JPH0617770B2 - Mark detector - Google Patents

Description

Description: (1) Description of the field to which the invention pertains The present invention relates to an alignment mark detector for a wafer mask or the like in a semiconductor manufacturing apparatus.

(2) Description of Prior Art FIG. 3 is a schematic diagram showing an outline of a mark detection device. 1 is a light source, a He-Ne laser, 2 is an optical scanner using a galvano mirror, 3 is a beam splitter, 4 is a position sensor, 5 is an optical detector, 6 is a wafer, 7 is a diffraction grating mark processed on the wafer. is there.

In such a structure, the optical scanner 2 is moved to the arrow b.
When rotated in the direction of, the light beam moves on the wafer 6 in the direction of arrow a in the figure. When the light beam travels and hits the mark 7, diffracted light (L) as shown in the figure is generated, hits the photodetector 5, and the mark signal is output from the photodetector 5. Since there is a fixed relationship between the control input voltage of the optical scanner 2 and the rotation angle, conventionally, mark position information is obtained from the control input voltage of the optical scanner 2 at the time when this mark signal is obtained. A reference control input voltage is periodically applied to the optical scanner 2, and at that time, the position sensor 4 reads the position of the light beam branched by the beam splitter 3 to correct the position detection error due to the drift of the optical scanner 2. .
When such a drift correction method is used, the position sensor 4 must detect the absolute position with high accuracy over the entire range in which the light beam drifts. For that purpose, the position sensor 4 uses a light position voltage conversion circuit. A complicated circuit as shown in FIG. 4 must be used. In FIG. 4, 4 is a two-dimensional position sensor, 11 and 12 are current-voltage conversion circuits, 13 is a subtractor, 14 is an adder, and 15 is a divider. Here, the high-precision subtractor and divider have the drawbacks that matching resistors and a large number of trimmers are required, the elements are expensive, and adjustment takes time.

Even in the position sensor circuit having such a configuration, when the ambient temperature changes and the dark current of the position sensor 4 increases, only the denominator input (D) of the divider 15 changes, so that the gain of the divider is reasonably priced. Changes, and a position detection error occurs, resulting in defective correction.

Further, there is an inconvenience that the drift amount of the optical scanner, which is periodically measured by the position sensor, is stored in the memory and must be read out every time the mark is detected to perform the correction calculation.

(3) Object of the Invention It is an object of the present invention to solve these drawbacks and to obtain a mark detector with high detection accuracy by a simple circuit configuration.

(4) Description of Configuration and Operation of the Invention A workpiece is a wafer and a mark is a diffraction grating mark. In the semiconductor manufacturing equipment, the wafer must be placed on the sample stage so that the mark originally processed on the wafer comes to a fixed position (hereinafter referred to as the mark reference position) in the coordinate system of the equipment. When the wafer is installed due to dimensional error or shaking during transportation, the mark is located slightly off the mark reference position. Therefore, the mark detector detects the coordinate position where the mark actually exists with the mark reference position as the origin.

Before describing the use of the position sensor as a position detection zero-cross comparator, the position sensor used in the present invention will be briefly described. The element used here is a non-divided semiconductor position detector, for example, S1544 or S1545 manufactured by Hamamatsu Photonics or PI manufactured by UDT.
For example, N-LSC / 5D. In these detectors, when the photocurrent generated at the point where the light strikes flows into the two electrodes, the photocurrent splits in inverse proportion to the distance from the position where the light hits to both electrodes, so the difference between the currents flowing out from the two currents. The position where the light hits can be known from this difference value. However, when the amount of light striking the detector changes, the difference in photocurrent with respect to the position also changes. Therefore, as shown in FIG. 4, the adder 14 calculates the sum of photocurrents, and the divider 15 calculates the difference in photocurrents as the sum of photocurrents. Unless the circuit is divided by, the position where the light hits cannot be detected regardless of the light intensity.

Further, even in such a circuit configuration, when the dark current changes, an error occurs as described above.

In the present invention, attention is paid to the fact that when light is applied to the centers of both electrodes of such a position sensor (hereinafter referred to as point O), an equal current flows through the two electrodes regardless of the intensity of the light, and is shown in FIG. A position sensor is used as a zero-cross comparator that outputs only the polarity of the difference between the currents flowing out from the two electrodes using such a circuit. In FIG. 5, 4, 11 and 12 are the same as in FIG. 4, and 17 is a voltage comparator. Also, FIGS. 6a and 6b show the relationship between the position where the light beam strikes and the output voltages 16 and 18 in the position sensor circuits shown in FIGS. 4 and 5, respectively. As can be seen from FIG. 6 (a), in the circuit of FIG. 4, the position where the light beam hits and the output voltage 16 have a linear relationship, and the position sensor which can read the position where the light beam hits directly from the output voltage value. When the dark current of No. 4 increases, the input / output relationship changes as shown by the dotted line, and a detection position error occurs. On the other hand, when the circuit of FIG. 5 is used (in the case of FIG. 6 (b)), the position where the light beam hits cannot be known from the output voltage, but the moment the light beam moves over the position sensor and passes through point O. Can be known from the change in output voltage, and the position of this O point can always be detected accurately without being affected by the amount of incident light or the amount of dark current. Therefore, if the position sensor is attached so that the light beam hits the point O of the position sensor when the light beam is at the mark reference position, when the light beam is scanned, the light beam will be the mark reference position no matter how the optical scanner drifts. The moment of passing the position can always be detected accurately by detecting the O point of the position sensor.

According to the present invention, the light beam is scanned at a constant speed using the zero-cross comparator to measure the time difference from the point O detection time to the mark signal detection time of the zero-cross comparator to know the distance from the mark reference position to the actual mark position. An embodiment of the mark detector is shown in FIG. Blocks 4, 11, and 12 in FIG. 1 are the same as those in FIG. 5, and 5 is the photodetector shown in FIG. Reference numeral 18 is a current-voltage conversion circuit similar to 11 and 12, 19 is a peak comparator, 20 is a time difference signal generation circuit, 21 is a reference clock generation circuit, 22 is an AND gate, and 23 is a counter.

FIG. 2 is a waveform diagram showing signal waveforms at various points in the block diagram shown in FIG. In Figure 2, (A) ~ (F)
The signal waveforms of the above are shown with the horizontal axis representing the beam position (-rotational axis of the optical scanner).

As shown in FIG. 6 (b), a signal as shown in FIG. 2 (A) appears at the output of the voltage comparator 17 when a beam is scanned by giving a control signal to the optical scanner. A signal as shown in FIG. 2B is obtained from the output of 18 according to the position of the mark, and the length indicated by Err in the figure is the error amount from the mark reference position to the actual mark position. Mark signal (B) is peak comparator
After entering 19 and transformed into a mark position signal (FIG. 2 (C)), it is input to the time difference signal generating circuit 20 together with the reference position signal (A).

The time difference signal generation circuit 20 outputs a pulse (D) having a time difference between the two input signals (A) and (C), and at the same time outputs a flag (24) indicating which signal is input first. This corresponds to the polarity of the mark position error.

The reference clock output from the reference clock generation circuit 21 is for converting the time difference signal (D) into distance information AND
The gate 22 becomes a pulse train (F) that is output only during the time difference signal (D) and enters the counter 23 where it is counted as a position error.

When this circuit (FIG. 1) is used, the absolute value of the error amount from the reference position to the actual mark position is displayed on the counter 23 at the end of one scan, and the polarity of the error is the time difference signal generation circuit 20.
Flag 24 is set.

Although FIG. 3 shows an example of one-dimensional mark detection, a two-dimensional mark detector can be realized by adding another pair of optical scanners and replacing the position sensor with a two-dimensional mark. In this case, the linearity of the position sensor itself is worse than in the one-dimensional case, and the peripheral circuits are required twice, so that the merit of implementing the present invention is further expanded.

Further, in the present invention, the optical scanner has to be driven at a constant speed, but in order to ensure the accuracy of the optical scanner, it is common to drive at a constant speed even in the conventional technique, and this does not cause any disadvantage.

(5) Description of Effects As described above, according to the present invention, it is possible to remove all the non-linearity error of the position sensor, the error due to the dark current change, and the error of the divider, which are the main causes of the error in the prior art. Not only can the position detection accuracy be greatly improved, but the circuit can be simplified.

Further, since the drift of the optical scanner does not affect the position of the mark detected by the present invention, there is no need to make a correction calculation after measurement, which is also advantageous in many respects.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a signal waveform diagram at various points in the block diagram shown in FIG. 1, and FIG. 3 is a schematic diagram of a mark position detecting device according to the present invention. Fourth
FIG. 5 is a conventional mark position detection circuit, FIG. 5 is a zero-cross comparator composed of a position sensor and a comparator, and FIG. 6 is a characteristic diagram showing the relationship between the beam position and the output voltage. (Explanation of symbols of main parts) 1 …… Laser 2 …… Optical scanner 4 …… Position sensor 5 …… Optical detector 7 …… Diffraction grating mark 11,12 …… Current-voltage conversion circuit 17 …… Comparator, 21 …… Reference clock generator 23 …… Counter

Claims (1)

[Claims] 1. A light beam sweeping means for sweeping a marked work piece by a light beam; a beam splitter for branching the light beam from the light beam sweeping means before entering the work piece; and the beam. Receives the light beam split by the splitter,
A position sensor that outputs a signal of a level corresponding to the light receiving position, a comparator that binarizes the output of the position sensor at a predetermined level, a detector that receives the light beam through the mark, and the detection Center detection means for detecting the center of the mark from the output signal of the detector, time difference measuring means for measuring the time difference between the time when the output of the comparator changes and the time when the center of the mark detection means detects the center of the mark. A mark detector comprising: