JPH06100711B2 - Automatic focusing device - Google Patents

Description

The present invention relates to an automatic focusing device.

[Conventional technology]

With the recent development of the semiconductor industry, there is an increasing demand for an inspection device equipped with a highly accurate focusing device for inspecting a wafer or the like. For example, JP-A-58-217909
In the light projection type automatic focusing device described in Japanese Patent Publication No. JP-A-2004-115, as shown in FIG. 6, the sample is irradiated with infrared laser light through a half cross section on one side of the pupil of the objective lens 2, and the laser light spot is spotted on the sample surface 1. The image 3 is formed, and the reflected laser light from the sample surface 1 is passed through the half section of the objective lens 2 on the opposite side of the pupil to the differential diode.
Detection is performed by the light receiving element consisting of D ₁ and D ₂ . The spot image 3 formed on the light receiving element surface moves on the light receiving element surface as shown by the arrow in FIG. 7 according to the displacement of the sample surface 1 in the optical axis direction. It is detected from the difference in the amount of light received by ₁ and D ₂ . In addition, M is a measuring light beam path, Mb is a measuring light beam bundle, Mr is a returned measuring light beam bundle, and 4 is a slit diaphragm. By the way, the size of the spot image 3 formed on the sample surface 1 usually has a diameter of several μm to several tens μm, although it varies depending on the magnification of the objective lens 2. When the spot image 3 having such a finite size has just come to the end of the minute unevenness of the sample, the step portion (FIG. 8 (c)), the inclined portion (FIG. 9 (c)), etc. As shown in FIGS. 9B and 9B, the spot image 3 is formed on the light receiving element in a state of being partially deformed by the shape of the sample surface. The light amount distributions on the light receiving element at this time are as shown in FIGS. 8 (a) and 9 (a), respectively. Incidentally, the spatial correspondence between the diagrams (c) and (b) is actually the opposite because the optical system is inserted between the sample surface 1 and the differential diodes D ₁ and D ₂ . For easy understanding, in FIGS. 8 and 9, the apparent spot image (FIG. (B)) and the sample surface (FIG. (C)) are drawn in correspondence with each other. In the figure (b), the width of the gap 5 between the differential diodes D ₁ and D ₂ is set to be substantially the same as the diameter of the spot image 3 in the focused state as shown by the alternate long and short dash line in the figure ( When Figure 7), the light of the differential diode D ₂ side in (a) view is not detected in the diode D _2, a differential diode
Only the light due to the blurred image on the D ₁ side is detected by the diode D ₁ . Although this state is actually focused on the high side of the sample surface 1, as long as the signals of the differential diodes D ₁ and D ₂ are detected, the sample surface 1 is a focused plane. Since it is recognized that the lower surface side of the sample surface 1 is moved to the focusing plane, the control system (not shown) is recognized as being positioned far below. However (as you can see clearly in the case of FIG. 8 in particular), when the lower surface side is in focus, the relative positional relationship between the differential diodes D ₁ and D ₂ and the blurred image is simply changed from the above state. Since the other conditions are the same, the focus is again on the high side of the sample surface 1. In this way, the servo system is oscillated at the boundary of the stepped portion of the sample surface 1 and the focusing operation becomes impossible. In the case of a flat mirror-like sample such as a raw wafer, such a step-like portion on the sample surface 1 exists only at the end portion of the outer peripheral portion, but in a wafer on which a pattern is formed or a general metal sample, etc. Such stepped portions are intermittently observed, which is a very problematic problem in practical use. In order to prevent this, it is possible to temporarily stop the focusing operation by detecting a mere decrease in the amount of reflected light at the step, but there is a drawback that the retractable range of the focusing operation becomes narrow, and as a result, the target The application range of the sample becomes narrow. Even if this method is adopted, the amount of reflected light may not be so much reduced in the case of a step portion having a high reflectance,
As a result, the step is not detected and the operation becomes inaccurate, which is not a fundamental solution.

Further, as a light receiving element, Japanese Patent Laid-Open Nos. 58-7112 and 60
Even if a semiconductor position detection element (PSD) such as that described in −42725 publication is adopted, oscillation of the servo system may be prevented, but a decrease in focusing accuracy at the stepped portion is unavoidable, and This is a serious problem for a sample having continuous steps.

[Problems to be solved by the invention]

In view of the above problems, the present invention provides an automatic focusing device capable of performing highly accurate focusing even on a sample having a large number of step portions and ensuring a wide range in which a focusing operation can be drawn. The purpose is to

[Means and Actions for Solving Problems]

In an automatic focusing device that projects a measurement light beam on the sample surface and detects the reflected light to automatically perform the focusing operation, the measurement light beam is imaged on the high and low surfaces of the step portion of the sample surface. The stepped portion is detected from the difference in the states, and the automatic focusing operation at the stepped portion is temporarily stopped to eliminate the adverse effect of the stepped portion during the automatic focusing operation. .

〔Example〕

Hereinafter, the present invention will be described in detail with reference to the illustrated embodiment. FIG. 1 shows an optical system according to the embodiment, and the basic principle of focusing is described in JP-A-60-42725. It is the same as the focus detection device. 11 is an objective lens, 12 is a lens barrel lens, 13 is a dichroic mirror, 14 is a reduction lens, 15 is an infrared light spot projection lens, 16 is a beam splitter,
Reference numeral 17 is a detection lens, 18 is a PSD (light receiving element), 19 is a prism, 20 and 21 are microlenses, and 22 and 23 are infrared light emitting elements such as laser diodes. These are focus detection optical systems. I am configuring. The infrared light emitted alternately from the light emitting elements 22 and 23 passes through the minute lenses 20 and 21, is reflected by the prism 19, passes through the beam splitter 16, and is projected onto the image plane A by the projection lens 15. Further, it is projected onto the sample B through the reduction lens 14, the dichroic mirror 13, the lens barrel lens 12, and the objective lens 11. The infrared light reflected by the sample B passes through the objective lens 11, the lens barrel lens 12, the dichroic mirror 13, the reduction lens 14 and the projection lens 15 again, and the beam splitter 16
Is reflected by the detection lens 17 to form a spot on the PSD 18. Then, when out of focus, the light emitting element 22 on the PSD 18
Since the spot images of 23 and 23 are formed at different positions, the spot images of the light-emitting elements 22 and 23 are alternately moved by blinking, and when focused, the spot images of the light-receiving elements 22 and 23 are both at the same position. The spot image does not move even when both of them blink, and the focusing operation is performed by moving the sample surface B so that the spot image does not move as a result. Further, it is possible to determine the front focus and the rear focus and measure the defocus amount from the relative positional relationship and the movement amount of the spot image accompanying the blinking.

24 is a prism having a half mirror surface, 25 is an infrared light cut filter, 26 is an eyepiece lens, 27 is a film surface, and these are observed together with the objective lens 11 and the lens barrel lens 12.
It constitutes the photographic optical system. 28 is a half mirror, 29 is a condenser lens, and 30 is an illumination light source, and these constitute an illumination optical system together with the objective lens 11 and the lens barrel lens 12.

Reference numeral 31 is a stage having a rack 32 and movable in the optical axis direction (vertical direction), and 33 is a motor for driving a stage 31 having a pinion 34 meshing with the rack 32 fixed to a rotating shaft.

In the automatic focusing optical system, the objective lens 11 or the lens barrel lens
Since the reduction lens 14 is arranged between the image plane A formed by 12 and the objective lens 11 or the lens barrel lens 12, even if the objective lens 11 is replaced with another objective lens which generally has different chromatic aberration. The amount of deviation from the image plane A of the infrared image can be made extremely small. Therefore, the residual deviation of the infrared light image can be sufficiently corrected by the correction circuit included in the focus detection circuit as described in Japanese Patent Application No. 60-271133, and the objective lens 11 can be completely focused even when it is replaced. It is possible to maintain the state.

FIG. 2 shows the automatic focusing circuit of the above-mentioned embodiment.
6 is a current-voltage conversion circuit, 37 is an adder, 38 is a divider, 39
Is a sample hold circuit, 40 is a bandpass filter, 41
Is a synchronous rectification circuit, 42 is a low-pass filter, 43 is a servo amplifier, 44 is a power amplifier, 45 is a rotation detection system linked to the motor 33 such as a tachogenerator and a rotary encoder, 46 is a signal conversion circuit of the rotation detection system 45, 47 is a switch circuit, 48 and 49 are sample and hold circuits, 50 is an adder circuit, 51 and
55 is a level discriminator, 52 is a logic circuit, 53 and 56 are reference voltage setting circuits, and 54 is an objective lens magnification detection circuit.

The circuit portion composed of the current-voltage conversion circuits 35 and 36, the adder 37, and the divider 38 is the position of the infrared spot image formed on the PSD 18 when the light emitting elements 22 and 23 emit light alternately. The signal processing circuit is the same as the signal processing circuit disclosed in Japanese Patent Laid-Open No. 60-42725. The output signal of which the output of this circuit portion is sampled and held by the sample and hold circuit 39 in accordance with the light emission timings of the light emitting elements 22 and 23 is as shown in FIG. The amplitude at this time corresponds to the deviation from the focus. The signal obtained by processing the output signal by the bandpass filter 40 and the synchronous rectification circuit 41 is the third signal, respectively.
As shown in FIGS. 3B and 3C, and by applying this to the low-pass filter 42, a DC signal as shown in FIG. 3D is obtained. The direct current component indicated by FE in FIG. 3 (d) corresponds to the defocus amount, FE = 0 at the time of focusing, matching the GND level, and when defocusing occurs in the opposite direction to the above, Three
A negative DC signal is generated as indicated by FE 'in FIG. Servo amplifier 43, power amplifier 4 based on this signal
The motor 33 is driven by 4 and the stage 31 is driven to focus the defocus amount FE to 0. At this time, the rotation state of the motor 33 is changed to the rotation detection system.
It is common to detect the signal at 45, convert the signal at the signal conversion circuit 46, and feed it back to the servo amplifier 43 to give the servo system a damping effect to improve the focusing accuracy and stability. is there.

Thus, the focusing operation is performed, and the light emitting elements 22 and 23 on the PSD 18 are
As shown in FIG. 4 (a), the infrared spot image due to is converged to one point, and the output of the sample hold circuit 39 at that time is the fifth point.
It matches the GND level as shown in FIG. The above is
This is a case where the light emitting elements 22 and 23, the light receiving element 18, the image plane A, and the sample plane B are arranged in a completely conjugate positional relationship. The positional relationship is arranged by, for example, slightly moving the projection lens 15. If they are displaced by a very small amount, such as by installing them, the infrared spot images of the light emitting elements 22 and 23 will be imaged at different positions in the focused state, as shown in FIG. 4 (b). , The fifth as the output of the sample hold circuit 39 at this time
A signal output as shown in FIG. In FIG. 4, the infrared spot images that are alternately formed by blinking are drawn together by changing the direction of the diagonal lines, but in reality, only one of them forms an image at a certain moment. It goes without saying that there is. Here, when the sample surface B is a normal plane, the defocus amount for each infrared spot is shown as shown in FIG. 5 (b) corresponding to the symmetry of the infrared spot image on the PSD 18. Δ
F1 and ΔF2 have the same size. However, when there is a step on the sample surface B and the spot hits it, the infrared spot image formed on the PSD 18 is the same as in the case of the prior art.
As shown in FIG. 7C, one, that is, the one that coincides with the in-focus plane is a point image, and the other, that is, the one that deviates from the in-focus plane is a blurred image. Correspondingly, the output signal of the sample hold circuit 39 is also the fifth image.
As shown in Figure (c), it is asymmetrical with respect to GND level.
Appears as an output signal with ΔF3 | ≠ | ΔF4 |. When the signals in FIG. 7C are sampled and held by the sample and hold circuits 48 and 49 at the lighting timings of the light emitting elements 22 and 23, DC components ΔF3 and ΔF4 are obtained as shown in FIG. 5D.
Are converted into DC level signals each having a. When these are added by the adder circuit 50, since ΔF3 and ΔF4 usually have different signs, the DC level signal | ΔF3 | − | ΔF4 | corresponding to the difference between the two signals is actually output. This is shown in FIG. 5 (e). This output is compared with the upper limit threshold (UTP) and the lower limit threshold (LTP) set by the reference voltage setting circuit 53 in the level discriminator 51, and the output indicating whether or not it is within the range of both is output to the logic circuit 52. Output. The logic circuit 52 is a switch circuit according to the result.
47 is turned on and off, and the focusing section drive motor 33 is operated to stop as required. The reference voltage setting circuit 53 is configured to generate a reference voltage corresponding to the objective lens magnification in accordance with the output of the magnification detection circuit 54, for example. or,
The output of the adder 37 is a signal corresponding to the total amount of received light of the PSD 18, that is, the reflectance of the sample surface B. This signal is input to the level discriminator 55 whose threshold is set by the reference voltage setting circuit 56, and the signal level is Is determined to be equal to or greater than a threshold value. According to the result, the logic circuit 52 turns on / off the motor 33 via the switch circuit 47. Similarly to the reference voltage setting circuit 53, the reference voltage setting circuit 56 also receives the output of the objective lens magnification detection circuit 54 to generate a reference voltage according to the objective lens magnification.

Since the automatic focusing device according to the present invention is constructed as described above, when the infrared spot hits the stepped portion of the sample surface B, the output | ΔF3 | The DC level signal of − | ΔF4 | becomes a value other than zero as shown in FIG. 5 (e), the step difference becomes large and exceeds the reference value determined by the reference voltage setting circuit 53, that is, | ΔF3 | When the level of − | ΔF4 | exceeds the range between UTPLTP, the motor 33 is stopped by the logic circuit 52 and the automatic focusing operation is temporarily stopped. When the position other than the infrared spot step portion is reached, the level of | ΔF3 | − | ΔF4 | returns to almost zero, so the motor 33 is driven again and the automatic focusing operation is restarted.

At the same time, in the level discriminator 55, it is judged from the output of the adder 37 whether the reflectance of the sample surface B is the reference value or more. When the reflectance is low, the logic circuit 52 stops the motor 33 and the automatic focusing operation is performed. To pause. In the apparatus of the present invention, the two infrared spots are deviated by a minute amount and
Since it is located above, it is not drawn clearly in the figure,
The circuit is configured to detect the reflectance of each spot and enable the automatic focusing operation only when both are above the reference value. It is necessary to suspend the automatic focusing operation when dust with a low rate is encountered.

The apparatus of the present invention operates as described above, and even if the infrared spot hits the stepped portion on the sample surface B, the focus position immediately before is maintained only by temporarily stopping the automatic focusing operation, and the infrared spot is the stepped portion. Since the automatic focusing operation is resumed again when it is out of the range, high-precision focusing is possible even when observing a sample with a large number of stepped portions without worrying about the oscillation and malfunction of the servo system as in the past. . Also, since the stepped portion is not detected due to the decrease in the amount of reflected light, it is not necessary to narrow the retractable range of automatic focusing to cope with the step with high reflectance, and secure a wide retractable range. There is a big advantage of being able to do it.

Although the number of projected light beams is two in the above embodiment, the basic configuration does not change even if the number of light beams is three or more.

Further, although the triangular prism is used as the means for synthesizing a plurality of measuring light beams, other means such as a triangular mirror can be used.

〔The invention's effect〕

As described above, the automatic focusing device according to the present invention is capable of performing highly accurate focusing even on a sample having a large number of step portions, and can secure a wide range in which the focusing operation can be pulled in. It has an important practical advantage of being able to target a sample.

[Brief description of drawings]

FIG. 1 is a diagram showing an optical system of an embodiment of an automatic focusing device according to the present invention, FIG. 2 is a diagram showing an automatic focusing circuit of the above embodiment, and FIG. 3 is a sample hold of the automatic focusing circuit. circuit
FIG. 4 is a diagram showing changes in the signal waveform from 39 to the low pass filter 43, FIG. 4 is a diagram showing an infrared spot image on the PSD 18 of the above embodiment, and FIG. 5 is a diagram showing the low pass filter 43 to the adder circuit of the automatic focusing circuit. Diagram showing changes in signal waveform up to 50,
FIG. 6 is a diagram showing an optical system of a conventional example, FIG. 7 is a diagram showing a spot image on a light receiving element of the conventional example, and FIGS. 8 and 9 are stepped portions of a sample surface in the conventional example, respectively. It is a figure which shows the light amount distribution on a light-receiving element at the time of respond | corresponding to an inclined part, the imaging state of a spot image, and a sample cross section. 11 …… Objective lens, 12 …… Body lens, 13 …… Dichroic mirror, 14 …… Reduction lens, 15 …… Infrared light spot projection lens, 16 …… Beam splitter, 17 …… Detection lens, 18 …… PSD, 19 …… Prism, 20,21 …… Reduction lens, 22,23 …… Infrared light emitting element, 24 …… Prism, 2
5 …… Infrared ray cut filter, 26 …… Eyepiece, 27
…… Film surface, 28 …… Half mirror, 29 …… Collimator lens, 30 …… Illumination light source, 31 …… Stage, 32 ……
Rack, 33 ... Motor, 34 ... Pinion, 35, 36 ... Current-voltage conversion circuit, 37 ... Adder, 38 ... Divider, 39 ...
… Sample hold circuit, 40 …… Band pass filter, 41 …… Synchronous rectification circuit, 42 …… Low pass filter,
43 …… Servo amplifier, 44 …… Power amplifier, 45 …… Rotation detection system, 46 …… Signal conversion circuit, 47 …… Switch circuit, 4
8,49 …… Sample and hold circuit, 50 …… Adding circuit, 51,5
5: Level discriminator, 52: Logic circuit, 53, 56: Reference voltage setting circuit, 54: Objective lens magnification detection circuit.

Claims (1)

[Claims] 1. An automatic focusing device for projecting a measuring light beam on a sample surface and detecting the reflected light beam to automatically perform a focusing operation, wherein a plurality of measuring light beams are provided and a step portion of a sample surface is provided. An automatic focusing device, characterized in that it is detected as a step portion from the difference in the image formation state of each of the measuring light beams on the high surface and the bottom surface, and the automatic focusing operation at the step portion is temporarily stopped. .