JPH0576765B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of the Invention] The present invention relates to a semiconductor printing apparatus, and more particularly to a semiconductor printing projection exposure apparatus that aligns a reticle and a wafer by detecting their positional relationship with an off-axis optical system. This invention relates to a reference distance correction method.

[Description of Prior Art] Conventionally, as this type of device, a device as shown in FIG. 1 has been proposed.

In FIG. 1, 1 is a moving stage, 2 is a projection lens, 3 is a reticle, 4 is a printing illumination device, and 5
1 is a wafer, 6 and 7 are motors, and 14 is a control box containing a control circuit.

The circuit pattern a on the reticle 3 is reduced onto the wafer 5 by the projection lens 2 and printed as shown in b1 in FIG. 2, for example. Subsequently, the motors 7 and 6 are driven to move the wafer 5 and bake it in steps b2, b3, and so on in FIG. 2. At this time, b, b1, b2, b3, ... are, for example, 1/5 of the circuit pattern a on the reticle 3.
A circuit pattern with reduced dimensions is transferred.

However, such a printing apparatus requires that the wafer and reticle be accurately aligned prior to each printing. Therefore, in the apparatus of the type shown in FIG. 1, laser interferometers 16 and 17 are arranged to detect the coordinates of the moving stage 1 in each moving direction (X, Y direction). 19 is a laser tube that sends laser light to the laser interferometer. As the moving stage 1 moves, pulses are output from the laser interferometer 16 in the X direction and the interferometer 17 in the Y direction, respectively. A control circuit (not shown) in the control box 14 integrates and counts these pulses to determine the amount of movement of the moving stage 1, thereby making it possible to accurately recognize the current coordinates of the moving stage 1. The resolution and linearity of the pulses of the laser interferometers 16 and 17 are sufficient for the required alignment accuracy, for example (0.05μ/1 pulse).
For example, when moving from a certain exposure area bn to the next exposure area bn+1 in FIG. Count accurately and drive the moving stage until the count value indicates the correct amount of movement.

A microscope 18 is provided outside the optical axis of the projection lens 2, and is used to detect errors in the loading position (placing error) when wafers are loaded on the moving stage 1.

In this apparatus, for example, if the position e in FIG. 3 is the standard wafer placement position, and f is the first exposure position for exposing chip b1, then if the wafer is placed correctly, the distance between e and f is By driving the moving stage 1 while measuring the distances x ₀ and y ₀ by the laser interferometers 16 and 17, the wafer 5 can be accurately moved to the position f.

However, it is rare for the wafer 5 to be placed correctly at the position e when it is loaded, and generally, the wafer 5 is placed at the position e', which is shifted by Δx and Δy, and further rotated by Δθ, as shown by the dotted line in FIG. has been done. For this reason, marks such as those shown in FIG. 4c and d are provided on the wafer 5 in order to measure the error during wafer loading using the microscope 18.

In FIG. 5a, it is assumed that the wafer is placed at h with deviations of Δx, Δy, and Δθ from the correct position g. The microscope 18 has a field of view 23 that is large enough to fit, for example, mark d within its field of view even when wafers are stacked with the maximum mechanical error of a stacking device (not shown), and the amount of deviation of mark d is Δx1 and Δy1 are detected by, for example, an imaging means such as a TV camera. Next, the moving stage is moved by a predefined mark interval l ₀ (FIG. 4) and the mark c is observed. FIG. 5b shows the relationship between the mark c and the field of view 23 of the microscope 18 after the stage has been moved. Similarly, the deviation Δx2 between the position where mark c should be visible (dotted line) and the position of mark c (solid line) that is actually observed when the wafer is placed correctly,
Measure Δy2.

Next, the wafer placement errors Δx, Δy, Δθ are calculated from the measured Δx1, Δy1, Δx2, Δy2 and the mark interval _l0 , and when moving the moving stage 1 to the first exposure position f, the wafer placement errors Δx, Δy, Δθ are calculated. The distances x ₀ and y ₀ are corrected by the mounting errors Δx, Δy, and Δθ, and then moved. Note that the θ direction is normally corrected by a rotation mechanism of a wafer loading chuck (not shown) of the moving stage 1.

In an apparatus having such a configuration, the attachment of the microscope 18 is required to be extremely precise and stable over a long period of time. For example, if the required alignment accuracy for each exposure is 0.1μ, the microscope 18
It had to be installed with the following accuracy, which was a major hindrance to the assembly and maintenance of the device. For example, on this order (less than 0.1μ)
In addition to precision during assembly and installation, temperature must also be strictly controlled. For example, once the power to the device is turned off, the temperature of the entire device will drop, and the next time the power is turned on, the temperature of the entire device will drop. There were some problems where the device could not be used for a long time until the temperature stabilized.

[Object of the Invention] The present invention provides a projection exposure apparatus that uses an off-axis microscope to align a wafer and a reticle, which greatly eases the requirement for precision in the mounting position of the off-axis microscope and facilitates assembly and maintenance of the apparatus. The purpose is to reduce the amount of time the device is unavailable due to turning the power on and off (downtime).

[Description of Embodiments] FIG. 6 shows an example of the configuration of an apparatus for carrying out the present invention, in which the apparatus of the method shown in FIG. This is an additional system that uses beam scanning to detect alignment marks on the wafer and reticle.

The on-axis alignment mark detection system shown in FIG. 6 oscillates a laser beam using a laser tube 13, and directs this beam into the projection lens 2 via a polygon mirror 12, a half mirror 10, an objective lens 9, and a mirror 8. This scans alignment marks (not shown) provided on the reticle 3 and wafer 5. Scattered light from the alignment marks on the reticle 3 and wafer 5 passes through a mirror 8, an objective lens 9, and a half mirror 10, and is detected by a photoelectric detector 11. A control circuit (not shown) in the control box 14 uses the output signal of the photoelectric detector 11 to determine the position of the alignment mark between the wafer and the reticle, that is, the normal position of the wafer when the moving stage 1 moves to the first exposure position. (f in Figure 3)
It is possible to calculate the amount of deviation from . Note that, hereinafter, this observation system will be referred to as TTL.

In the above configuration, after turning on the power of the apparatus and before loading the first wafer of the lot, test wafers as shown in FIG. 7 are first supplied to the apparatus. At this time, a control circuit (not shown) in the control box 14
The operation mode (control program) is switched to the adjustment mode (or microscope position measurement mode) by input from the input/output terminal 21.

In this adjustment mode, the control circuit first
Marks c and d on the wafer are each marked with a microscope 18.
The moving stage 1 is driven to come within the field of view 23, and the deviation amounts Δx1, Δy1, and Δx2 of each mark are determined.
Measure Δy2. Then, the wafer placement errors Δx, Δy, and Δθ are determined in exactly the same manner as in the case of the conventional apparatus, and the movable stage 1 is moved to the first exposure position f while correcting the wafer placement errors Δx, Δy, and Δθ.

The test wafer 25 is a stable one with good flatness, such as a glass wafer, and as shown in FIG. 7a, the first exposure area 26 is
A mark i for the TTL observation system, as shown in Fig. 7, is provided near the . The reticle shown at 27 in FIG. 7b also has a mark j as shown in FIG. 7d for the TTL observation system at a location corresponding to mark i on the wafer near area k for the circuit pattern. is provided.

Now, the control circuit moves the moving stage 1 to the first position.
After driving to the exposure position, the TTL observation system is driven to scan the mark j on the reticle 27 and the mark i on the wafer 25 with a laser beam.
Marks i and j are observed overlappingly as shown in Figure 7e in the TTL observation system, and the laser beam is 28
These marks are scanned as shown in , and the intervals Δl1 to Δl4 between each mark are measured. Microscope 18
If the marks are in the correct position, the marks are correctly aligned (Δl1 = Δl2 = Δl3 = Δl4), but the microscope 18 may move away from the normal position due to temperature, aging, installation error, etc., for example Δx', Δy' If it is misaligned,
Interval Δl1 between each mark in the TTL observation system
~Δl4 also deviates by the amount multiplied by the magnification of each observation system. The control circuit stores this error in a storage device (not shown) and ends the adjustment mode.

Next, the wafers from the lot are loaded, and after the adjustment mode is finished, the marks c and d on the wafers are observed with the microscope 18, and the wafer placement errors Δx, Δy,
Δθ is calculated, and when moving the moving stage 1 to the first exposure position f while correcting the mounting errors Δx, Δy, and Δθ, the deviation amounts Δx′ and Δy′ stored in the memory circuit are taken into consideration. to drive the moving stage 1.

[Modified Example of Embodiment] In the above embodiment, a method of scanning with a laser beam was used as a TTL observation system for measuring the position of the microscope 18, but other means such as a TV camera were used to perform alignment from the video signal. A method that reads marks may also be used.

In addition, the reticle of the above-mentioned embodiment has a masking mechanism of the reticle during normal printing.
If you mask this area, there is no need to prepare a special reticle for testing, and you can use the regular printing reticle with marks J, and in this case, there is no need to replace the reticle. .

Further, in the above embodiment, the TTL observation system is one-lens, but if it is made two-lens and the marks on the reticle and wafer are two-lens, errors in the rotational direction can also be determined with high accuracy.

In addition, in the above, a test wafer is used when driving the TTL observation system, but a mark for TTL observation is provided near the first exposure area of each wafer or the first wafer of a lot. Alternatively, the positional shift of the microscope 18 may be measured.

Also, if the position of the TTL observation system is stable,
Marks on the reticle are not required. That is, in this case, the positional deviation of the microscope 18 can be measured by observing only the mark position on the wafer with respect to the TTL observation system. Generally, this type of equipment transfers the reticle pattern onto the wafer (for example, 1/1
5) Since reduction printing is performed, the tolerance on the reticle side may be larger than that on the wafer surface side.

[Effects of the Invention] As explained above, according to the present invention, positional errors due to off-axis microscope temperature, aging, mounting, etc. are measured by a TTL observation system provided on the wafer, and the subsequent By taking measures such as correcting the feed amount to one exposure position or the first exposure position, the positional accuracy of the off-axis microscope can be relaxed, and the assembly, adjustment and maintenance of the apparatus can be facilitated, and furthermore, Downtime due to power on/off can be reduced.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing an example of a conventional projection exposure apparatus that performs alignment using an off-axis observation system, Fig. 2 is a diagram showing how to print onto a wafer in the apparatus of Fig. A diagram for explaining off-axis observation and wafer feeding methods in the device shown in FIG. 1, FIG. 4 is a diagram for explaining detection of the wafer position detection mark, FIG. 6 is a schematic configuration diagram showing a projection exposure apparatus according to an embodiment of the present invention, and FIG. 7 is a diagram used in the apparatus shown in FIG. 6. FIG. 2 is a schematic diagram of a pattern formed on a wafer and reticle. In the figure, 1 is a moving stage, 2 is a projection lens, and 3
is a reticle, 4 is a printing illumination device, 5 is a wafer,
6 and 7 are motors, 8 is a mirror, 9 is an objective lens,
10 and 15 are half mirrors, 11 is a photoelectric detector,
12 polygon mirror, 13 laser tube, 1
4 is a control box, 16 and 17 are laser interferometers, 18 is a microscope, and 21 is an input/output terminal.

Claims (1)

[Claims] 1 The step of measuring the positional error using an off-axis microscope installed off the optical axis of the projection optical system that projects the pattern of the original onto the substrate, and the step of measuring the positional error based on the measured positional error and a pre-stored reference distance. moving the stage on which the substrate is mounted toward the pattern projection position of the projection optical system; and after moving the stage toward the pattern projection position, determining the distance between the mark on the original plate side and the mark on the substrate side. A method for correcting a reference distance for a projection exposure apparatus, comprising the steps of: measuring a positional error between the marks via the projection optical system; and correcting the reference distance by the measured positional error between the marks. .