JP5231927B2 - Microprojection inspection device - Google Patents

Description

  The present invention measures the height of micro-projections such as wafer bumps (rectangular bumps, ball bumps), ball grid arrays (BGA), liquid crystal display (LCD) substrates, spacers used in filters, etc. The present invention relates to a microprojection inspection apparatus for detecting defects.

  In order to cope with the downsizing, high functionality, and high speed of digital information devices in recent years, bumps are used as a structure that can mount I / O terminals with high density and excellent high-frequency characteristics when connecting semiconductor chips. The application of flip chip mounting has been expanding.

  At this time, if variations occur in the size of the bumps formed on the chip, there is a possibility that a connection failure to the mounting substrate or a short circuit between the bumps may be caused.

  For this reason, it is important to judge the quality of the bumps formed on the chip by inspection before mounting, but the number of bumps may be several thousand per chip, and it is very difficult to visually inspect. Have difficulty. For this reason, an apparatus for automatically inspecting this is required, and the technology for that purpose has been made public.

  Here, optical inspection techniques are all described, but for example, Patent Document 1 discloses a technique related to height measurement by white interferometry. Patent Document 2 discloses a technique related to height measurement by the confocal method. Patent Document 3 discloses a technique related to height measurement by triangulation. Patent Document 4 discloses a technique related to height measurement by the knife edge method.

JP 2001-066612 A Japanese Patent No. 0336858 Japanese Unexamined Patent Publication No. 3180198 Japanese Patent Laid-Open No. 2002-22415

  However, in the methods of Patent Documents 1 and 2, it is necessary to take a plurality of images in which the focal position is changed in height detection, and there is a problem that the inspection throughput is reduced accordingly. The methods disclosed in Patent Documents 3 and 4 are advantageous in terms of inspection throughput because it is not necessary to change the focal position during detection. However, along with the recent progress in higher mounting density, the bump size (pitch) has also been miniaturized. For prismatic bumps, the bump width is 15 to 10 μm (pitch 30 to 20 μm), and for ball bumps, the bump diameter is 50 to 30 μm (pitch 100). In the methods of Patent Documents 3 and 4, it has become difficult to measure these fine bumps. This will be described below.

  Here, regarding the knife edge method among the above-described conventional techniques, the detection principle and the problem to cope with bump miniaturization will be described in detail. The knife edge optical system is a system that is also used as an optical system for detecting fluctuations in disk substrate height position in an optical disk drive. FIG. 1 is a diagram for explaining the principle. In FIG. 1, the illumination optical system is omitted, and only the detection optical system is described.

  A reflected light beam 110 from the surface of the sample 100 to be inspected is collected by the detection light collecting means 101 to form a detection light beam 111. This is focused on the optical sensor 104 by the image condensing means 103 to form a detection light spot 113. At this time, the detection light beam is shielded by the beam shielding means 102 installed immediately after the detection condensing means 103. Thus, a semicircular light beam 112 is formed.

  As shown in FIG. 1 (a), when the height position (z-direction position) of the sample 100 to be inspected is at the focal position of the detection optical system, the focal position 113a of the detection light spot is also the position of the optical sensor 104. A detection light spot 114a on the surface 104 is formed at the center of the optical sensor 104 in the xy direction.

  On the other hand, when the specimen 100 to be inspected moves in the positive direction of the z position as shown in FIG. 1B, the focal position 113b of the detection light spot is also in the positive direction of the z position as compared to the case of FIG. Will be moved to. At this time, the detection light spot 114b on the surface of the optical sensor 104 is out of focus, and is formed in a state of moving to the minus side in the x direction due to the effect of the beam shielding means 102.

  Further, when the sample 100 to be inspected moves in the minus direction of the z position as shown in FIG. 1C, the focal position 113c of the detection light spot is also minus in the z position as compared with the case of FIG. Will be moved to. At this time, the detection light spot 114c on the surface of the optical sensor 104 is out of focus, and is formed in a state of being moved to the plus side in the x direction by the effect of the beam shielding means 102.

The movement amounts of the detection light spots 114b and 114c on the optical sensor 104 in FIGS. 1B and 1C reflect the movement amount in the height direction (z direction) of the sample 100 to be inspected. In this way, it is possible to detect the movement direction (plus / minus) in the height direction and the movement amount of the sample to be inspected as the movement direction and movement amount of the detection light spot 114 on the photosensor surface. The detection of the moving direction and the moving amount of the detection light spot on the optical sensor surface is performed by using, for example, a two-segment sensor for the optical sensor 104, the output A from the first region 104a, and the output B from the second region 104b. It becomes possible by comparing. Here, when the amount of movement in the height direction of the sample to be inspected is Δz,
△ z∝ (A−B) / (A + B) ... Formula (1)
Can be expressed as In Fig. 1 (a), the amount of movement of the specimen to be inspected in the height direction is ∆z = 0, the output of the 2-part sensor corresponding to this is A = B, and the right side of equation (1) is also 0 .

  In FIG. 1 (b), the amount of movement Δz> 0 of the specimen to be inspected is A, and the output of the two-divided sensor corresponding to this is A> B, and the right side of equation (1)> 0 is there. In FIG. 1 (c), the amount of movement Δz <0 in the height direction of the sample to be inspected, and the output of the two-divided sensor corresponding to this is A <B, and the right side of equation (1) is <0. The movement amount is expressed in relation to the absolute value of (A−B).

  Here, the denominator (A + B) on the right side of equation (1) means normalization of the amount of spot movement by the total amount of light reaching the optical sensor 104, thereby reducing the influence of the reflectance of the sample to be inspected. Thus, only the amount of movement of the sample to be inspected can be detected.

  FIG. 2 explains the prior art of the bump height inspection apparatus using this knife edge optical system. The bump height inspection apparatus in FIG. 2 forms an illumination light spot on the surface of the specimen 100 to be inspected by a laser light source 201, a polygon mirror 202, an fθ lens 203, an illumination-side imaging lens 204, a half mirror 205, and an objective lens 206. The reflected light from the surface of the specimen 100 to be inspected is collected by the illumination optical system, the objective lens 206, the half mirror 205, the knife edge 207, the detection-side imaging lens 208, the cylindrical lens 209, and the optical sensor 210. It comprises a detection optical system that forms a detection light spot on 210. In addition, the sample 100 to be inspected is placed on the stage 250 and can move in the x, y, and z directions.

  Here, in the illumination optical system, the laser light source 201 is an illumination light beam generating means, and generates a predetermined illumination light beam. The polygon mirror 202 is beam deflecting means for deflecting the illumination light beam so as to reciprocate in the y direction. The fθ lens 203 and the illumination-side imaging lens 204 function to uniformly form an illumination light spot at each position of the detection area 2001 of the sample 100 to be inspected. The illumination light beam is reflected back by the half mirror 205 to form an illumination spot on the sample 100 to be inspected via the objective lens 206.

  In the detection optical system, the objective lens 206 is detection condensing means for condensing the reflected light from the sample 100 to be inspected to form a detection light beam. The detection light beam is transmitted through the half mirror 205 to be a beam. A semi-circular light beam is formed by the knife edge 207 serving as the shielding means and the detection-side imaging lens 208 serving as the image condensing means, and this is condensed to form a detection light spot on the optical sensor 201.

  At this time, since the illumination light spot is scanned on the specimen 100 to be inspected, the beam scanning is also performed at the positions of the knife edge 207 and the detection-side imaging lens 208. The scanning direction and the edge of the knife edge 207 are also scanned. It is necessary to arrange so that the directions coincide. The cylindrical lens 209 functions to collect the scanning detection light beam on the optical sensor 210. If the cylindrical lens 209 is not provided, the optical sensor 210 needs to be large enough to receive all the scanning detection light beams. However, a small optical sensor can be used by condensing with the cylindrical lens as shown in FIG. Is possible.

  The optical system is configured as shown in FIG. 2, and the specimen 100 is moved on the stage 250 so that the detection area 2001 of the optical system covers the examination area of the specimen 100, so that the optical specimen can be rapidly moved onto the specimen. The formed bump can be inspected.

  However, the conventional system described in Fig. 2 has a problem that it is difficult to cope with bump miniaturization in recent years due to restrictions on the optical system. This will be described below.

First, the structure of the wafer bump to be inspected will be described. FIG. 3 shows the structure of a prismatic bump. Square pillar bumps are a method often used in driver IC chips for flat panel displays such as liquid crystal displays or plasma displays.

  3A is a semiconductor wafer on which a plurality of IC chips 302 are formed. FIG. 3B is a diagram showing the arrangement of the prismatic bumps 303 in one chip 302. FIG. 3C is a diagram showing a cross-sectional structure of the bump. An electrode layer (UMB = Under Bump Metal) 305 is formed on the wiring pattern 304 formed on the uppermost portion of the semiconductor chip 302 (illustration of the cross-sectional structure is omitted). And prismatic bumps 303 are formed thereon. Reference numeral 306 denotes a protective layer formed of an insulator. The prism bumps are formed by a gold plating process in a wafer state before being cut into chips. In recent years, the miniaturization of chips has progressed in order to increase the mounting density and improve the mass productivity of driver ICs. At the same time, the bump width 3031 and the bump pitch 3032 have been miniaturized.

  FIG. 4 shows the structure of the ball bump. Ball bumps are a method often used in the implementation of high-performance ICs such as microprocessors and image processing LSIs. Reference numeral 401 in FIG. 4A denotes a semiconductor wafer on which a plurality of IC chips 402 are formed. FIG. 4B is a diagram showing the arrangement of ball bumps 403 in one chip 402. Ball bumps are formed on the entire surface of the chip and are capable of forming many input / output terminals required for high-performance ICs.

  FIG. 4C is a diagram showing a cross-sectional structure of the bump. An electrode layer (UMB = Under Bump Metal) 405 is formed on the rewiring pattern 404 formed on the uppermost portion of the semiconductor chip 402 (the cross-sectional structure is not shown). And a ball bump 303 is formed thereon. Reference numeral 306 denotes a protective layer formed of an insulator. Ball bumps are formed by a solder reflow process in a wafer state before being cut into chips. In recent years, with the improvement in performance of microprocessors and ICs for image processing, the number of input / output terminals has increased, and along with that, the bump width 4031 and bump pitch 4032 have been miniaturized.

  FIG. 5 shows problems at the time of prismatic bump height inspection in the conventional method. In FIG. 5, 100 is a sample to be inspected, 303 is a prismatic bump formed on the sample to be inspected, 101 is a detection condensing means, 102 is a beam shielding means, and 103 is an imaging condensing means. The edge direction of the beam shielding means 102 is shown as extending in the y direction.

  FIG. 5A is a diagram for explaining a detection state when a plurality of prismatic bumps are arranged in the y direction. In this case, the reflected beam 110 from the surface of the sample 100 to be inspected is directly incident on the detection condensing means 101 and reaches the detection condensing means 103 and the beam shielding means 102. The detection is performed without any problem.

  On the other hand, FIG. 5B is a diagram for explaining a detection state when a plurality of prismatic bumps are arranged in the x direction. The reference numerals in the figure are the same as those in FIG. In this case, since the reflected light beam 110 from the sample to be inspected is reflected by the side wall of the prismatic bump 303 and reaches the detection condensing means 101, the height position of the upper surface of the sample 100 to be inspected cannot be detected correctly. , This may be erroneously detected at the position of 501 for example.

  As described with reference to FIG. 2, in the prior art, the edge direction of the beam shielding means 102 needs to coincide with the beam scanning direction, and therefore the difference in measurement accuracy due to the directionality of the prismatic bumps as described with reference to FIG. Will occur. The difference in detection accuracy due to the difference in the xy direction of the prismatic bumps in the prior art tends to increase as the space of the prismatic bumps becomes narrower with the recent miniaturization of the prismatic bumps. It is in. In order to solve this without changing the optical system, for example, it is necessary to cope with the double inspection method in which the sample to be inspected is rotated 90 ° and inspected again. The problem arises that height must be sacrificed.

In the prior art, a difference in measurement accuracy due to directionality also occurs during ball bump inspection. This will be described with reference to FIG. In FIG. 6A, 100 is a sample to be inspected, 403 is a ball bump formed on the sample to be inspected, 101 is a detection condensing means, 102 is a beam shielding means, and 103 is an image condensing means. The edge direction of the beam shielding means 102 is shown as extending in the y direction as in FIG.

  FIG. 6B shows the directions of the light beams of the illumination beam 6021 and the reflected beam 6041 when the top portion 6031 of the ball bump 403 is detected. The illumination beam 6021 is incident on the ball bump 403 in the vertical direction, and the reflected beam 6041 is also reflected in the vertical direction. On the other hand, FIG. 6C shows the directions of the light beams of the illumination beam 6022 and the reflected beam 6042 when the slope 6032 of the ball bump 403 is detected. The illumination beam 6022 is incident on the ball bump 403 perpendicularly, while the reflected beam 6042 is emitted obliquely.

  FIG. 6D illustrates an example in which detection points on the ball bumps are taken in the y direction, and five points from 603a1 to 603a5 are evenly taken around the bump top portion 603a3. Similarly, FIG. 6 (e) shows an example in which the detection points on the ball bump are taken in the x direction, and five points from 603b1 to 603b5 are equally taken around the bump top portion 603b3. . Here, it is assumed that the intervals between the detection points in FIGS. 6D and 6E are the same. The interval in this case is an interval projected on the diameter.

  In FIG. 6 (f), the reflected light beam from each detection point shown in FIG. 6 (d) becomes the detection light beam through the detection condensing means 101, and reaches the position of the beam shielding means 102 and the imaging condensing means 103. It shows the position when it arrives. The reflected light from the bump top portion 603a3 reaches the center 605a3 of the imaging condensing means 103, and the reflected light beams emitted obliquely from the bump inclined surfaces 603a2 and 603a4 reach 605a2 and 605a4 moved in the y direction.

  Although 605a3 and 605a2 and 605a4 also have beam position movements, both are on the y-axis, and therefore can be partially shielded by the beam shielding means 102, and corresponding detection points 603a2, 603a3, It is possible to detect the height of 603a4. However, the reflected light from the detection points 603a1 and 603a5 whose inclination is larger than a certain degree cannot enter the detection condensing means 101 or the imaging condensing means 103 (605a5 and 605a5), and the reflected light at that position is high. Detection is impossible.

  In FIG. 6 (g), the reflected light beam from each detection point shown in FIG. 6 (e) becomes a detection light beam through the detection condensing means 101, up to the position of the beam shielding means 102 and the imaging condensing means 103. It shows the position when it arrives. The reflected light from the bump top portion 603b3 reaches the center 605b3 of the imaging condensing means 103, and the reflected light beams emitted obliquely from the bump inclined surfaces 603b2 and 603b4 reach 605b2 and 605b4 moved in the x direction.

  Unlike the case of FIG. 6 (f), only the beam shielding means 102 can partially shield the beam 605b3, and the beam 605b2 moved in the x direction is not shielded by the beam shielding means 102 at all. 605b4 is completely shielded by the beam shielding means 102, and it is impossible to detect the height of the corresponding detection points 603b2 and 603b4. Furthermore, the reflected light from the detection points 603b1 and 603b5 whose inclination has increased to a certain degree cannot enter the detection condensing means 101 or the imaging condensing means 103 (605b1 and 605b5), and the reflected light at that position is high. Detection is impossible.

  As described with reference to FIG. 2, since the edge direction of the beam shielding means 102 needs to coincide with the beam scanning direction in the prior art, the difference in measurement accuracy due to the direction of the ball bump described with reference to FIG. Will occur. The difference in detection accuracy due to the difference in the xy direction of the ball bump in the prior art occurs because the curvature of the bump surface increases (the radius of curvature decreases) with the recent miniaturization of the ball bump. There is a tendency to increase the possibility. In order to solve this without changing the optical system, for example, it is necessary to cope with the double inspection method in which the sample to be inspected is rotated 90 ° and inspected again. The problem arises that you have to sacrifice height

  The present invention has the following configuration in order to solve the above-described problems and achieve the object of performing high-throughput and high-accuracy bump height inspection.

That is, the invention according to claim 1 is an illumination condensing unit that condenses the illumination light beam from the illumination light beam generating unit to form an illumination light spot on the surface of the sample to be inspected, and the illumination light spot is inspected. An illumination optical system having a beam deflecting unit that deflects the illumination light beam so as to scan the surface of the sample, and a detection condensing unit that condenses the reflected light from the surface of the sample to be inspected to form a detection light beam A detection optical system, signal processing means for calculating and outputting height information of microprojections on the sample to be inspected based on an output signal output from an optical sensor , and mounting the sample to be inspected Te, a stage means for relatively moving with respect to the illumination optical system and the detection optical system, and Luz stage control means to drive control the stage means, the beam deflecting means drive control to ruby over beam deflection control Have means and In the microprojection inspection apparatus, the detection optical system includes means for splitting the detection light beam to form a first detection light beam and a second detection light beam, and the first detection light beam partially. A first beam shielding means for shielding the first semicircular beam to form a first detection light spot by condensing the first semicircular beam. Optical means, a first optical sensor that detects the first detection light spot, and a second beam shield that partially shields the second detection light beam to form a second semicircular beam Means, second imaging condensing means for condensing the second semicircular beam to form a second detection light spot, and a second optical sensor for detecting the second detection light spot has the door, said signal processing means, or wherein the first output signal from the optical sensor a second optical sensor Are those from the output signal and outputs the calculated height information of the minute projections of the inspection on the specimen, the first beam shielding means has a knife edge, the second beam shielding means knife has an edge, straight portion of the first beam knife edge of the knife edge and the second beam shielding means shielding means the first semicircular beam straight portion and the second semicircular beam Are arranged so as to be orthogonal to each other, and the detection optical system includes beam deflection means of the illumination optical system, and the beam deflection means allows the detection light beam to pass in a direction opposite to the illumination light beam. This is a microprojection inspection apparatus.

  As a result, the height inspection of the minute protrusions on the sample to be inspected can be performed at high speed with uniform sensitivity regardless of the direction.

According to a second aspect of the present invention, there is provided illumination condensing means for condensing the illumination light beam from the illumination light beam generating means to form an illumination light spot on the surface of the sample to be inspected, and the illumination light spot to be covered. An illumination optical system having a beam deflecting means for deflecting the illumination light beam so as to scan the surface of the inspection sample, and detection condensing for condensing the reflected light from the surface of the inspection sample to form a detection light beam A detection optical system having means, signal processing means for calculating and outputting height information of a microprojection on the sample to be inspected based on an output signal output from an optical sensor , and mounting the sample to be inspected to a stage means for relatively moving with respect to the illumination optical system and the detection optical system, and Luz stage control means to drive control the stage means, the beam deflecting means drive control to ruby over beam deflection With control means In the small projection inspection apparatus, the detection optical system includes means for splitting the detection light beam to form a first split light and a second split light, and a semicircular first detection of the first split light. A first prism whose reflecting surface is a reflecting surface that is divided into a light beam and a semicircular second detection light beam, and a first detection light spot formed by condensing the first detection light beam. detecting the first and imaging focusing means, and a second imaging condensing means for forming a second detection light spot by focusing the second detection light beam, the first detection light spot A first optical sensor; a second optical sensor for detecting a second detection light spot; and a second detection light beam having a semicircular third detection light beam and a second detection light beam. A second prism whose inclined surface is a reflecting surface, and a third detection light spot by condensing the third detection light beam. A third imaging condensing means for forming a fourth image focusing means for forming a fourth detection light spot by focusing the fourth detection light beam, said third detection light spot And a fourth optical sensor for detecting a fourth detection light spot, wherein the signal processing means includes the first optical sensor, the second optical sensor, and a third optical sensor. The height information of the microprojections on the sample to be inspected is output from the output signals from the optical sensor and the fourth optical sensor, and is output. The top side of the first prism and the first prism The prisms of the two prisms are orthogonal to each other, and the detection optical system includes a beam deflecting unit of the illumination optical system, and the beam deflecting unit causes the detection light beam to travel in a direction opposite to the illumination light beam. It is a microprojection inspection apparatus characterized by passing .

  This makes it possible to perform high-speed inspection of the height of minute protrusions on the sample to be inspected with uniform sensitivity regardless of direction, and at the same time reduce the influence of disturbances such as thermal fluctuations and vibrations. High-precision inspection can be performed.

  According to a third aspect of the present invention, there is provided the microprojection inspection apparatus according to the first and second aspects, wherein the beam deflecting means is a polygon mirror.

  According to a fourth aspect of the present invention, there is provided the microprojection inspection apparatus according to the first or second aspect, wherein the beam deflection means is a galvanometer mirror.

  According to a fifth aspect of the present invention, there is provided the microprojection inspection apparatus according to the first or second aspect, wherein the optical sensor means is a two-divided sensor.

  According to a sixth aspect of the present invention, there is provided the microprojection inspection apparatus according to the first or second aspect, wherein the optical sensor means is a position sensor.

  According to a seventh aspect of the present invention, in the first and second aspects, the illumination optical system means includes an illumination light beam diameter control means for changing the beam diameter of the illumination light beam. This is a projection inspection apparatus. This makes it possible to flexibly set the detection range and sensitivity according to the inspection target.

  According to an eighth aspect of the present invention, in the seventh aspect, the illumination light beam diameter control means operates so as to change the beam diameter of the illumination light beam according to the size of the minute projection to be inspected. This is a microprojection inspection apparatus. This makes it possible to flexibly set the detection range and sensitivity according to the inspection target.

  According to a ninth aspect of the present invention, in the first and second aspects, the detection optical system means further includes a detection light beam diameter control means for changing the beam diameter of the detection light beam. This is a projection inspection apparatus. This makes it possible to flexibly set the detection range and sensitivity according to the inspection target.

  According to a tenth aspect of the present invention, in the ninth aspect, the detection light beam diameter control means operates so as to change the beam diameter of the detection light beam according to the size of the minute protrusion to be inspected. This is a microprojection inspection apparatus. This makes it possible to flexibly set the detection range and sensitivity according to the inspection target.

  According to an eleventh aspect of the present invention, in the ninth and tenth aspects, the detection light beam diameter control means is a Keplerian beam expander, and a pinhole is provided at a condensing point in the beam expander. Is a microprojection inspection apparatus characterized by the above. As a result, the influence of stray light from the sample to be inspected is reduced, and highly accurate detection is possible.

  According to a twelfth aspect of the present invention, there is provided a microprojection inspection apparatus according to the first and second aspects, wherein the illumination light beam generating means is a laser light source.

  According to a thirteenth aspect of the present invention, there is provided the microprojection inspection apparatus according to the first or second aspect, wherein the illumination light beam generating means is a light source having a plurality of wavelengths. As a result, errors due to speckle due to the coherence of illumination light can be reduced, and highly accurate detection can be achieved.

  According to a fourteenth aspect of the present invention, in the first aspect, the signal processing means compares the first and second output signals output from the first and second photosensors with a predetermined value and is normal. This is a microprojection inspection device characterized by performing abnormality determination and outputting only normal signals as height information. As a result, the height inspection of the minute protrusions on the sample to be inspected can be performed at high speed with uniform sensitivity regardless of the direction.

  According to a fifteenth aspect of the present invention, in the second aspect, the signal processing unit compares the first to fourth output signals output from the first to fourth optical sensors with a predetermined value and is normal. This is a microprojection inspection device characterized by performing abnormality determination and outputting only normal signals as height information. As a result, the height inspection of the minute protrusions on the sample to be inspected can be performed at high speed with uniform sensitivity regardless of the direction.

  According to a sixteenth aspect of the present invention, in the first aspect, the microprojection inspection apparatus has information storage means for storing design information of the sample to be inspected, and the signal processing system includes the beam deflection control means. A signal to be output as height information from the first and second output signals based on a comparison between the inspection position information of the inspection sample obtained from the stage control means and the design information of the inspection sample. Is a microprojection inspection apparatus characterized by the above. As a result, the height inspection of the minute protrusions on the sample to be inspected can be performed at high speed with uniform sensitivity regardless of the direction.

  According to a seventeenth aspect of the present invention, in the second aspect, the microprojection inspection apparatus has information storage means for storing design information of the sample to be inspected, and the signal processing system includes the beam deflection control means. A signal to be output as height information from the first to fourth output signals based on a comparison between the inspection position information of the inspection sample obtained from the stage control means and the design information of the inspection sample. Is a microprojection inspection apparatus characterized by the above. As a result, the height inspection of the minute protrusions on the sample to be inspected can be performed at high speed with uniform sensitivity regardless of the direction.

  According to the present invention, with the above-described configuration, it is possible to perform a bump height inspection at a high speed with a uniform sensitivity regardless of the direction even for a fine bump.

  Hereinafter, the contents of the present invention will be described in detail according to examples.

  The contents of the first embodiment will be described with reference to FIG. The bump height inspection apparatus in FIG. 7 includes a laser light source 201, a first half mirror 221, a polygon mirror 202, an fθ lens 203, an illumination-side imaging lens 204, a folding mirror 205, and an objective lens 206. An illumination optical system that forms a light spot on the surface, an objective lens 206 that collects reflected light from the surface of the specimen 100 to be inspected, a folding mirror 205, an illumination-side imaging lens 204, an fθ lens 203, a polygon mirror 202, and a first The detection light beam reflected by the first half mirror 221 is split by the second half mirror 222 through the half mirror 221, and the first knife edge 2071 and the first detection-side imaging lens 2081 A detection optical system that forms a detection light spot on the first optical sensor 2101, and forms a detection light spot on the second optical sensor 2102 by the second knife edge 2072 and the second detection-side imaging lens 2082 together. Consists of. In addition, the sample 100 to be inspected is placed on the stage 250 and can move in the x, y, and z directions.

  Here, in the illumination optical system, the laser light source 201 is illumination light generating means, and generates a predetermined illumination light beam. The polygon mirror 202 is a beam deflecting unit that deflects the illumination light beam so as to reciprocate in the y direction. The fθ lens 203 and the illumination-side imaging lens 204 are part of the illumination light condensing means that functions to uniformly form an illumination light spot at each position of the detection region 2001 of the sample 100 to be inspected. The illumination light beam is reflected back by the folding mirror 205, and forms an illumination spot on the sample 100 to be inspected via the objective lens 206. Here, the objective lens 206 is also a part of the illumination light condensing means.

  In the detection optical system, the objective lens 206 is a part of detection light beam forming means for condensing the reflected light from the sample 100 to be inspected to form a detection light beam, and the detection light beam is reflected by the folding mirror 205. Then, the light reaches the first half mirror 221 via the illumination-side imaging lens 204, the fθ lens 203, and the polygon mirror 202. Here, the illumination-side imaging lens 204, the fθ lens 203, and the polygon mirror 202 are also part of the detection light beam forming means. In this way, by returning the detection light beam to the same optical path as that of the illumination optical system, unlike the prior art described with reference to FIG. 2, it is possible to form a detection light beam that is not deflected at the knife edge position. This makes it possible to install the beam shielding means without being restricted by the direction of beam deflection.

  The detection light beam reflected from the first half mirror 221 is divided by the second half mirror 222 to form two detection light beams, and the first detection light beam is a first beam shielding means. The first semicircular light beam is formed and condensed by the knife edge 2071 and the first imaging condenser lens 2081 which is the first imaging condenser means. A detection light spot is formed on the two-divided sensor 201.

  The second detection light beam is formed into a second semicircular shape by a second knife edge 2072 as second beam shielding means and a second imaging condenser lens 2082 as second imaging condenser means. A light beam is formed and condensed to form a detection light spot on the second two-divided sensor 2012 that is the second optical sensor. At this time, the first knife edge 2071 and the second knife edge 2072 are arranged so that the edge directions are perpendicular to each other. By constructing such an xy knife-edge optical system, the height of the bump that does not depend on directionality is compensated for by compensating for the xy direction difference in measurement accuracy of the single knife-edge optical system described in FIGS. Inspection is possible.

  In the embodiment shown in FIG. 7, the beam deflecting means is described as a polygon mirror, but it goes without saying that it may be constituted by, for example, a galvanometer mirror.

  In the embodiment shown in FIG. 7, the optical sensor is described as a two-divided sensor. However, the optical sensor is composed of, for example, an optical position sensor, a one-dimensional CCD light receiving element, a two-dimensional CCD light receiving element, or a CMOS light receiving element. It goes without saying that it is also good.

  In the embodiment shown in FIG. 7, the case where the two detection light beams divided by the second half mirror 222 are shielded by the first and second beam shielding means orthogonal to each other has been described. Further, a third and a fourth half mirror are provided to form four detection light beams. In addition to the first and second beam shielding means and the imaging lens / light sensor, the first and second beam shieldings are performed. Needless to say, it is possible to further eliminate the direction difference in measurement accuracy by installing the third and fourth beam shielding means and the imaging lens / light sensor having a 45 ° inclination with respect to the means.

  The optical system is configured as shown in FIG. 7, and the specimen 100 is moved on the stage 250 so that the detection area 2001 of the optical system covers the inspection area of the specimen 100, so that the optical specimen can be rapidly moved onto the specimen. The formed bump can be inspected. At this time, the stage driving unit 701 functions to drive the stage 250 in an arbitrary direction and detect the state of the stage (current position / velocity of x, y, z, etc.). Further, the deflector driving means 702 functions to drive the polygon mirror at an arbitrary number of rotations and detect the state (rotation speed, rotation angle, etc.) of the polygon mirror.

  Further, the signal processing means 703 takes in the output signals from the optical sensors 2101 and 2102 and A / D converts them, and based on the processing equation as described in the equation (1), height information of each point of the sample to be inspected. Is calculated. These controls and signal processing are comprehensively controlled by the overall control means 700 so that inspection can be performed in an optimum state for each type of sample 100 to be inspected. The storage device 710 also stores information on the sample to be inspected, such as sample size, chip arrangement / bump arrangement on the sample, inspection recipe to be executed on the inspected sample, and inspection results of each inspected sample. It is something to keep.

  A second embodiment of the inspection apparatus according to the present invention will be described with reference to FIGS. In these drawings, components having the same reference numerals as those shown in the above are the same as those described above, and thus description thereof is omitted here.

FIG. 8 shows a configuration in which the height is detected by two knife edges. A portion where light is shielded by the beam shielding means 1021 in FIG. 8A is transmitted through the beam shielding means 1022 in FIG. 8B, and a portion where light is transmitted through the beam shielding means 1021 in FIG. The beam shielding means 1022 in FIG. 8B is configured to shield, and the detection light spots 1061 and 1062 are formed on the respective optical sensors 1041 and 1042 by the respective imaging condensing lenses 1031 and 1032.
At this time, for example, by using a two-divided sensor as the photosensor, the output A from the first region 104a and the output B from the second region 104b of the first photosensor and the first region 104c of the second photosensor Is compared with the output D from the second region 104d to determine the movement direction (plus / minus) of the height direction of the sample 100 to be inspected and the amount of movement of the detection light spot 114 on the optical sensor surface. It is possible to detect the movement direction and the movement amount. For example, when the amount of movement of the specimen 100 in the height direction is Δz,
Δz∝ {(A−B) + (D−C) / {(A + B) + (C + D)} (2)
Can be expressed as FIGS. 8A and 8B show a state of Δz> 0, where A> B in the first photosensor and D> C in the second photosensor.

  The advantage of using a two-system knife-edge optical system composed of two beam shielding means having two different beam shielding positions and two optical sensors will be described with reference to FIG. FIG. 9 shows a detection state of the knife edge optical system when the reflected light beam 110 from the sample 100 to be inspected is tilted. In an actual inspection apparatus, there are various disturbance factors that cause the reflected light beam 110 from the sample 100 to be inspected to be inclined, such as thermal fluctuations and vibrations, or manufacturing errors on each surface of the polygon mirror itself. The configuration shown in FIG. 8 can reduce the influence of these disturbance factors. This will be described below.

  FIG. 9A shows a case where the reflected light beam 110 is tilted from the state of FIG. At this time, the detection light spot 1061 formed on the optical sensor 1041 is offset by reflecting the inclination of the reflected light beam 110. Therefore, even if the height of the sample 100 to be inspected is detected using only the output A from the first region 104a and the output B from the second region 104b of the first photosensor, Both the influence of the height of the specimen 100 to be inspected and the influence of the inclination of the detection light beam 110 are used, and if only this is attempted to detect the height of the specimen 100 to be inspected, a large detection error will occur. End up.

  Here, the configuration shown in FIG. 9B is also provided, and the output A from the first region 104a and the output B from the second region 104b of the first photosensor, and the first region 104c of the second photosensor. When the height detection is performed based on the output C and the output D from the second region 104d based on the expression (2) described above, the influence of the inclination of the detection light beam 110 is as follows. Output A from the first region 104a of the first photosensor due to the offset of the first optical sensor and decrease in output D from the second region 104d of the second photosensor due to the offset of the detection light spot 1062 Therefore, in the numerator {(A−B) + (D−C)} of the height information calculation formula of the sample 100 to be inspected shown in the equation (2), the influence of the inclination of the detection light beam 110 is small, It becomes possible to effectively detect the height information of the sample 100 to be inspected.

  FIG. 10 shows an example of the configuration of an optical system for realizing the two-system knife-edge optical system configuration described with reference to FIGS. The reflected light beam 110 from the sample to be inspected is condensed by the detection condensing means 101 to form the detection light beam 111. Immediately after the imaging condensing means 103, a prism whose both slopes are reflecting surfaces. 1020 is installed to form a detection light spot on the first optical sensor 1041 and the second optical sensor 1042 from the two semicircular beams 1121 and 1122 formed by reflection on both slopes of the prism. It is. By doing so, it is possible to realize a two-system knife-edge optical system with a simple configuration.

  FIG. 11 is a diagram for explaining an embodiment of the present invention in which the two-system knife-edge optical system using the prism described in FIG. 10 is mounted. In the figure, the same reference numerals as those in the previous figure (mainly FIG. 7) are the same as the previous explanations, and the explanation is omitted here. Of the two detection light beams divided by the second half mirror 222, the first detection light beam is divided into two by the first prism 231 installed behind the first imaging condenser lens 2081. A semicircular light beam is formed and condensed to form a detection light spot on the first optical sensor 2101 and the third optical sensor 2103.

  The second detection light beam is formed and condensed by the second prism 232 installed behind the second imaging condenser lens 2082 to form a second semicircular light beam. Then, a detection light spot is formed on the fourth optical sensor 2104. At this time, the first prism 231 and the second prism 232 are arranged so that the edge directions are at right angles. With this configuration, the xy direction difference in the measurement accuracy of the knife edge optical system described in FIGS. 5 and 6 is compensated for at the same time, and at the same time, the height of the influence of the tilt of the reflected light beam described in FIG. 9 is small. A detection optical system can be configured.

  A third embodiment of the inspection apparatus according to the present invention will be described with reference to FIGS. In these drawings, components having the same reference numerals as those shown in the above are the same as those described above, and thus description thereof is omitted here.

  FIG. 12 is a diagram for explaining the beam diameter, measurement range, and sensitivity of the detection light beam 111. 12A shows a case where the beam diameter of the detection light beam 111 is large, and FIG. 12B shows a case where the beam diameter of the detection light beam 111 is small. The relationship between the beam diameter φa of the detection light beam 111 in FIG. 12A and the beam diameter φb of the detection light beam 111 in FIG. 12B is φa> φb.

  As shown in FIG. 12 (a), when the beam diameter of the detection light beam 111 is large, it is condensed on the optical sensor 104 by the imaging condensing means 103 at a large condensing angle (θa). The detected light spot 114 is reduced. As a result, the detection of the movement of the detection light spot 114 accompanying the change in the height of the specimen 100 (not shown) is detected by using the two-segment sensor for the optical sensor 104 and the output A and the first output from the first region 104a. When compared with the output B from the two regions 104b, the output A and B change greatly with a slight movement of the detection light spot 114, and the height change of the sample 100 to be inspected can be detected with high sensitivity. On the other hand, even if the height of the sample 100 to be inspected is slightly changed, the entire detection light spot 114 enters the first region 104a or the second region 104b of the two-divided sensor, and the measurement range becomes narrow. .

  On the other hand, as shown in FIG. 12B, when the beam diameter of the detection light beam 111 is small, it is condensed on the optical sensor 104 by the imaging condensing means 103 at a small condensing angle (θb). , The formed detection light spot 114 becomes larger. As a result, the detection of the movement of the detection light spot 114 accompanying the change in the height of the specimen 100 (not shown) is detected by using the two-segment sensor for the optical sensor 104 and the output A and the first output from the first region 104a. When the comparison is made by comparing the output B from the two regions 104b, the changes in the outputs A and B are moderate with respect to the movement of the detection light spot 114. As a result, the detection sensitivity of the height change of the sample 100 to be inspected is lower than in the case of FIG. 12A, but the measurement range can be widened.

  The size of bumps to be inspected is not limited to a few tens of μm, and some products have a size of several hundreds of μm, so the inspection device can flexibly change the detection sensitivity and detection range. It is important to be able to handle various products by configuring. A configuration for realizing this will be described with reference to FIGS.

  FIG. 13 is a diagram for explaining a configuration in which the illumination light beam diameter control means 1302 is installed in the illumination optical system in the embodiment shown in FIG. The illumination light beam diameter control means 1302 is, for example, a motor-driven zoom beam expander, and is configured such that the magnification can be arbitrarily changed by the beam diameter adjustment mechanism control means 1301. By controlling the beam diameter of the illumination light beam, the beam diameter of the detection light beam can be controlled as a result, and the optimum detection sensitivity / range can be set according to the inspection object. At this time, information on the relationship between the optimum detection sensitivity / range and beam diameter according to the inspection object is stored in the storage device 710 and reflected in the setting of the inspection recipe to be executed for the sample to be inspected, so that the overall control means 700 Is comprehensively controlled so that inspection can be performed in an optimum state for each type of sample 100 to be inspected.

  FIG. 14 is a diagram for explaining the configuration in which the detection light beam diameter control means 1303 is installed in the detection optical system in the embodiment shown in FIG. The detection light beam diameter control means 1303 is, for example, a motor-driven zoom beam expander, and is configured such that the magnification can be arbitrarily changed by the beam diameter adjustment mechanism control means 1301. By directly controlling the beam diameter of the detection light beam, it is possible to set the optimum detection sensitivity / range according to the inspection object. At this time, information on the relationship between the optimum detection sensitivity / range and beam diameter according to the inspection object is stored in the storage device 710 and reflected in the setting of the inspection recipe to be executed for the sample to be inspected, so that the overall control means 700 Is comprehensively controlled so that inspection can be performed in an optimum state for each type of sample 100 to be inspected.

  A fourth embodiment of the inspection apparatus according to the present invention will be described with reference to FIGS. In these drawings, components having the same reference numerals as those shown in the above are the same as those described above, and thus description thereof is omitted here.

  FIGS. 15A and 15B are diagrams for explaining the generation of stray light in the inspection, taking the inspection of the ball bump as an example. Since the surface of the substrate on which the bumps are formed has a complicated structure, reflected light from each structure is generated as stray light, which may cause detection errors. In FIG. 15 (a), 110 is a reflected light beam reflected only on the surface of the ball bump, which correctly transmits the height information of the bump, but 110a is generated by multiple reflection between a plurality of bumps. This is a reflected light beam of stray light, which causes detection errors. Further, as shown in FIG. 15 (b), in addition to a plurality of bumps, a reflected light beam 110b of stray light that is multiply reflected by the substrate is also generated, which also causes a detection error.

  FIG. 16 is a diagram for explaining an embodiment for reducing the detection error due to the stray light. The detection light detection light beam diameter control means 1303 described with reference to FIG. 14 is configured by a Kepler type beam expander, and a pinhole is installed at a condensing point in the beam expander. The Kepler-type beam expander is a beam in which a first convex lens 130a having a focal length f1 and a second convex lens 130b having a focal length f2 are separated by a distance (f1 + f2), and incident on the beam expander. When the diameter is φ1, the optical element functions so that the beam diameter to be emitted becomes φ2 = φ2 = f2 / f1 × φ1. In addition, a condensing point is formed between the first and second convex lenses.

  In this embodiment, the detection light beam before the beam expander is composed of normal reflected light 1111 and stray light component 1112. This is condensed by the first convex lens 130a, and a pinhole 130c is installed at the condensing point. Accordingly, the normal reflected light 1111 passes through the pinhole, but the disturbed stray light component 1112 is shielded, and the detection light beam 1113 emitted from the beam expander can be formed of the normal reflected light component. With this configuration, the influence of the stray light component can be reduced, and highly accurate detection is possible.

  A fifth embodiment of the inspection apparatus according to the present invention will be described with reference to FIG. In the figure, those having the same reference numerals as those in the previous drawings are the same as those described above, and thus the description thereof is omitted here.

  In the embodiment shown in FIG. 7, the configuration in which illumination light is generated by one laser light source 201 has been described. However, single-wavelength laser illumination has high coherence, for example, an error caused by a speckle pattern caused by bump surface roughness. May occur.

  A configuration for eliminating this will be described with reference to FIG. A laser light source 2011 that generates laser light of wavelength λ1 and 2012 that generates laser light of wavelength λ2 are installed, and two laser lights are transmitted by a wavelength separation mirror 223 that transmits light of wavelength λ1 and reflects light of wavelength λ2. Inspection is performed as one illumination light beam. As a result, the coherence of the illumination light can be reduced, and errors caused by the interference can be reduced. Although the light source is described as the laser light in the embodiments of FIGS. 7 and 17, the light source is not limited to this, and it is needless to say that white light or the like may be used.

  A sixth embodiment of the inspection apparatus according to the present invention will be described. The first embodiment (FIG. 7) and the second embodiment (FIG. 11) of the present invention have two knife edge optical systems in which the edge directions of the beam shielding means are orthogonal to each other. As described in FIG. 6, depending on the state of the inspection object, detection abnormality may occur in each optical system. For stable inspection, it is necessary to judge / remove these abnormal values and extract only normal output signals. In implementing this, a configuration is considered in which the signal processing means 703 compares the output signal of the optical sensor (104 in FIG. 1) with a predetermined value to perform normality / abnormality determination and extract only the normal signal. It is done.

A specific example in which normality / abnormality determination is performed in comparison with a predetermined value will be described below. Assuming that the optical sensor 104 is a two-divided sensor, when the output from the first region 104a is A and the output from the second region 104b is B,
(1) Set the judgment value i that determines whether the detection light beam is incident on the optical sensor.
If A + B <i, the detection light beam is not incident on the optical sensor = abnormal, and the output signal is excluded. This is to determine that the detection light beam has deviated from the optical system, as indicated by 605a1 and 605a5 in FIG. 6F and 605b1 and 605b5 in FIG. 6G.
(2) Set a judgment value j to determine the deviation of the detection light beam incident on the optical sensor. If A−B> j or B−A> j, the detection light beam is incident on the optical sensor. The output signal is excluded by judging that the deviation is abnormal. This is to determine the deviation of the detection light beam such that the knife edge does not function as shown in 605b2 and 605b4 in FIG. 6 (g).

  This can be applied to each of the two photosensors 1041 and 1042 in FIG. 7 and the four photosensors 1041, 1042, 1043, and 1044 in FIG. In this way, it is possible to perform normality / abnormality determination of the output from each optical sensor and to inspect only the normal signal.

A seventh embodiment of the inspection apparatus according to the present invention will be described. The sixth embodiment is an example in which the output signal from each optical sensor is selected using the output signal itself, but the output signal may be selected in advance based on the structure of the sample to be inspected. Is possible. This will be described below.

As shown in FIGS. 3 and 4, the structure on the sample to be inspected (in this case, the semiconductor wafer substrate), that is, the arrangement of chips and bumps on the substrate, the height and size of the bumps, etc. It is known in advance based on information. These design information and output signal selection conditions determined from the design information are stored in the storage device 710, and can be obtained from the stage position obtained from the stage control means 701 and the deflector control means and the deflection angle of the deflector during the inspection. The output signal is selected by comparing with the current inspection position. As a result,
The inspection can be performed by the normal signal extracted based on the design information.

  In the sixth and seventh embodiments, examples of signal selection by the respective methods have been described, but it goes without saying that both may be used together.

  In the above description, the inspection of the bump formed on the semiconductor chip has been described. However, the present invention is not limited to the spacer used for a liquid crystal display (LCD) substrate or a filter, a partition of a plasma display, or the like. Needless to say, the present invention can be applied to a microprojection inspection for measuring the height of a microprojection in an electronic device product or the like and detecting the defect.

The figure which shows the principle of the optical height detection by a knife edge system. The figure which shows the structure of the prior art in a microprojection inspection apparatus. The figure explaining the prism pillar bump formed on the semiconductor wafer. The figure explaining the ball bump formed on the semiconductor wafer. The figure which shows the subject of the prior art in the case of test | inspecting a prismatic bump. The figure which shows the subject of the prior art in the case of test | inspecting a ball bump. The figure which shows the 1st schematic structure of the microprojection inspection apparatus of this invention. The figure which shows the structure of the 2 system knife edge system of this invention The figure which shows the detection state when an optical axis inclines due to disturbance etc. in the 2 system knife edge system of this invention. The figure which shows the structure of 2 system | strain knife edge system using the prism of this invention. The figure which shows the 2nd structure of the microprojection inspection apparatus of this invention. The figure explaining the relationship between a detection light beam diameter, a measurement range, and sensitivity in this invention. The figure which shows the structure which provided the illumination light beam diameter control means of this invention. The figure which shows the structure which provided the detection light beam diameter control means of this invention. The figure explaining the stray light generated when illuminating a ball bump. The figure which shows the structure of the Kepler type beam expander which installed the pinhole in the detection light beam diameter control means of this invention. The figure which shows the structure which installed the several light source from which a wavelength differs in the illumination light beam generation means of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Sample to be examined, 101 ... Detection condensing lens, 102 ... Beam shielding means, 103 ... Imaging condensing lens, 104 ... Optical sensor, 110 ... Reflected light beam, 111 ... Detection light beam, 112 ... Semicircular beam , 114 ... detection light spot, 201 ... laser light source, 202 ... polygon mirror (beam deflection means), 203 ... fθ lens, 204 ... illumination-side imaging lens, 205 ... half mirror, 206 ... objective lens, 2001 ... optical system detection Area 208: Detection-side imaging lens 207 Knife edge (beam shielding means) 209 Cylindrical lens 210 Optical sensor 250 Stage 301 301 401 Semiconductor wafer 303 Square prism bump 403 Ball bump , 700 ... Overall control means, 701 ... Stage control means, 702 ... Deflector control means, 703 ... Signal processing means, 231, 232 ... Prism, 1301 ... Beam diameter adjustment mechanism control means, 1302 ... Illumination side beam diameter adjustment mechanism, 1303 ... Detection side beam diameter adjustment mechanism, 110 ... Positive Normal reflection light, 110a, 110b ... stray light (abnormal reflection light), 1111 ... normal detection beam, 1112 ... abnormal detection beam, 130c ... pinhole, 2012 ... second laser light source.

Claims (2)

Illumination condensing means for condensing the illumination light beam from the illumination light beam generating means to form an illumination light spot on the surface of the sample to be inspected, and the illumination so that the illumination light spot scans the surface of the sample to be inspected An illumination optical system having beam deflecting means for deflecting the light beam;
A detection optical system having detection condensing means for condensing the reflected light from the surface of the sample to be inspected to form a detection light beam;
Signal processing means for calculating and outputting height information of microprojections on the sample to be inspected based on an output signal output from the optical sensor;
Stage means for placing the sample to be inspected and moving it relative to the illumination optical system and the detection optical system, stage control means for driving and controlling the stage means, and driving and controlling the beam deflection means In a microprojection inspection apparatus having a beam deflection control means,
Means for splitting the detection light beam to form a first detection light beam and a second detection light beam;
First beam shielding means for partially shielding the first detection light beam to form a first semicircular beam; and the first detection light by condensing the first semicircular beam. A first imaging and focusing means for forming a spot; a first optical sensor for detecting the first detection light spot;
A second beam shielding means for partially shielding the second detection light beam to form a second semicircular beam; and a second detection light for condensing the second semicircular beam. Second imaging and focusing means for forming a spot; and a second optical sensor for detecting the second detection light spot;
The signal processing means calculates and outputs height information of a microprojection on the sample to be inspected from an output signal from the first optical sensor and an output signal from the second optical sensor. ,
The first beam shielding means has a knife edge, the second beam shielding means has a knife edge, and the knife edge of the first beam shielding means and the knife edge of the second beam shielding means are The linear part of the first semicircular beam and the linear part of the second semicircular beam are arranged to be orthogonal to each other;
The detection optical system includes a beam deflecting unit of the illumination optical system, and the beam deflecting unit passes the detection light beam in a direction opposite to the illumination light beam .
Having storage means for storing design information of the specimen to be inspected;
The signal processing means is based on a comparison between the inspection position information of the inspection sample obtained from the beam deflection control means and the stage control means and the design information of the inspection sample. A microprojection inspection apparatus , wherein any one of output signals is selected and height information is calculated and output . Illumination condensing means for condensing the illumination light beam from the illumination light beam generating means to form an illumination light spot on the surface of the sample to be inspected, and the illumination so that the illumination light spot scans the surface of the sample to be inspected An illumination optical system having beam deflecting means for deflecting the light beam;
A detection optical system having detection condensing means for condensing the reflected light from the surface of the sample to be inspected to form a detection light beam;
Signal processing means for calculating and outputting height information of microprojections on the sample to be inspected based on an output signal output from the optical sensor;
Stage means for placing the sample to be inspected and moving it relative to the illumination optical system and the detection optical system, stage control means for driving and controlling the stage means, and driving and controlling the beam deflection means In a microprojection inspection apparatus having a beam deflection control means,
The detection optical system splits the detection light beam to form first split light and second split light;
A first prism whose reflecting surface is a reflecting surface that divides the first split light into a semicircular first detection light beam and a semicircular second detection light beam, and the first detection light beam A first imaging condensing means for condensing to form a first detection light spot, and a second imaging light collection for condensing the second detection light beam to form a second detection light spot. A light means, a first light sensor for detecting the first detection light spot, a second light sensor for detecting a second detection light spot,
A second prism whose reflecting surface is a reflecting surface that divides the second split light into a semicircular third detection light beam and a semicircular fourth detection light beam; and the third detection light beam. A third imaging condensing means for condensing to form a third detection light spot, and a fourth imaging light collection for condensing the fourth detection light beam to form a fourth detection light spot. A light means; a third light sensor for detecting the third detection light spot; and a fourth light sensor for detecting a fourth detection light spot;
The signal processing means includes height information of a microprojection on the specimen to be inspected from output signals from the first optical sensor, the second optical sensor, the third optical sensor, and the fourth optical sensor. Is calculated and output,
The top side of the first prism and the top side of the second prism are orthogonal to each other,
The detection optical system includes a beam deflecting unit of the illumination optical system, and the beam deflecting unit passes the detection light beam in a direction opposite to the illumination light beam .
Having storage means for storing design information of the specimen to be inspected;
The signal processing means is configured to output first to fourth output signals based on a comparison between inspection position information of the inspection sample obtained from the beam deflection control means and the stage control means and design information of the inspection sample. A microprojection inspection apparatus characterized by selecting any one and calculating and outputting height information .

Images