JPH0269603A - Method for aligning mask and wafer - Google Patents

Abstract

PURPOSE:To improve exposing accuracy by utilizing interference fringes having two wavelengths with light having two kinds of wavelengths different from exposure wavelength to fit a pattern projection optical system as alignment light to measure the deviation in the position in the height and cross directions between a mask and a wafer. CONSTITUTION:A mask mark 11 is placed on a point A and the alignment light having Doppler-shifted wavelengths lambda1, lambda1+DELTAlambda is made incident as luminous flux having spherical waves to condense on an incident pupil of a projection lens 20 by using an acoustic optical modulator. Here, diffracted light of zero-order, + or - first-order,... is generated from the mask mark 11 and the first-order diffracted light is projected on two points B and C on the wafer 30 through the projection lens 20 with positional information of the mark 11 of the point A. Wafer marks 31b and 31c with the same grating shape are formed on the above positions of points B and C and moire patterns are produced on the marks and the light again diffracted is led by a mirror 41, etc., and detected by a detector 42. The relative deviation in the position of the height directional position can be measured by carrying out prescribed signal processing by a signal processing circuit 43 based on an output signal of the above-mentioned detector 42.

Description

Detailed Description of the Invention [Invention 1] (Industrial Application Field) The present invention relates to a method for aligning two objects, and particularly relates to a method for aligning two objects, and in particular, the direction of the grooves between a mask and an evaporator used for pattern transfer. The present invention relates to a mask/wafer alignment method suitable for alignment in the height direction.

(Prior Art) In recent years, with the miniaturization and high integration of semiconductor devices such as LSIs, optical reduction projection exposure devices with high resolution performance have been widely used as pattern exposure means. The resolving power R of this device is expressed by Rayleigh's formula %, and the depth of focus σ is expressed by σ-±λ/2NA'...■, where k is a constant and λ is the wavelength of light. , NA is the lens numerical aperture.

In order to improve the resolution from the above equation 0, the lens numerical aperture NA must be increased. In fact, in the conventional equipment, the exposure wavelength was λ-43 [inn, NA-0, 35°σ-±1.8 μm, but recently it has been changed to NA-0,
5σ-±0.9. As the lens numerical aperture NA increases, the depth of focus decreases according to the above equation 0, and therefore the height of the wafer must be determined by API with a high viscosity.

The height of the wafer can be measured using an air micrometer (Mechanical Engineering Handbook, 6th edition PJ7-183) or by passing light through a slit (in contrast, it is incident diagonally and reflected on the wafer). An oblique-incidence height detector that uses the fact that the light is imaged on the detector and the coupling position on the detector changes depending on the top and bottom of the wafer (1985 Seiso Society Spring Conference Academic 5i1 Conference Proceedings) P20g) is widely used.

However, with an air micrometer, the nozzle cannot be installed directly below the projection lens, and the height outside the pattern area must be set at 71p1, so the height of the exposure area and the height of the measurement position may be different. There is sex. In addition, in the case of a light oblique incidence type height detector, height measurement cannot be performed with high viscosity because all of the objects are fixed due to changes in the reflectance of the pattern within the surface to be measured. In other words, in either method, the big problem is that the height of the wafer cannot be measured with high precision.

Note that it is necessary to align the wafer in the lateral direction with respect to the mask with high viscosity, but the lateral position of the wafer can be aligned with high precision using a diffraction grating mark or the like.

(Problem to be solved by the invention) As described above, with the conventional technology, it is difficult to measure the height of the pattern exposed surface (wafer) with high precision due to changes in reflectance due to the pattern and the structure of the air micrometer. can be,
There has been a problem in that the reduction in depth of focus accompanying the increase in lens numerical aperture NA cannot be adequately coped with.

The present invention has been made in consideration of the above circumstances, and its purpose is to accurately determine the vertical and lateral positions of the wafer relative to the mask, regardless of the pattern formed on the wafer. It is an object of the present invention to provide a method for aligning a mask and a wafer that can be aligned and contribute to improving exposure accuracy.

[Structure of the Invention] (Means for Solving the Problems) The gist of the present invention is to align light that zeros out two wavelengths different from the wavelength matching the pattern projection optical system in an alignment method using a grid pattern. used as light,
The purpose of this method is to measure the crystallographic direction misalignment and lateral misalignment between a mask and a wafer by using two-wavelength "1'-interference fringes."

That is, in the present invention, prior to projecting and exposing a desired pattern formed on a mask onto a wafer through a projection lens,
Positioning of the mask and wafer to determine the (11λ·1 positional deviation of the wafer in the direction and height direction of the wafer with respect to the mask, and to determine the position of the mask and the wafer based on the measurement information) In the method, lattice-like marks are provided on the mask and the wafer, respectively, and alignment light having two different wavelengths from the exposure light is irradiated onto the mask mark, and the mask formed by the light irradiation is The diffracted light from the mark is irradiated onto the pair of wafer marks through the projection lens, producing -1- interference fringes of two wavelengths on the pair of wafer marks, and the reflected diffracted light from the pair of wafer marks is irradiated. The amount of positional deviation in the taste direction is determined based on the phase difference between the two reflected diffracted lights, and the average value of the two anti-η・1 diffracted lights at position 11 is calculated. It's a shame for those who tried to find it.

(Function) According to the present invention, 2 f! Detect the anti-f1 diffracted light obtained by shining the alignment light that precludes the growth of I on the mask mark q, + and shining the resulting diffracted light on the wafer mark J1.1. Then, the mask and the sea urchin 1 can be aligned in the direction of the germ direction and the position of the direction of the drawing. Here, the phases of the two reflected diffraction lights change by the same amount and in the same direction as the wafer changes in the direction of one island.
As the wafer changes in the lateral direction, it changes by the same amount in the opposite direction. Therefore, regardless of changes in the orientation of the wafer, the positional deviation in the height direction is determined from the phase difference, and 1110, 1
In addition, it becomes possible to determine the lateral positional shift from the 1' average value of the phase regardless of changes in the height direction of the wafer.

Therefore, it is possible to align the mask and wafer with high viscosity without adding a special mechanism for height detection and without being affected by the pattern formed on the exposed wafer. It can be carried out.

(Example) Hereinafter, details of the present invention will be explained with reference to illustrated embodiments.

FIG. 1 is a schematic diagram showing a reduction projection exposure apparatus used in a method according to an embodiment of the present invention. In this device, a reticle (mask) IO and a wafer 30 are placed above and below the projection lens 20.
will be installed. The projection lens 20 is connected to the mask 10 so that a quadrature relationship is maintained with respect to the exposure wavelength. The wafer 30 is positioned with high accuracy by setting the object-image interval.

In this state, light emitted from point A of the mask 10 when the mask 1o is irradiated with exposure light forms an image on the wafer 30 through the projection lens 20. On the other hand, when alignment light different from the exposure wavelength is used, such as He-Ne laser light, the light emitted from point A of the mask 10 is transmitted to the wafer 3o.
Instead of making a connection at the top, make a connection at point D, as shown by the broken line in the figure.

In this embodiment, a grid mark (mask mark) 1 is placed at point A.
1, and doppler-shifted alignment light having wavelengths λ1 and λ2−λ1+Δλ using a B-port optical modulator (not shown) enters the projection lens 20 as a spherical wave light beam to be focused on the entrance pupil of the projection lens 20. A point mask mark 11
, O-order ±1st-order diffracted light is generated, and the ±1st-order diffracted light passes through the projection lens 20 with the position information of the mask mark 11 at point A to two points B and C on the wafer 30. irradiated. At the two points B and C, a moving moiré pattern (ie, a periodic pattern image depending on the angle of incidence θ) is generated. At this time, the magnification β of the image formed at point D with respect to the alignment light λ1 of the projection lens 20 and the distance Ω between the focal point and the wafer are determined by design.

Therefore, the distance 2r between B and C is determined by the width P of the grid mark 11 formed at point A as shown in the following equation.

Here, λ is the wavelength of the alignment light, n is the diffraction order (±1 in the case of the figure), and θ is the transmission diffraction angle.

Grid marks (wafer marks) 31b and 31c having the same shape are formed at points B and C on the wafer 30, and a moire pattern is generated on these marks. Then, the light diffracted again by the marks 31b and 31c, for example, in the case of the figure, the reflected diffracted light returning vertically is taken out, guided by a mirror 41, etc., and detected by a detector 42. The light incident on the marks 31b and 31C has an incident angle θ'(janθ'-(
1/β) tan θ), the phase of the light detected by the detector 42 changes corresponding to the relative displacement in the lateral direction between the wafer 30 and the mask 10 and the displacement in the height direction of the wafer 30. do. This phase change changes by the same amount and in the same direction for the two detection signals with respect to a change in the height of the wafer 30 in the j direction. However, the drawing h of the wafer 30 and mask 10
For a phase λ·1 displacement in the ° direction, it is assumed that the two IMHIs of the two detection units 1° change by the same amount and in opposite directions.

Therefore, ]/2 of the pitch 11 of the two detection signals does not depend on the change in the width direction of the wafer 30, but becomes the displacement of the pitch 11 in the lateral direction between the wafer 30 and the mask 10.

, 2 detections (No. 1 + 11 ゛1; uniform field is for wafer 30 and mask 10 towards FJ71) I + change 1 to indicate the height displacement of wafer 30 become.

Therefore, by performing predetermined signal processing by the signal processing circuit 43 based on the output signals of the detection units 111 and 442, it is possible to measure the positional misalignment between the positions in the height direction and the position in the lateral direction.

FIG. 2 is a diagram showing a specific circuit configuration fA'j related to the above signal processing. The alignment light is irradiated onto the mask mark 11 on the mask 0, and the half mirror 51 is irradiated with the alignment light.
The light is reflected by the photodetector 52 and detected. This detector 52 provides a reference signal. On the other hand, the reflected and diffracted lights from the wafer marks 31b and 31c on the wafer 30 are detected by photodetectors 53 and 54, respectively. Detector 53
The detection signal of the first position 1 along with the reference signal of the detector 52
1! of the reflected diffracted light from the wafer mark 31b. χ・1 for α light
The phase difference is detected. In addition, the detection 15 of the detector 54
The signal is supplied to the second phase difference detection circuit 56 together with the reference signal of the detector 52, and this circuit 56 detects the wafer mark -3.
The phase difference of the reflected diffracted light from 1c to the reference light is increased by 7-l'l. position +II difference detection circuit 55.56-3
Detection 1. No. J is the adder circuit 57 and ?, respectively. h&'p
A circuit 58 is provided. Then, the adder circuit 57 outputs the average value of the phases of the two signals (more precisely, twice the average value), and the subtractor circuit 58 outputs the phase l (+I-Grf) of the two signals. A value twice the phase difference) is output. Therefore, based on the output values of the force 1 calculating circuit 57 and the subtracting circuit 58, Hl in the height direction and in the bl, i and j directions.
Measuring the k·1 displacement is the iiJ function.

The specific example shown in FIG. 1 is not further shown in FIG. '3. The alignment light enters the mask mark 11 on the mask 10 through a lens 45 and a condenser lens 46, for example.

This light is made to enter the center of the entrance pupil of the lens 20 as a spherical wave.

The mask mark 11 is a one-dimensional diffraction grating as shown in FIG. One is provided.

The ±1st-order light generated from the mask mark 11 passes through the entrance pupil and is irradiated onto the wafer marks 31b and 31c, producing moiré fringes. The wafer marks 31b31c are provided at positions corresponding to χ·l with respect to the mask mark 11, respectively, and include a one-dimensional diffraction grating shown in FIG. 4(a), a two-dimensional diffraction grating shown in FIG. The pattern is a checkered lattice shown in (e) or a combination of these. These selections may be made in an optimal manner depending on the usage conditions of the exposure apparatus. For example, in order to perform positioning using alignment light while transferring a mask pattern, the inventors used marks shown in FIG. 4(e) to two-dimensionally diffract the reflected diffracted light. , the mirror 41 is directed outward from the mask pattern so as not to block the exposure illumination light. Furthermore, if necessary, it is also possible to make the mask mark 1 a two-dimensional grid or a checkered grid.

Further, the alignment light irradiation position on the wafer 30 is within the pattern exposure area, and the detected alignment light is the diffracted light of the wafer marks 31b and 31c, and is always detected under constant conditions regardless of the pattern of the exposed wafer. i signal can be applied.

Furthermore, since the wavelength of the He-Ne laser beam changed by one week is used as the alignment light, the resist on the wafer 30 may be exposed to light during alignment detection (also, there is almost no absorption within the resist, which is sufficient for signal detection). You can get the amount of light.

Note that in the explanation so far, only one direction (for example, the Lateral direction of wafer 30 (X, Y
direction) can be accurately determined. Further, as a means for correcting the positional deviation of the mask and wafer, it is sufficient to minutely move a stage or the like on which the wafer 30 is placed, as is well known.

As described above, according to the method of this embodiment, the wafer 30 relative to the mask 10 (by using a mechanism for AI determining the positional deviation in the seedling direction, and by improving the wafer mark and signal processing circuit, etc.) It is possible to simultaneously determine 4111 displacements in the height direction.

In this case, it is possible to easily solve various problems with conventional height detection, such as a decrease in detection degrees due to changes in pattern reflectance, without providing a special mechanism for height detection. This enables highly accurate height detection.

Therefore, the mask and wafer can be aligned with high precision, which can contribute to improving exposure precision.

Further, since light having a different wavelength from the exposure light is used as the alignment light, it is possible to prevent the resist on the wafer 30 from being exposed to light due to alignment. Furthermore, since the mask mark is provided on the dicing line of the chip, the element formation area does not decrease due to the formation of the alignment mark, and the chip area can be utilized effectively. Further, since the positional deviation of the wafer 30 in the horizontal direction and the height direction relative to the mask 10 can be measured simultaneously, there is an advantage that the time required for aligning the mask and the wafer can be shortened.

Note that the present invention is not limited to the embodiments described above, and can be implemented with various modifications without departing from the gist thereof. For example, as the mask mark or wafer mark, a grid mark as shown in FIG. 4 may be appropriately selected and used. Similarly, as the alignment light, a wavelength different from that of the exposure light and a wavelength that does not expose the resist on the wafer may be appropriately selected.

[Effects of the Invention] As described in detail above, according to the present invention, light having a wavelength of 2 fI different from the exposure wavelength suitable for the pattern stepped optical system is used as the alignment light, and interference fringes of two wavelengths are utilized. By determining the total height and lateral positional deviations between the mask and the wafer, it is possible to determine the height and (seedling direction) of the wafer applied to the mask, regardless of the pattern formed on the exposed wafer. It is possible to align the position with high precision,
This will improve the exposure viscosity and bring it closer to the blade.
Becomes J Noh.

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram (1η diagram) showing a reduction projection exposure apparatus used for one embodiment of the present invention, and Fig. 2 shows the main parts of the above apparatus (
FIG. 3 is a perspective view showing a more specific version of FIG. 1, and FIG. 4 is a block diagram showing an example of a grid mark on a mask or a wafer. 10...-Reticle (mask) 11...t8 f-
Shape mark (mask mark) 2 ()... Projection lens, 3
0 Uj-ha, 3] b, 3] c1,,l] Square mark (wafer mark) 41...Reflection mirror 4
252, ~. 54...Detector, 51...Half mirror 55, 56...Phase difference detection circuit, 57...Addition circuit, 58...Subtraction circuit. Figure 2 Applicant's agent Patent attorney Takehiko Suzue Figure 4

Claims (3)

[Claims] (1) Prior to projecting and exposing a desired pattern formed on a mask onto a wafer through a projection lens, the relative positional deviation of the wafer in the horizontal direction and height direction with respect to the mask is measured, and the measured information is In the mask/wafer alignment method for correcting the position of the mask/wafer based on the above, grating marks are provided on the mask and the wafer, respectively, and have two wavelengths different from those of the exposure light. irradiating the mask mark with alignment light, and irradiating each of the pair of wafer marks with diffracted light from the mask mark generated by the light irradiation through the projection lens;
generating interference fringes of two wavelengths on a pair of wafer marks, detecting each reflected diffraction light from the pair of wafer marks,
A mask characterized in that the amount of positional deviation in the height direction is determined based on a phase difference between the two reflected diffracted lights, and the amount of positional deviation in the lateral direction is determined based on the average value of the phases of the two reflected diffracted lights.・How to align the wafer. (2) Using the alignment light irradiated to the mask mark as a reference light, detect the phase difference T_1, T_2 between this reference light and the two reflected diffraction lights, and calculate the height direction position shift from (T_1-T_2)/2. Find the amount, (T_1+T_2)/
2. The method for aligning a mask and wafer according to claim 1, wherein the amount of lateral positional deviation is determined from . (3) The mask/wafer according to claim 1, wherein a one-dimensional diffraction grating, a two-dimensional diffraction grating, a checkerboard pattern, or a combination thereof is used as the mask mark and the wafer mark. Alignment method.