteristics of the printer, it can transmit P-polarized light and reflect S-polarized light.
, And the direction of the active layer of the semiconductor laser and a half wave
Select the position of the long plate as shown in Fig. 2 or 3 and
-The major axis side and the minor axis side of the field are
It can be set without touching. Also, in the optical path after the two-beam combination, the reflection side
Is a S-polarized light beam whose main ray is 6 and is 6'on the transmission side.
Is a P-polarized light, so both are circularly polarized
In order to do so, the quarter wave plate 9 is installed. this is
The lattice spacing of the physical optical element that is the alignment mark is the wavelength
When the neighboring ones are included, it depends on the grating direction and the polarization direction.
If you want to avoid the phenomenon that the diffraction efficiency shows different characteristics,
This is a particularly preferable means. Lattice spacing is much longer than the wavelength
In this case, the quarter wave plate 9 need not be set. Furthermore, another
As an embodiment of the present invention, as shown in FIG.
It is also possible to combine the partial reflecting mirror 10 and the synthetic lens 11 in the excess part.
Yes. Also, in FIGS. 1 to 4, the chief ray of each light beam is
Since the image size on the sensor surface is not increased, it is almost parallel.
Preferably. The one in which the light flux is not heavy will be specifically described.
The emission angle of the semiconductor laser should be full-width half-maximum intensity and flat with the active layer.
Go to the side θ , The vertical side is θ_⊥Is shown for convenience,
The distance between the light source and the principal plane of the light source side of the condenser lens is g.
You. A so-called collimator that forms a geometrical optical source image at infinity
In the above condition, g is the focal length f of the condenser lens.
You. On the other hand, let the light intensity distribution of a semiconductor laser be I (x, y).
And then e^-2Ω is the radius determined by the strength_X,as well as
ω_yThen, in generalIs represented by

 Therefore, the radius of the half-value intensity on the principal plane of the condenser lens is gt
an {θ / 2} and gtan {θ_⊥/ 2}, so x-axis and y-axis
Respectively θ Direction and θ_⊥If you think according to the direction,That is, ω 1.7 tan (θ / 2, ω_⊥1.7 tan (θ_⊥/ 2).

 Thus ω , Ω_⊥After obtaining the
Contact vol.22, No7 ('84), P53-P65
An arbitrary lens system can be assembled using mathematical expressions for the characteristics of the laser beam.
It is possible to consider the intensity distribution when they are combined.

In the configuration shown in FIGS. 1 to 4, the alignment mark has a beam alignment direction (x direction), where the beam waist radius of the irradiation beam in the alignment direction is ω ₀ , and the alignment mark size in the beam alignment direction is When the projection system irradiates the beam, the amount (that is, the irradiation width) that is added to the extra width and uncertain width (generally about 5 to 20 μm on each side) that is widened with a margin, is 2a. The chief ray spacing of
If it is 2t, it is preferable to set it as cosh (4at / ω ₀ ² ) exp {2 (a / ω ₀ ) ² }. This is a signal as shown in Fig. 8 and the illuminance at the center A of the irradiation part is And the illuminance at the irradiation point C is Therefore, the illuminance is equal to that in the setting satisfying the above expression, which means that the energy loss is small and the uneven illuminance is small. Practically Moderate conditions are preferred.

As is clear from the numerical evaluation of the distribution, when the lower limit is exceeded, the beam spacing is relatively narrow and the illuminance around the mark is reduced. Conversely, when the upper limit is exceeded, the beam spacing is relatively low. The illuminance in the central portion is widely reduced, which deviates from the appropriate range. As a concrete example of the numerical values, if the beam waist size ω ₀ is set to 170.1 μm and the illumination area radius a is set to 140 μm, the beam interval 2t is 210 μm.
And the illuminance at the position point, which is almost the peak value in FIG. Therefore, the ratio of: or: is 1: 0.89, and there is little unevenness, and the energy irradiated in 2a is
High efficiency can be achieved at 66%.

The above is an example of two light sources, but as shown in FIG. 8, three light sources may be used depending on the mark size and the characteristics of the semiconductor laser. In FIGS. 2 and 3, an optical system (usually referred to as a beam expander) having an angular magnification for reducing a beam by an afocal system is added to the system after the light flux combination.
Providing or vice versa is optional depending on design needs.

Further, it is obvious that the present invention can be applied to other position detecting devices for detecting a light beam polarized by a physical optical element, for example, the device described in the conventional example.

〔The invention's effect〕

As described above, according to the present invention, when detecting the position of an object by detecting the light beam deflected by the physical optical element, it becomes possible to mitigate the deterioration of accuracy even if the light beam shift occurs.

[Brief description of drawings]

FIGS. 1 (a), (b) and (c) are respectively a structural diagram, a principle diagram, a mark portion detailed diagram of the position detecting device of the first embodiment of the present invention, and FIGS. 2 (a) and (b) are FIG. 4 is a diagram of a light flux combining section of the position detecting device according to the second embodiment of the present invention, and FIGS. 3A and 3B are diagrams of a light flux combining portion of the position detecting device according to the third embodiment of the present invention. Is a principle diagram of a position detecting device according to a fourth embodiment of the present invention, FIG. 5 is a diagram illustrating a light emission pattern of a semiconductor laser, FIG. 6 is a diagram illustrating conventional problems, and FIG. 7 is a diagram illustrating the present invention. FIG. 8 is a diagram for explaining the concept of the effect, FIG. 8 is a diagram of an example of the light intensity distribution of the light fluxes synthesized by the present invention, and FIG. 9 is a principle diagram of the position detecting device of the fifth embodiment of the present invention. FIG. 11 is a block diagram of a conventional position detecting device, and FIG. 11 is a principle diagram of the same. In the figure, 1,1 ′, 1 ″ …… Semiconductor laser 2,2 ′, 2 ″ …… Condenser lens (or collimator lens) 3,3 ′ …… Half mirror 5,5 ′ …… Mark center line 6, 6 ', 6 "... Principal ray of projected light beam 7 ... Polarizing beam splitter 8 ... Half wave plate 9 ... Quarter wave plate 10 ... Partial reflector 11 ... Synthetic lens is there.

Claims (1)

(57) [Claims] 1. A means for irradiating a physical optical element provided on an object to be aligned with a light beam, and a light beam irradiated by the irradiation means and deflected by the physical optical element to detect the object. And a means for detecting a position, wherein the irradiation means forms a light beam having a cross-sectional shape corresponding to the shape of the physical optical element by partially overlapping a plurality of light beams, and irradiates the light beam. Position detection device.