JPH0346219A - Aligner - Google Patents

Abstract

PURPOSE:To make it possible to conduct a highly precise instrumentation by a method wherein the alignment condition of a wafer and the alignment marks of a mask, an appropriate gain of all the alignment marks on the wafer are computed individually, and an instrumentation signal is regulated properly. CONSTITUTION:A mask 66, having an accurate alignment mark at a plurality of places, and a wafer 68 having the alignment mark same as the mask 66 are integrally carried by use of a carriage 72. The relative positional relation between the mask 16 and the wafer 68 is detected based on the output of a photoelectric conversion device 70, the amount of inclination theta of the optical axis and the scanning axis of the carriage of a projection optical system is computed, a carriage driving mechanism 77 and control valves 79 and 80 are controlled, and the inclination adjustment and magnification adjustment of the optical axis of the projection optical system 67 and the carriage scanning axis are conducted. To be more precise, the level of signal to detect all alignment marks is adjusted for each place. As a result, the irregularity of signal on the point of instrumentation of the wafer can be made uniform, and the accuracy of instrumentation can also be improved.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a semiconductor printing apparatus, and more particularly to a reflective projection exposure apparatus.

[Conventional technology]

In the conventional reflection type projection exposure apparatus, published patent publication 1986
-140826 and Sho 60-140827 are known.

(Problems to be Solved by the Invention) In these conventional techniques, when measuring the positional deviation between the mask and the wafer, a laser scanning type auto-alignment technique is used. As shown in FIG. 7, a plurality of alignment marks are provided on the mask and wafer symmetrically with respect to the center line of the wafer. In order to observe these marks, there are two signal detection systems consisting of 64, 65, 69, and 70 in the direction orthogonal to the plane of the paper in FIG. The amount of deviation due to this is measured simultaneously for one pair. To observe each alignment mark scattered on the wafer,
This is performed by moving the carriage shown in FIG. 172 in the Y direction and driving the signal detection system in the X direction by a drive system (not shown). As shown in FIG. 4, inside the mark detection circuit 84, the analog signal obtained from the photoelectric conversion element 70 in FIG. is being carried out.

At this time, the gain adjustment of the variable gain amplifier 201 is performed only by adjusting the alignment marks A and A' of the mask and wafer shown in FIG. It was fixed to the gain amplifier circuit and used as a gain for signal amplification when measuring other alignment marks.

However, in actual process wafers, due to non-uniform resist coating and variations within the same wafer during the process, 70 photoelectric conversion elements may be measured at other measurement points within the wafer, such as at B and B' in FIG. There may be cases where the voltage level output from the device is different.

In this case, if the amplifier gain is the same as that of the first measurement points A and A', the signal output of the variable gain amplifier is too small to be able to binarize the signal, or it is too large and becomes saturated, making it impossible to perform accurate measurements. was there.

[Means (and effects) for solving the problem] The present invention includes:
In a reflective projection exposure apparatus, when the alignment state of the wafer and mask alignment marks is measured, and the magnification and the tilt of the optical axis of the projection optical system and the carriage scanning axis are calculated from the measured values, and the carriage drive is controlled, Instead of finding the gain at one location (pair of alignment marks) on the wafer surface and applying that gain to all alignment marks on the wafer surface, we now calculate the appropriate gain for each country for all alignment marks on the wafer.
By adjusting the measurement signal, it is possible to perform highly accurate measurements.

(Embodiment) FIG. 1 is a schematic configuration diagram of a reflection projection type exposure apparatus according to an embodiment of the present invention. 6
1. A polygon mirror 62 is arranged.

Here, the laser light source 60. The condenser lens 61 and the polygon mirror 62 are shown rotated by 90' around the Z axis after being arranged in a direction perpendicular to the drawing (X direction). The laser beam is scanned in the X direction by the rotation of the polygon mirror 62.

Along the optical path of the laser beam ℃ reflected by the polygon mirror 62, an f/θ characteristic lens 63, a dividing prism 76 for dividing the laser beam ℃ into two in the X direction, and two half mirrors 64 (one is two objective lenses 65, a mask 66 having accurate alignment marks 101 (described later) at multiple locations, a reflective projection optical system 67, accurate alignment marks 101 at multiple locations; A wafer 08 having an alignment mark 102 (described later) is arranged.

The mask 66 and the wafer 68 are integrally carried by a carriage 72, and the carriage 72 is movable in the Y direction in the figure along a fluid bearing guide 78 by a carriage drive mechanism 77. When the circuit pattern of the mask 66 is exposed onto the wafer 68, the X
Illuminated by a slit-shaped light source (not shown) extending in the direction,
Carriage 72 moves.

The reflected light from the mask 66, the projection optical system 67, and the wafer 68 is reflected by two half mirrors 64, and two condenser lenses 69 and two charge conversion elements 70 are arranged along the optical path of the reflected light. The output of the charging conversion element 70 is applied to an arithmetic processing circuit 71, which detects the relative positional relationship between the mask 66 and the wafer 68 based on the output of the photoelectric conversion element 70, and also calculates magnifications Mx, My according to arithmetic expressions to be described later.
and calculates the amount of inclination θ between the optical axis of the projection optical system 67 and the carriage scanning axis, controls the carriage drive mechanism 77 and control valves 79 and 80, adjusts the inclination of the optical axis of the projection optical system 67 and the carriage scanning axis, and adjusts the magnification. Make adjustments.

The arithmetic processing circuit 71 includes a central processing unit (CPU) 81 and a CPU 8 that control the operation of the arithmetic processing circuit 71.
RO where the program of operation sequence 1 is stored.
M82, RAM 83 for storing arithmetic processing data, mark measuring circuit 84 for measuring the interval between marks on the mask 66 and wafer 68 based on the output of the photoelectric conversion element 70, carriage drive circuit 85 for controlling the carriage drive mechanism 77, arithmetic processing Drive data set circuit 86 that converts and sets data from digital values to analog values, control valve 7
It is equipped with a control valve drive circuit 87 for driving 9.80, a human output interface 88, and the like.

91 is a switch for setting the measurement and adjustment mode, 92 is a switch for inputting the number of repeated measurements, and 90 is an operation panel to which these switches 91 and 92 are attached. Switches 91 and 92 are connected to the input/output interface 88 of the arithmetic processing circuit 71 and change the state of the switches 91 and 92 to C.
It can be input to PU81.

FIG. 2 is a schematic cross-sectional view of the carriage 67 and guide 78 shown in FIG. 1, viewed from the left.

The arm 25 floats on the guide member 78 via an air bearing as a floating element. 31 is an air bearing pad. This pad is connected to a high-pressure air source (not shown) to allow high-pressure air to flow out.

The right arm 25 is provided with six pads on the top, bottom, left and right, and the left arm is provided with two pads on the top and bottom. That is, the air bearing on the right side forms an omnidirectional restraining bearing, and the air bearing on the left side forms a vertical restraining bearing. Therefore, the left air bearing has lower rigidity in both the vertical and horizontal directions than the right air bearing. 50 is a spherical seat provided on the left arm.

FIG. 3(a) shows the alignment marks used on the mask 66 and the wafer 68, and the marks Iota, 101b
, 101c, and 101d.
1 is formed on the mask 66, and the marks 102a, 102b
An alignment mark 102 is formed on the wafer 68. Actually, as shown in FIG. 7, similar alignment marks are formed at multiple locations.

Measurement and adjustment operations in the above configuration will be explained. Referring to FIG. 1, laser light .degree. C. emitted from a laser light source 60 is scanned by a polygon mirror 62 and divided into two parts by a dividing prism 76 in the X direction. The two divided lights pass through a half mirror 64 and head towards a mask 66 and the like.

The alignment mark 101 of the mask 66 and the alignment mark 102 of the wafer 68 are scanned by the laser beam A via the optical system 67, and the scattered light from the marks 101 and 102 reaches the half main channel 64, where some of the light is 2
It is reflected in the direction of two charging conversion elements 70.

When the alignment mark 101 of the mask 66 and the alignment mark 102 of the wafer 68 are overlapped as shown in FIG. 3(a), the alignment mark 102a of the wafer 68 exists between the alignment marks 101a and 101b of the mask 66 and alignment mark 101c
, 101d, and the laser beam is scanned in the scanning direction A (X direction) from left to right. The photoelectric conversion element 70 detects pulsed light at the locations where the laser beam intersects the alignment marks 101 and 102 in FIG. 3(a), and generates an output voltage with a waveform as shown in FIG. 3(b). occurs.

Here, the internal configuration of the mark measuring circuit 84 will be explained a little with reference to FIG.

201 is a variable gain amplifier circuit that amplifies or attenuates the output from the photoelectric conversion element 70; 202 is a signal level measurement circuit that reads the signal output level of the variable gain amplifier;
03 is a slice data output circuit for binarizing the signal output, and 204 is a comparator 206 that compares the voltage of the signal output (analog signal) with the slice data and binarizes it, and a comparator 206 outputs the binarized pulse signal (digital signal). This is a pulse interval measurement circuit that measures the interval between signals.

FIG. 5 is a diagram showing an output obtained by amplifying the output voltage of the photoelectric conversion element 70 in FIG. 3(b) by a variable gain amplifier circuit 201. (a), (b).

Cc ) is a diagram showing the difference in gain of variable gain amplifiers. If the amplifier gain is too high than the appropriate amplifier gain, (C
) becomes saturated.

Therefore, the signal level measuring circuit 202 measures the peak value of the signal and adjusts the variable gain amplifier circuit so that the signal level is appropriate. so-called A
GC function is added.

Next, the explanation returns to FIG. 3. FIG. 3(e) is a diagram showing a measurement pulse signal obtained by binarizing the signal output of the variable gain amplifier. W, , W, . . . W are intervals of pulse signals, and the mark measuring circuit 84 of the arithmetic processing circuit 71 calculates the time intervals W1°W2 . By measuring .

That is, the deviation in the X direction in FIG. 3(a) is ΔX.

If the deviation in the X direction is Δy, then Δx = (W 1- W 2 + W 4- W s
) / '...(1) Δ y −(−Wl +W2 +W 4
-Ws)/4...(2) In the matched state, W, -W, ÷W4=W,
Therefore, Δx1Δy are both (zero).

As shown in FIG. 6, alignment marks 101 and 102 at two locations separated by a distance Ill C+ in the X direction of the mask 66 and wafer 68 are measured, respectively, and the amount of misalignment (
1) and (2), and the amount R, is (Δx Rr
*ΔyRI), Ll (Δx L l , ΔyL+)
Then, the horizontal (X direction) magnification Mx is given by the following equation.

... (3) Also, measure the alignment marks 101 and 102 at two locations separated by a distance of 11 DI in the Y direction, and calculate the amount of misalignment R2 by (
ΔxR2, Δ3/R2), L2 (Δx L 21 Δ
yL2), the vertical (Y direction) magnification My is given by the following equation.

... (4) These calculations are performed by the calculation processing circuit 71.

FIG. 7 shows the relationship between the X'-axis and y-axis of the optical system 67 in the horizontal plane, and the y-axis and X-axis of the carriage 72 in the traveling direction.

In this case, the image plane amount is generated by the angle θ formed between the x and y axes of the optical system 67 and the scanning axes x and y. 10mm in Figure 7
Arrows 81 drawn at intervals indicate the direction and magnitude of transfer distortion. At this time, the one-character pattern 82 as shown in FIG. 8(a) is an image of the optical system 67 that is inverted left and right on the wafer as shown in FIG. 8(b), and in the y-axis direction. The image of the pattern is tilted by an angle of 2θ from the y' axis.

Next, returning to Figure 's6, the measurement of the above angle θ will be described.

Two sets of alignment marks 101 at the bottom of the figure in the X direction.
102 is located on the optical axis of the two objective lenses 65, and when the laser beam .degree. C. is scanned in the direction A in the figure, the two photoelectric conversion elements 70. The matching state is detected by the arithmetic processing circuit 71, and the left-side matching deviation amount ΔxL, .DELTA.yLI and the right-side deviation amount ΔxR+ + ΔyR+ are determined. Subsequently, the carriage drive mechanism 36 is activated by a command from the arithmetic processing circuit 71.
The carriage 72 is moved by a distance i o + in the Y direction via the laser beam, and the alignment state of another alignment mark 101, 102 is measured by scanning the laser beam in the direction B in the figure, and the deviation amount ΔxL2° is similarly measured. ΔyL 21 ΔxR2, Δy
Obtain R2. Here, alignment marks 101 .
The left and right deviation angle θ8 of 102 is approximately expressed by the following equation.

θ, −(1/2G)・(ΔyL+−ΔyRt+Δ
yL2-ΔyR2) (5), and the vertical deviation angle θy of the alignment marks 101 and 102, which are separated by a distance l1lIDl in the Y direction, is approximately expressed by the following equation.

θy − (1/2D)・(ΔxR, +ΔxL4−Δx
R, -ΔXL2)... (6) These calculations are performed by the arithmetic processing circuit 71, and the control valve 79 is adjusted while the carriage is moving so as to reduce this deviation angle, thereby eliminating the transfer distortion error. The amount to be adjusted is θ=(θwooθ,)12 (7), where θ is a component of misalignment and the fact that the optical axis and the scanning axis are not parallel in the horizontal plane.

If a plurality of angles θ to be adjusted are determined, the transfer distortion error can be more accurately resolved.

Then, such measurements are performed at multiple locations on the wafer and mask.

At that time, the flow of the present invention when measuring multiple alignment marks at multiple locations is shown in FIG. 9. First, in FIG. 7, the mask and wafer signal levels of A and A' are adjusted. , determine the gain to obtain an appropriate signal level. After that, the amount of misalignment between the mask and the wafer is measured.

Next, move the carriage and observe B and B' in FIG. Therefore, the signal levels of the mask and wafer are adjusted, the gain is determined to achieve an appropriate signal level, and the amount of deviation is measured. After that, the same operation can be repeated until there are no more alignment marks to measure. After measuring all measurement points, based on the measurement data, perform the above M, , ,
Calculate θ.

By controlling the posture of the carriage based on this calculation result, the transfer distortion between the mask and the wafer is corrected.

That is, in the present invention, alignment marks provided at multiple locations on the mask and wafer (in this embodiment, a pair of alignment marks are provided at each location on the mask and on the wafer) are measured at each location. The signal level of the detected signal is adjusted at each location by AGC.

(Other Examples) As a measurement means, laser scanning alignment technology was used to adjust the analog signal level, but
Even when the TV alignment method is used as the measurement means, the same problem can be solved by adjusting the signal at all measurement points.

(Effects of the Invention) As described above, it is possible to equalize not only uneven coating of resist on a wafer, but also signal variations at measurement points on a wafer in a wafer process.

This prevents unmeasurability with small signals (wafers with low reflectivity).

Further, it is possible to prevent deterioration of measurement accuracy due to a deterioration of the S/N ratio due to a large signal (particularly a wafer with a high reflectance, example A11).

[Brief explanation of drawings]

FIG. 1 is a schematic cross-sectional view of one embodiment of the present invention. 2rXJ is a cross-sectional view of the mask and wafer carriage mechanism of FIG. 1; FIG. 3 is a top view of the alignment marks on the mask and wafer, and the output signal from the charging element. FIG. 4 is an internal block diagram of the mark measuring circuit 84 in FIG. 81. FIG. 5 is an explanatory diagram of the signal level of the alignment signal. FIG. 6 is a top view of the mask and wafer. FIG. 7 is an explanatory diagram of transfer errors between a mask and a wafer. FIG. 8 is an explanatory diagram of transfer distortion between a mask and a wafer. FIG. 9 is an explanatory diagram (flow chart) of gain adjustment according to the present invention. 66... Mask 67... Reflective projection optical system 68... Wafer 70... Photoelectric conversion element 71
... Arithmetic processing circuit 72 ... Carriage mechanism 77 ... Carriage drive mechanism 78 ... Fluid bearing guide

Claims (1)

[Claims] (1) In an exposure apparatus that transfers and exposes a pattern on a mask onto a wafer by moving a carriage that integrally mounts a mask and a wafer having alignment marks at a plurality of locations relative to an imaging optical system, the mask and a scanning means for scanning an alignment mark on a wafer with a laser, a detection means for detecting a signal generated from the alignment mark during laser scanning by the scanning means, and an amplification of the output according to the output of the detection means. An adjustment means for adjusting the level; and an attitude control means for controlling the attitude of the carriage during movement based on the output of the detection means processed at the amplification level adjusted by the adjustment means, the adjustment means comprising: An exposure apparatus characterized in that an amplification level of the output of the detection means is adjusted for each of a plurality of measurement points on the mask and the wafer.