JP2827312B2 - Position detection device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a position detecting device, for example, in an exposure apparatus for manufacturing a semiconductor element, on a first object surface such as a mask or a reticle (hereinafter, referred to as a “mask”). Position detection suitable for performing relative positioning (alignment) between a mask and a wafer when performing proximity exposure for exposing and transferring a fine electronic circuit pattern formed on a second object surface such as a wafer. It concerns the device.

2. Description of the Related Art Conventionally, in an exposure apparatus for manufacturing a semiconductor, relative positioning between a mask and a wafer has been an important factor for improving performance. In particular, in recent aligners in an exposure apparatus, for high integration of semiconductor elements,
For example, one having a positioning accuracy of sub-micron or less is required.

In many alignment apparatuses, a so-called alignment pattern for alignment is provided on a mask and a wafer surface, and both alignments are performed using positional information obtained from the alignment patterns. As an alignment method at this time, for example, the amount of deviation between the two alignment patterns is detected by performing image processing, or alignment is proposed as proposed in U.S. Pat. No. 4,037,969 or Japanese Patent Application Laid-Open No. 56-157033. This is performed by irradiating the zone plate with a light beam from a light source such as a laser using a zone plate as a pattern, and detecting the position of a condensing point on a predetermined surface of the light beam emitted from the zone plate.

Generally, an alignment method using a zone plate has a feature that relatively high-precision alignment can be performed without being affected by a defect in an alignment pattern, as compared with a method using a simple alignment pattern.

When using a physical optical element such as a Fresnel zone plate to perform relative positioning between a mask and a wafer, the light source is a He-Ne laser or semiconductor laser with high luminance, good directivity, and good coherency. Etc. are used.

However, a light beam from such a laser is irradiated onto a mask or alignment mark provided on a wafer surface for alignment, and signal light from the alignment mark is received by a sensor arranged on a predetermined surface to perform alignment. If
Interference occurs between the reflected light from the alignment pattern on the mask surface and the reflected light from the alignment pattern on the wafer surface, or interference caused by scattered light from the alignment pattern on the mask surface and the wafer surface So-called speckles occur on the sensor surface.

In addition, unnecessary light such as alignment signal light and scattered light may cause interference, and similarly, optical noise may be generated. When such optical noise increases, the S / N ratio of the output signal from the sensor deteriorates, which is a major cause of lowering the positioning accuracy.

In particular, when the distance between the mask and the wafer is very small as in a proximity type exposure apparatus, which is one of the methods for manufacturing a semiconductor device, reflected (diffractive) light from the mask and reflected (diffractive) light from the wafer. And it becomes very easy to interfere.

(Problems to be Solved by the Invention) In view of the above-mentioned drawbacks of the conventional example, the present invention has a high accuracy with almost no influence of interference (reflection) by reflected (diffraction) light from each of a mask and a wafer in proximity exposure. Another object of the present invention is to provide a position detecting device capable of detecting a relative position between a mask and a wafer.

For example, according to the present invention, a light beam is made incident on a first object on which a pattern to be transferred is formed and an alignment mark provided on a second object surface on which the pattern is to be transferred, and a light beam from the alignment mark is utilized by using the light beam from the alignment mark. When detecting a relative position between the first object and the second object, a predetermined surface on which a sensor or the like is arranged by adjusting a wavelength width of a light beam from a light source, for example, a light beam from a laser, reducing coherency of the light beam It is an object of the present invention to provide a position detection device capable of reducing the occurrence of speckles above, improving the sensor S / N ratio, and enabling highly accurate position detection.

(Means for Solving the Problems) The position detecting apparatus of the present invention is arranged such that a first object having a pattern to be transferred and a second object to which the pattern is to be transferred are opposed to each other in order to perform proximity exposure. In a position detecting device for detecting the relative position between the first object and the second object, when the wavelength of the light flux from the light source means is λ nm for the first object and the second object, the full width at half maximum of the spectrum is (λ ^2/20
A light beam of 0000) nm to 60 nm is irradiated, light beams from the first object and the second object are detected by light beam detecting means, and the first object is detected by position detecting means using an output signal from the light beam detecting means. A relative position between the first object and the second object is detected.

In particular, in the present invention, the light source unit is configured to include a laser light source and a light source drive control unit that controls a driving current value of the laser light source to control a full width at half maximum of a spectrum of a generated light beam, or the light source unit. Is composed of a super luminescent diode.

(Example) This is one of the semiconductor element manufacturing methods using the position detection device of the present invention. In the proximity exposure, assuming that the distance g between the mask and the wafer is, for example, 50 μm, the optical path difference between the reflected (diffracted) light from the mask and the reflected (diffraction) light from the wafer is 2 g = 100 μm.

On the other hand, the spectrum width is relatively wide as compared with other laser light sources. That is, a semiconductor laser having a short coherence length, for example, HL7832G (trade name; manufactured by Hitachi, Ltd.)
The spectral width of the laser is the full width at half maximum Δλ = 2 mm,
At this time, the coherent distance l _c ≒ 312 μm. Therefore, the coherence length l _c is to become larger than the optical path difference 2g, reflected from the mask (diffraction) becomes coherent relationship between light and reflected from the wafer (diffraction) light, hence the interference therebetween may occur.

On the other hand, when a very short coherence light source, such as a white light source, is used as a light source for illuminating a physical optical element such as a Fresnel zone plate, the position detecting light flux converged on the sensor surface by the physical optical element is a light flux. Blurring due to chromatic aberration due to the spread of the spectrum is generated. Therefore, the spot diameter on the sensor surface is not sufficiently reduced, and the S / N ratio is deteriorated due to low energy density.

Therefore, the present invention is characterized in that the relative position between the mask and the wafer is detected by using a light beam having coherence that gives the best position detection accuracy in proximity exposure. Hereinafter, specific examples will be described.

FIG. 1A is a schematic diagram of a position detecting device according to a first embodiment of the present invention using a zone plate.

In the figure, a parallel light beam emitted from a light source 10 passes through a half mirror 74, and is condensed at a converging point 78 by a converging lens 76. Then, a mask alignment pattern 3M on a mask M surface as a first object and a support A wafer alignment pattern 4W on the surface of the wafer W as a second object placed on the table 62 is irradiated. These alignment patterns 3M and 4W are configured by reflection type zone plates, and form light converging points on a plane orthogonal to the optical axis including the light converging point 78. At this time, the amount of shift of the condensing point on the plane is detected by guiding the light to the detector (also referred to as a sensor) 8 by the condensing lenses 76 and 80.

Then, based on the output signal from the detector 8, the control circuit 84
The drive mechanism 64 drives the mask M and the wafer W to perform relative positioning between the mask M and the wafer W.

FIG. 1A is an explanatory diagram showing an image forming relationship of a light beam from the mask alignment pattern 3M and the wafer alignment pattern 4W shown in FIG. 1B.

In the figure, a part of the luminous flux diverging from the focal point 78 is diffracted from the mask alignment pattern 3M, and a focal point 78a indicating the mask position is formed near the focal point 78.
Further, a part of the other light flux passes through the mask M as the zero-order transmission light, and enters the wafer alignment pattern 4W on the wafer W without changing the wavefront. At this time, the luminous flux is diffracted by the wafer alignment pattern 4W, and then the mask M
Is transmitted as zero-order transmitted light, and is condensed in the vicinity of a converging point 78 to form a converging point 78b representing a wafer position. In the figure, when the light beam diffracted by the wafer W forms a focal point, the mask M acts as a simple transparent state. Thus, in this embodiment, one converging point is formed by one zone plate.

The position of the focal point 78b due to the wafer alignment pattern 4W thus formed is determined along the plane orthogonal to the optical axis including the focal point 78 according to the deviation Δσ of the wafer W with respect to the mask M. The deviation amount Δ of the amount corresponding to Δσ
σ '.

In the position detection device shown in FIG. 1, when calculating the relative displacement, light from a mask and a zone plate provided on the wafer surface are independently imaged on a predetermined surface to be evaluated, and both the image formations are performed. The amount of deviation from the relative positional relationship as a reference for the image point, that is, when there is no deviation in the mask / wafer, is determined. The relative positional relationship serving as a reference at this time is determined by design values, but may be determined by trial firing.

 Next, the light source 10 used in the present embodiment will be described.

Generally, laser light has high coherence. For example, commercially available He-
The Ne laser has three to four longitudinal modes, and the coherence length is about several tens of cm. When the semiconductor laser is driven with an optical output of 1 mW or more, the coherence length becomes several hundred cm or more. Therefore, when the laser light having good coherency is used as the light source in the alignment apparatus having the configuration as in the present embodiment, the reflected light from the mask surface and the wafer surface, the scattered light from the alignment mark, and the alignment signal light Unnecessary light and the like other than the above interfere with each other to generate interference fringes and speckle patterns on the sensor surface.

For example, as shown in FIG. 2, a light beam 47 from the light source enters the sensor 8 as an alignment signal 47 'via the mask M and the alignment patterns 3M and 4W on the wafer W. Light reflected (diffracted) from the surface enters the sensor 8 as unnecessary light 47 ", and interferes with the alignment signal light 47 'to generate interference light.

The interference light at this time causes a reduction in the S / N ratio of the sensor 8 and a significant reduction in the positioning accuracy.

Next, the intensity distribution of the interference pattern on the sensor surface in this embodiment will be described with reference to FIG.

As shown in FIG. 3 (A), when the alignment signal light 47 'for positioning and the unnecessary light 47 "overlap with interference on the sensor 8, the light intensity distribution on the surface of the sensor 8 generally increases. Third
As shown in FIG. 7B, the light has interference light noise that changes at random.

The complex amplitude of the signal light at a certain point on the sensor Complex amplitude of unwanted light Then, the composite of the complex amplitude is Here, As is the amplitude of the signal light, AN is the amplitude of the unnecessary light, ω is the frequency of the light, φ ₁ is the initial phase of the signal light, and φ ₂ is the initial phase of the unnecessary light. Is the square of Becomes As described above, when the signal light and the unnecessary light are combined, in addition to the intensity As ² when only the signal light is received and the intensity AN ² when only the unnecessary light is received, the intensity due to the interference effect of the third term is obtained. That is, a term of 2 · As · ANcos (φ ₂ −φ ₁ ) appears. In the above case, the signal light and the unnecessary light are completely coherent, but the third term is given by .2 · As · ANcos · (φ ₂ −φ
₁ ) · γ (0 ≦ γ ≦ 1), γ = 1 when two lights are completely coherent, γ = 0 when they are completely incoherent, and 0 <when they are partially coherent.
Assuming that γ <1, the strength at that time can be obtained appropriately.

Generally, in the case of proximity exposure, the distance g between the mask alignment mark 3M and the wafer alignment mark 4W is 10
μm to several hundred μm. In this range, when the semiconductor laser is output at several mW, the degree of coherence γ is γ
≒ 1. Therefore, when the signal light and the unnecessary light overlap on the sensor surface, the intensity of the light thereat becomes A ² = As ² + AN ² + 2 · As · ANcos (φ ₂ −φ ₁ ) · γ (γ ≒ 1), and γ ≒ When the value is 1, the effect of the third term cannot be ignored. For example, the phase difference φ ₂ −φ ₁ between the signal light and the unnecessary light is λ / 2 ≒ 0.4 μm when λ = 0.8 μm.
A ² is randomly between the ^{A 2 = As 2 + AN 2} -2 · As · AN and ^{A 2 = As 2 + AN 2} +2 · As · AN to change by π at a relative positional change of two beams degree Will fluctuate.

When two or more light beams having coherence with each other overlap on the sensor 8 as shown in FIG. 3B in this manner, the light intensity of the sensor 8 varies by a mask-wafer interval of about 0.1 μm. However, since the shape greatly changes under the influence of the interference effect, the position detection accuracy is significantly deteriorated.

On the other hand, when γ = 0, that is, when the signal light 47 ′ and the unnecessary light 47 ″ are completely incoherent, the light intensity distribution on the sensor surface is as shown in FIG. Since the intensity is 47 "and the influence of interference is eliminated, even if the signal light 47 'and the unnecessary light 47" overlap on the sensor surface, the influence of the unnecessary light can be removed by electrical signal processing. For example, the signal component of the unnecessary light 47 "is subtracted from the sum signal of the signal light 47 'and the unnecessary light 47" by knowing the intensity distribution of the unnecessary light 47 "in FIG. A signal output can be obtained.

Next, the condition of the spectral width (full width at half maximum of the spectrum) Δλ for the coherence of the reflected light from the mask and the reflected light from the wafer to be γ = 0 when the distance between the mask and the wafer is g will be described.

Generally, there is a relation of l _c = λ ² / Δλ between the coherence length l _c of light, the wavelength λ, and the spectral width Δλ. For example the aforementioned HL7832G as a semiconductor laser is driven below the threshold, the spectral width [Delta] [lambda] = 16 nm, the coherence length l _c of when the center wavelength 0.79μm in (both measured with an optical spectrum analyzer TQ8345 of by Advantest Corporation) 40 [mu] m This was in good agreement with the theoretical value λ ² / Δλ = 39 μm. In the following discussion, this theoretical formula l _c = λ ² / Δλ is used.

From l _c = λ ² / Δλ, Δλ = λ ² / l _c . On the other hand, the condition for preventing the reflected light from the mask and the reflected light from the wafer from interfering with each other is as follows.
l _c ≦ 2g. Therefore, the condition required for the spectrum width to prevent interference when the mask-wafer distance g is given is Δλ ≧ λ ² / 2g. Here, in order to prevent interference, it is understood that the wider the spectral width is, the better, and the shorter the wavelength, the better.

In the case of a proximity exposure apparatus, it is usually necessary to prevent interference even when the distance between the mask and the wafer becomes 100 μm or less. Conditions to prevent interference in spacing 100μm becomes spectral width Δλ ≧ λ ² / 2g than ^{Δλ ≧ λ 2/200000 ...... (} 1) ( provided that [Delta] [lambda], the unit of lambda is nm).

On the other hand, if the spectral width of the light source is increased to shorten the coherence distance, the physical optical element has wavelength dependence, so that the signal spot on the sensor surface becomes large and the amount of light incident on the sensor is constant. Because of this, the peak value of the signal decreases. Since the noise level does not change, if the spectrum width of the light source is widened beyond a certain condition, the S / N is reduced, and sufficient accuracy in position detection cannot be obtained.

In various position detection devices using physical optical elements such as zone plates, the light output of the light source and the diffraction efficiency of the physical optical element differ depending on each method. As long as it is detected, the S / N due to the spread of the imaging spot when the spectrum width of the light source due to the wavelength dependence of the physical optical element is wide
Is a problem.

This will be described with reference to FIG. 4 which is a partial schematic diagram of the method shown in FIG. 1 (A).

In FIG. 4, 3M (4W) is a physical optical element on a mask (or wafer). Reference numeral 78 denotes an image forming point of the signal spot, and a surface 78a having this point is conjugate to the sensor surface position.
a is the size of the physical optical element, w is the spot diameter when the light source has no spectral spread, w 'is the spot diameter at the image forming position when the light source has a spectral spread of Δλ,
w ″ indicates the spot diameter at the conjugate position 78a with the sensor surface when the light source has a spectral spread of Δλ, and Δ indicates the amount of change in the focal length of the physical optical element 3M when the wavelength changes by Δλ.

In FIG. 4, when the light source has no spectrum spread, the spot diameter w at the conjugate position 78a with the sensor surface is described as w = (1.4λf) / a, where λ is the wavelength of the light source.
Here, f is the focal length of the physical optical element 3M (4W) at the wavelength λ.

On the other hand, consider the case where the light source has a spectrum spread of Δλ. Wavelength from λ When the focal length of the physical optical element 3M changes to And the spot position changes from the sensor position 78 as a result. And the spot diameter w 'there is Becomes

At this time, if the spot diameter at the conjugate position 78a with the sensor surface is w ″, Than It is.

This wavelength Is approximately equal to the spot diameter of the light beam having the spectral width Δλ.

If the spectral width of the light source is The change in spot diameter at the sensor surface conjugate position 78a Is described.

Now, assuming that the total signal light amount on the sensor surface is the same with and without the spectral spread, the peak value of the signal output becomes inversely proportional to the spread of the spot diameter because the spot diameter spreads from w to w ″. / w ″ times. In other words, the energy density of the signal light on the sensor decreases, and the S / N
Deteriorates and sufficient accuracy cannot be obtained.

Next, the upper limit of the spectrum width required for obtaining sufficient accuracy will be described in consideration of the decrease in S / N due to the spread of the spectrum width.

In FIG. 1 (A), enlarged projection is performed on the sensor 8 by the lenses 76 and 80, and the physical optical element 3M is used to determine the position.
(Or 4M) is preferably reduced to some extent.

More specifically, the spot diameter is set to (1.4λ
f) / a, which is preferably 10 μm or less.

On the other hand, in the production of a physical optical element such as a Fresnel zone plate, the longer the f, the larger the pattern pitch can be.
Although it is more advantageous as the length is longer, considering the above conditions, it is generally considered that the focal length f of the physical optical element is 1 mm or less. In an optical semiconductor device as a light source, particularly in a semiconductor laser, an SLD, or the like, a wavelength in the vicinity of 0.8 μm is considered as the most common wavelength region. The size a of the physical optical element is larger than the occupied area allowed for the alignment mark on the wafer.
A thickness of about 100 μm is generally considered.

As a general example in the position detection method of FIG. 1A, f = 1 mm, λ = 0.85λ, and spectrum width Δλ = ± 3
Assuming 0 nm, the ratio w ″ / w of the spread of the spot diameter when the spectral width Δλ = ± 30 nm is about 1.25 as compared with the case where the spectral width is not spread, which indicates that the spread of the spectral width is about ± 30 nm. If it is spread, the peak value of the signal light on the sensor is reduced by about 20%, so that the S / N ratio is deteriorated, which has an unfavorable effect on obtaining sufficient position detection accuracy.

Therefore, in the position detection method as shown in FIG. 1A, it is generally preferable that the light source has a spectral width of 60 nm or less. That is, the condition is Δλ ≦ 60 (nm) (2). (1), (2) one condenser by ^{λ 2/200000 ≦ Δλ ≦ 60} ( unit nm) in system to form a spot effect of interference light of physical optic element in a proximity exposure from the equation, the spot enlargement Is obtained as an appropriate range for reducing the decrease in the S / N ratio due to

Next, the light source drive control means 100 controls the oscillation wavelength width of the light beam from the light source 10 to be within the above-described range, lowers the coherency of the light beam, and influences the S / N of the interference light generated on the sensor surface. An embodiment of a method for reducing the deterioration of the ratio will be described.

When using a laser, for example, as a light source and reduce the effects of the interference light generated on the sensor surface driven at 1.1 times or less of the drive current to the threshold current value I _th upon driving of the laser ing.

FIG. 5A is a schematic diagram showing a driving region of a driving current at this time using a semiconductor laser in this embodiment. When the laser threshold current value of the area S1 and the area S2 are in the drawing was I _th, the driving current value is divided by the boundary point at which the 1.1 × I _th. In general, S1 indicates a region having a small coherency, and S2 indicates a region having a large coherency as compared with the region S1.

The as described above the laser is driven by the following 1.1i _th driving current as shown in areas S1 thereby achieving a reduction in the coherency in the present embodiment.

FIG. 5 (B) is an explanatory diagram of one embodiment showing the oscillation characteristics when a semiconductor laser HL8314G (trade name; manufactured by Hitachi, Ltd.) is used as a light source in this embodiment.

In the figure, the horizontal axis represents the current value applied to the semiconductor laser, and the vertical axis represents the light output at each current value. When the optical output of the semiconductor laser exceeds a threshold current value I _th (about 46 mA), it suddenly starts increasing. As shown in the figure, the semiconductor laser HL8314G (hereinafter “H
L8314G. " ) Has a threshold current value I _th of about 46 mA. In the same figure, the driving current value I is 40 mA, 46 m for reference.
Examples of the spectral distribution of the oscillation wavelength at A, 48 mA, and 54 mA are shown at 41, 42, 43, and 44, respectively.

Although the spectral distribution of the semiconductor laser light shown in the figure is a multi-mode up to about the threshold current I _th, I = 54mA
When an applied current of (≒ 1.2I _th ) (optical output is 4.5 mW) is applied, a single mode laser such as HL8314G oscillates in a single mode. It is a multi-mode oscillation up to I = 1.1I _th in HL8314G. The coherent length measured with an optical spectrum analyzer (TQ8345 manufactured by Advantest Co., Ltd.) is about 60 μm in the spectrum diagram 41 when I = 40 mA (spectrum width is 10.5 nm), and the spectrum diagram 42 when I = 46 mA (spectrum width is 4.5nm) about 120μm, I = 48mA
In Figure 43 (spectral width 0.45 nm), about 12
The spectrum at 44 μm and I = 54 mA (spectrum width 0.3 nm or less) was about 1 mm or more. The light output is I =
It was about 0.1 mW in the spectrum of FIG. 41 at 40 mA, and about 0.5 mW in the spectrum of FIG. 42 at I = 46 mA. FFP (Far Field Pa
ttern), when I = 40 mA or more, θ _N = 30 °,
was θ _⊥ = 20 ° constant.

In this embodiment, the semiconductor laser that oscillates in a single mode above the threshold current value is driven at a threshold current value or less in the present embodiment, whereby the spectrum is multi-mode, and high brightness is maintained while maintaining the FFP directivity of the emitted light. It is used as a light beam with low coherency and high S / N ratio with low coherence.

In addition, in the present invention, a super luminescent diode (SLD, also referred to as “super radiant diode”) having low coherency can be used as a light source.

6 (A) and 6 (B) show the directional characteristics of the emitted light in the direction perpendicular and parallel to the active layer surface of SLD, L330z (trade name; manufactured by Hamamatsu Photonics), and FIG. The emission spectrum distribution when the current value I is I = 120 mA is shown, and the current-light output characteristics of the SLD are shown in FIG.

SLDs have better directivity of light than general light emitting diodes, have a wide light emission spectrum width despite high light output, and have relatively low coherency, so they are used as a light source for the alignment device according to the present invention. It is suitable.

The full width at half maximum of the emission spectrum width shown in FIG. 6C is about 10 nm, which has coherence corresponding to the spectrum width when I = 40 mA of the semiconductor laser shown in the previous embodiment. At this time, the coherent length is about 60 μm, and when the distance between the mask and the wafer is set to 30 μm or more, the deterioration of the S / N ratio of the position detection signal due to the interference between the position detection signal light and unnecessary light is good. Can be prevented.

If a current of about I = 130 mA is passed, the spectrum width is almost the same and a value of 2 mW or more can be obtained as the optical output. In addition, the size of the light emitting portion of the SLD can be reduced to a small area of 6 μm × 2 μm or less, and it can be used as a light source having high brightness and small coherence in combination with good directivity of the emitted light beam. Has features.

FIG. 7 is a schematic view of a main part of another embodiment when the present invention is applied to an exposure apparatus for manufacturing a semiconductor by a so-called proximity method.

In the figure, M is a mask, and W is a wafer, which respectively correspond to a first object and a second object for performing relative positioning. Three
M 'is a mask alignment pattern on the mask M surface and corresponds to the first physical optical element, and 4W' is a wafer alignment pattern on the wafer 4 surface and corresponds to the reflective second physical optical element.

In this embodiment, the alignment patterns 3M 'and 4W'
Constitute a so-called uneven alignment system composed of physical optical elements each having a condensing and diverging action. In the figure, a light beam emitted from a light source 10 having a variable wavelength width is converted into a parallel light beam by a light projecting lens system 11, and the mask alignment pattern 3M 'is irradiated via a half mirror 12. The mask alignment pattern 3M 'converts the incident light beam to a point Q on the wafer W.
Consists of a zone plate for focusing. The light beam condensed at the point Q is on the way to the point Q and enters the wafer alignment pattern 4W '.

The wafer alignment pattern 4W 'is composed of a reflection-type zone plate. After reflecting the incident light beam and passing it through the mask M and the half mirror 12, the detection surface 9 of the detector 8 is detected.
Focusing on top.

The position of the center of gravity of the converging point is obtained by the position detection circuit 24, and a position signal is sent to the control device 23.

Here, the center of gravity is defined as a zero vector when a value obtained by multiplying a position vector of each point in the detection plane from this point by the light intensity of each point in the light beam detection plane is integrated over the entire detection plane. It is a point.

The control device 23 evaluates the amount of wafer displacement and sends a positioning signal to the wafer stage controller 22. Then, the wafer stage 21 is moved by the wafer stage controller 22 to correct the positional deviation between the mask and the wafer. Reference numeral 100 denotes a light source drive control unit that controls the oscillation wavelength width of the light beam from the light source 10 as described later. 15 is an aligner body, 16 is a mask chuck, 17 is a membrane, and 20 is a wafer chuck.

With the above configuration, when the wafer W shifts by Δδ with respect to the mask M, the exit angle of the light beam that has been subjected to the lens (light focusing) action by the mask alignment turn 3M ′ and the wafer alignment pattern 4W ′, respectively. And the focal point shifts on the detection surface 9. When the emission angle is small, the mask M and the wafer W are shifted by Δσw in the parallel direction, and the focal length of the alignment mark of the mask is set to a
w, assuming that the distance from the converged light flux passing through the alignment mark 4W 'of the wafer to the converging point is L, the center-of-gravity shift amount Δδw of the converging point on the detection surface 9 is Becomes For example, as an appropriate example, if the distance g between the mask M and the wafer W is g = 30 μm and aw = 214.7228 μm L = 18657 μm, a sensitivity of -100 times can be obtained.

That is, the present embodiment has a feature that if the measurement can be performed with an accuracy of Δδw = 1.0 μm, the positional deviation amount Δσw can be evaluated up to 0.01 μm.

At this time, the focal length bw of the wafer is -186.570 μm.

A specific method for detecting the amount of displacement will be described. At the time of mask setting, the center of gravity position where there is no misalignment between the mask and the wafer due to trial printing is determined as a reference position, and at the time of position detection, how much the center of gravity position deviates from the reference position in the x direction is detected. The relative displacement amount of the mask / wafer can be obtained from the equation (3). As described above, in this embodiment, the physical optical elements 2 formed on the respective surfaces of the mask and the wafer are used.
One condensed point is formed by the two.

Also in the present embodiment, the light source drive control unit 100 controls the oscillation wavelength width of the light beam from the light source 10 so as to fall within a range that reduces the influence of interference and the deterioration of the S / N ratio due to the increase in the spot diameter. As the light source 10, the aforementioned semiconductor laser,
SLD and the like. The light source drive control means 100 controls the light source so as to have a spectrum width as specifically described below.

In the position detection device as shown in FIG. 7, an allowable spectrum width is shown for obtaining sufficient position detection accuracy. An acceptable reduction in the energy density of the signal light on the sensor is that the signal level is reduced by 20% from the peak value as described above.

In FIG. 7, the aw = 214.7228 μm and the bw = −186.57, respectively, as shown in the numerical examples of the focal lengths of the mask and the wafer.
0 μm, mask-wafer gap δ = 30 μm, wafer-sensor distance L = 18657 μm, and light source wavelength SLD
It is set to a general value of 0.85 μm.

Under this condition, the full width of the spectrum is 40 nm, and the half value is 20 nm. Therefore, in the following, the calculation is performed with Δλ / 2 = 20 nm.

The focal length f of the Fresnel zone plate changes the wavelength by Δ
Assuming λ, the changed focal length f ′ is Aw = 214.7228 μm, bw = −182.9118 μm Substituting λ = 0.85 μm gives aw = 209.6705 μm, bw = −178.6080 μm, Thus, the geometrical optical point is 30202.9 μm.

At this time, the sensor naturally remains at the value of 18657 μm shown above, so that the spot on the sensor has the diameter a in the alignment direction of the exit window of the Fresnel zone plate of the wafer.
The spot is only 0.3823a. Therefore, if a is 100μ
Assuming m, the value of 0.3823a is 38 μm. This is λ = 0.8
Since the wave optical spot diameter at 5 μm is 1.4λF (F is the F number and F = f / a), 1.4 × 0.85 × 18657/100 ≒ 222 [μm], which is the sum of 222 μm and 38 μm It changes to about 260 μm. That is, it means that the peak value is spread by 17%, and the decrease in the peak value is inversely proportional to the spread of the spot diameter, so that the peak value is reduced by about 15%.

In other words, when the spectrum width is ± 20 nm, the peak value is 15%
About a degree. If you remove from this numerical example, the peak value will be 20
The decrease by about% is when the half width of the spectrum is ± 30 nm.
That is, when the spectrum width is ± 30 nm, the peak value is about 80%, and the necessary level for securing S / N of signal detection can be secured.

If the spectrum width exceeds this range, the above-mentioned peak value will be 80% or less, and it will be difficult to secure accuracy especially when estimating a wafer having a large scattering component and a low S / N when the diffraction efficiency of the alignment signal light is low.

Therefore, the half width of the spectrum of the light source is preferably ± 30 nm or less. . Therefore, Δλ ≦ 60 (nm) (4) is a condition for reducing the deterioration of the S / N ratio due to the enlargement of the spot diameter. Conditions for reducing the influence of interference are the same as those in the embodiment of FIG. 1A, and are given by equation (1). Thus (1), (4) from equation ^{λ 2/200000 ≦ Δλ ≦ 60} ( unit nm) is a condition in such a device.

In this embodiment, a light source having a wide spectral width, such as a xenon lamp or a white lamp, can be used as a light source having low coherency.

However, these light sources generally have low brightness, are extremely large in comparison with semiconductor devices having an area of several tens of μm, and have a problem that the spectral width is too large and the spot diameter is not reduced.

For this reason, in the present embodiment, an SLD or the like is used as a light source, and the coherency of the light beam is reduced by the above-described method, thereby achieving a highly accurate alignment device.

(Effect of the Invention) As described above, according to the present invention, the first object and the second object are aligned using the light beams from the alignment marks provided on the first object plane and the second object plane. At this time, by controlling the oscillation wavelength width of the light beam from the light source illuminating the alignment mark as described above, the speckle pattern generated on the sensor surface is reduced, and the S / S of the output signal from the sensor is reduced.
It is possible to achieve a position detection device suitable for proximity exposure that has improved the N ratio and enabled high-luminance alignment.

[Brief description of the drawings]

1 (A) and 1 (B) are schematic views of a main part of a first embodiment of the present invention, FIG. 2 is an explanatory view of a part of FIG. 1, and FIGS.
(B) and (C) are schematic diagrams for explaining light intensity distributions of signal light and unnecessary light on the sensor surface, FIG. 4 is a partial schematic diagram for explaining a spot forming state, and FIG. FIG. 5B is an explanatory diagram showing characteristics of a semiconductor laser, FIG. 6 is an explanatory diagram showing characteristics of a superluminescent diode, and FIG. 7 is a diagram showing the present invention. It is the principal part schematic of 2nd Example of this. In the figure, M is a mask (first object), W is a wafer (second object), 3M is a mask alignment mark, 4W is a wafer alignment mark, 10 is a light source, 11 is a light projecting lens system, 12 is a half mirror, 8 Denotes a detector, 24 denotes a position detection circuit, 23 denotes a control device, 22 denotes a controller, 21 denotes a stage, and 100 denotes light source drive control means.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁶ , DB name) G01B 11/00-11/30 H01L 21/30 G03F 9/00

Claims (3)

(57) [Claims] A first object having a pattern to be transferred and a second object to which the pattern is to be transferred are disposed in close proximity to each other for performing proximity exposure, and a relative position between the first object and the second object is provided. the position detecting device for detecting a, when the wavelength of the light beam from the light source means to said first object and second object and [lambda] nm, FWHM of the spectrum (λ ^2/200000) nm~60nm
Of the first object or the second object is detected by the light beam detecting means, and the first object and the second object are detected by the position detecting means using an output signal from the light beam detecting means. A position detecting device for detecting a relative position of the position. 2. The apparatus according to claim 1, wherein said light source means includes a laser light source and a light source drive control unit for controlling a drive current value of said laser light source to control a full width at half maximum of a spectrum of a generated light beam. 2. The position detecting device according to 1. 3. The position detecting device according to claim 1, wherein said light source means is constituted by a super luminescent diode.