Inspection apparatus and method of semiconductor element

The semiconductor chip inspection device addresses positioning inaccuracies by using defined reference positions and controlled movements to capture and evaluate images, ensuring accurate and efficient inspection of chip end faces.

JP2025144241AActive Publication Date: 2025-10-02OPTO SYST CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2024043925
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2024-03-19
Publication Date
2025-10-02
Estimated Expiration
2044-03-19

AI Technical Summary

Technical Problem

Existing semiconductor chip inspection methods struggle with inaccurate positioning due to thickness variations and uneven mounting surfaces, leading to inefficient and potentially incomplete inspection of chip end faces.

Method used

The inspection device employs a planar imaging device to position end surface imaging devices at defined reference positions, allowing horizontal and oblique movements to capture multiple images, and utilizes a central control unit to evaluate and determine the optimal image based on data from an evaluation line extending from the chip's top surface to its end faces.

Benefits of technology

This approach ensures accurate and efficient inspection of both chip end faces by defining the second reference position for each chip, enabling quick and reliable determination of inspection quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2025144241000001_ABST
    Figure 2025144241000001_ABST
Patent Text Reader

Abstract

To provide an inspection apparatus allowing more exact and prompt inspection.SOLUTION: The inspection apparatus comprises: a flat face imaging device HR imaging a chip top face 10c of a semiconductor chip; end face imaging devices SLa, SLb imaging chip end surfaces 10a, 10b; and an operation controller CTL positioning the end face imaging devices SLa, SLb at a first reference position P0, then evaluating a photographic image and stopping it at a second reference position Ps, then inclining and moving the end face imaging devices SLa, SLb toward the semiconductor chip 10 while allowing the chip end surfaces 10a, 10b to be repeatedly imaged. The plurality of photographic images is evaluated from a point of view whether a portion connecting the chip top face 10c and the chip end surfaces 10a, 10b can be clearly detected, and then an optimum image is determined.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical Field]

[0001] The present invention relates to an inspection method and an apparatus for quickly and reliably inspecting chip end faces in the manufacturing process of semiconductor chips such as semiconductor laser LDs (Laser Diodes). [Background technology]

[0002] LD chips generally have a double heterostructure in which an active layer that emits light at the laser emission wavelength is sandwiched between a P-type cladding layer and an N-type cladding layer. When a forward voltage is applied between the P-type and N-type layers, both end faces of the active layer function as reflecting mirrors, causing the laser light to travel back and forth within the active layer while being amplified, resulting in stimulated emission.

[0003] Edge-emitting lasers can be broadly classified by their reflection structure into Fabry-Perot lasers, which use semiconductor cleavage planes as reflectors, distributed feedback lasers (DFB lasers), which form a diffraction grating in the waveguide, and distributed Bragg reflector lasers (DBR lasers), which form diffraction gratings before and after the active region.

[0004] For semiconductor chips with such a configuration, it is necessary to inspect the induced emission apertures formed on the chip end faces for scratches or dust, for example, at the final stage of manufacturing. Therefore, the applicant has proposed an invention that enables quick and reliable inspection of the quality of the induced emission apertures (Patent Document 1). [Prior art documents] [Patent documents]

[0005] [Patent Document 1] WO2023012966 (filed August 5, 2021) Summary of the Invention

[0006] In the above invention, the imaging device is tilted away from the semiconductor chip from a somewhat forcibly defined starting position of the tilt movement, and multiple images are acquired during this movement, from which the best focused image is selected.The quality of the stimulated emission aperture is then determined based on the selected best image. [Problem to be solved by the invention]

[0007] However, in an inspection environment where a large number of semiconductor chips are placed on a mounting surface that is not necessarily perfectly flat, the above configuration makes it impossible to move the imaging device to the optimal position, which differs for each semiconductor chip, when the tilt movement begins.

[0008] In other words, since there is some thickness error in the thickness of each semiconductor chip, and since some unevenness (for example, on the order of several hundred microns) is unavoidable in the mounting surface on which the semiconductor chip is attached, the optimal starting point for tilt movement actually differs for each semiconductor chip to be inspected.

[0009] Therefore, the tilt movement of the imaging device may start from an inaccurate movement start position, and in such a case, the end face of the semiconductor chip cannot be accurately determined.

[0010] In particular, even if the imaging device captures a large image of the semiconductor chip when it is closest to the semiconductor chip, as in the case of cited document 1, there is a risk that the semiconductor chip may be overlooked as the imaging device moves upward in the tilted direction. Also, there are cases where it is necessary to determine the quality of both end faces of the semiconductor chip, and in such cases, accuracy and speed of the quality inspection are also desired.

[0011] The present invention has been made in view of the above-mentioned problems, and an object of the present invention is to provide an inspection method and an inspection device that realize more accurate inspection more quickly. [Means for solving the problem]

[0012] The semiconductor element inspection device according to the present invention is configured to include a plane photographing device capable of photographing the chip top surface of a semiconductor chip to be photographed that is placed on a light-reflective or light-transmitting mounting surface; an end surface photographing device capable of photographing the chip end surface of the semiconductor chip while moving back and forth in an oblique direction between a position diagonally above the semiconductor chip and the semiconductor chip; and an operation control device that positions the end surface photographing device at a first reference position based on an image photographed by the plane photographing device, moves the end surface photographing device in a horizontal direction from the first reference position, evaluates the photographed image of the semiconductor chip during movement, stops the end surface photographing device at a second reference position, and then obliquely moves the end surface photographing device toward the semiconductor chip. An inspection device, wherein a first reference position is a position including a second reference position, either beyond or within the first reference position in a planar positional relationship with the semiconductor chip, and the second reference position is a position at which an entire image of the semiconductor chip is captured within the photographing screen of the end face photographing device, and the operation control device is configured to include: a photographing means for photographing a plurality of images while illuminating the chip end face while tilting the end face photographing device from the second reference position; and a determination means for extracting data on an evaluation line extending from the top surface of the chip to the chip end face for the plurality of photographed images photographed by the photographing means, evaluating the change in the data on the evaluation line, and determining the optimal image.

[0013] Furthermore, a semiconductor element inspection method according to the present invention includes a plane imaging device capable of imaging an upper surface of a semiconductor chip to be imaged that is placed on a light-reflective or light-transmitting mounting surface; an end surface imaging device capable of imaging a chip end surface of the semiconductor chip while moving back and forth in an oblique direction between a position obliquely above the semiconductor chip and the semiconductor chip; and an operation control device that executes a first step of positioning the end surface imaging device at a first reference position based on an image captured by the plane imaging device, a second step of moving the end surface imaging device in a horizontal direction from the first reference position, evaluating the image captured of the semiconductor chip during movement, and stopping the end surface imaging device at a second reference position, and a third step of moving the end surface imaging device at an oblique angle toward the semiconductor chip. a first reference position that is a position including a second reference position within or beyond the first reference position in a planar positional relationship with the semiconductor chip, and the second reference position is a position that captures an entire image of the semiconductor chip within an imaging screen of the end face imaging device; and the operation control device is configured to include: an imaging means that captures multiple images while illuminating the chip end face while tilting the end face imaging device from the second reference position; and an identification means that extracts data on an evaluation line extending from the top surface of the chip to the chip end face from the multiple images captured by the imaging means, evaluates the change in the data on the evaluation line, and identifies an optimal image. [Effects of the Invention]

[0014] According to the present invention, the second reference position at which the end surface photographing device starts tilting toward the semiconductor chip is defined for each semiconductor chip as an inspection target, so that the photographing means functions properly and the identifying means can be accurately determined. Also, by adopting the configurations of claims 2 and 3, both side surfaces of the semiconductor chip can be inspected efficiently, quickly, and accurately. [Brief explanation of the drawings]

[0015] [Figure 1] 10 is a diagram illustrating a procedure for determining a first reference position Po and a second reference position Ps. [Figure 2]10 is a diagram illustrating the arrangement of semiconductor chips and the operation of an end surface imaging device. [Figure 3] 1 is a diagram illustrating the configuration of an end surface imaging device and the depth of field. [Figure 4] 10 is a diagram for explaining the maximum value of the photographing pitch Pi. [Figure 5] 10 is a flowchart illustrating an interrupt process that is started every time a motor that realizes tilt movement rotates by a predetermined angle Φ. [Figure 6] 10 is a flowchart illustrating the entire inspection operation. [Figure 7] 10 is a diagram illustrating an embodiment in which the placement surface is a light-transmitting sheet. [Figure 8] 8 is a photographed image in the embodiment of FIG. 7. DETAILED DESCRIPTION OF THE INVENTION

[0016] The present invention will be described in detail below based on the embodiments shown in Figures 1 to 3. In the following embodiments, semiconductor lasers LD (hereinafter referred to as semiconductor chips 10) are used as examples of semiconductor elements, and the stimulated emission apertures at the chip end faces 10a, 10b of a large number of semiconductor chips 10 aligned vertically and horizontally as shown in Figure 2 are inspected.

[0017] As shown in Figure 1, the inspection equipment EQU of the embodiment is configured to include a movable holding device BS that holds semiconductor chips 10...10, a planar imaging device HR that photographs the chip top surface 10c of the semiconductor chip 10 that is the inspection target from directly above, an end surface imaging device SL that moves horizontally from the holding device BS on a horizontal movement line L0 of height VR from a first reference position Po to a second reference position Ps, and when it reaches the second reference position Ps, moves at an inclination on an inclined movement line LN toward the semiconductor chip 10 at a predetermined inclination angle θ, thereby photographing the semiconductor chip 10, and a central control unit CTL that controls the operation of the planar imaging device HR, the end surface imaging device SL, etc., and determines whether the semiconductor chip 10 is good or bad based on the image captured by the end surface imaging device SL.

[0018] Here, a large number of semiconductor chips 10 are attached to the mounting surface 40 of the holding device BS in the aligned arrangement shown in Fig. 2(b) or in the staggered alignment position shown in Fig. 2(c). The holding device BS is configured to be movable so that the semiconductor chip to be inspected is positioned directly below the planar imaging device HR by appropriately moving on a horizontal plane based on the control of the central control device CTL.

[0019] Although not particularly limited, in this embodiment, both longitudinal end faces 10a, 10b of each semiconductor chip 10 are the inspection surfaces. Therefore, in this embodiment, in addition to the first end face imaging device SLa, a second imaging device SLb is provided. Furthermore, the mounting surface 40 is configured to have optical reflectivity, so that the inspection light 20 is reflected by the mounting surface 40, and the reflected light 30 that is re-reflected by the chip end faces 10a, 10b returns to the end face imaging devices SLa, SLb (see FIG. 2(d)).

[0020] These end surface imaging devices SLa and SLb are configured to include an illumination unit that irradiates inspection light 20 toward the semiconductor chip 10 that is the inspection target, an incident-light microscope that receives reflected light 30, a tilting movement unit that tilts and moves the illumination unit and the microscope, and a horizontal movement unit that moves the illumination unit and the microscope horizontally.

[0021] 3(c) is a diagram illustrating the inclined movement unit that realizes reciprocating movement in the inclined direction (Z direction) and the horizontal movement unit that realizes horizontal movement in the horizontal direction (X direction). Here, the illumination unit is provided with a mechanical shutter, and based on the ON / OFF operation of the shutter, it is controlled whether or not the illumination unit in operation actually irradiates the inspection target with inspection light. Note that, instead of the ON / OFF operation of the shutter, the illumination unit operation can be turned ON / OFF; however, this has a slower response speed and is not suitable for high-speed flashing operation.

[0022] As shown in Figures 1 and 3(c), the Z direction of the present invention forms an inclination angle θ with respect to the X direction. The inclination angle θ is determined taking into consideration the arrangement pitch of the semiconductor chips 10 in the illumination direction (W1, W2 shown in Figure 2) and the difference in elevation of the inspection range on the chip end faces 10a, 10b, and is preferably determined in the range of 20° to 40°. In other words, this inclination angle θ is an angle at which the chip end faces 10a, 10b of the inspection target can be photographed without being obstructed by adjacent semiconductor chips 10, and the staggered arrangement of Figure 2(c) has the advantage that the inclination angle θ can be set smaller than the aligned arrangement of Figure 2(b).

[0023] As shown in FIG. 3(c), the end surface imaging devices SLa and SLb are configured to include a stepping motor Mx that realizes reciprocating movement in the X direction, a horizontal slide table 45 that can move reciprocatingly in the X direction based on the rotation of the stepping motor Mx and a first ball screw mechanism, a slide unit 44 that incorporates a stepping motor Mz and a second ball screw mechanism, a slide table 43 that can move reciprocatingly in an inclined direction along the slide unit 44, and an imaging main body 42.

[0024] In this configuration, the stepping motor Mx and the horizontal slide base 45 constitute a horizontal moving part, and the ride unit 44 and the slide base 43 constitute an inclined moving part. Note that Fig. 3(c) merely shows a simplified schematic configuration of the horizontal moving part and the inclined moving part, and the actual end surface imaging device SL does not exactly correspond to the configuration of Fig. 3(c).

[0025] Continuing the explanation based on the above, the central control unit CTL monitors the imaging screen of the planar imaging device HR and moves the holding device BS on a horizontal plane to position the semiconductor chip 10, which is the inspection target on the mounting surface 40, directly below the planar imaging device HR.

[0026] Next, the end face imaging device SL is positioned at a first reference position Po on the horizontal movement line L0 corresponding to the coordinate position of the plane imaging device HR (see FIG. 1). Note that the first end face imaging device SLa is positioned at the first reference position Po corresponding to the chip end face 10a, and the second end face imaging device SLb is positioned at another first reference position Po corresponding to the chip end face 10b.

[0027] Here, the first reference position Po is uniformly defined based on the coordinate position of the plane imaging device HR (the position O directly above the inspection target), and the relative separation distance DST between the plane imaging device HR and the inspection target is determined, for example, as DST = {VR / tan(θ)} + Δ. Here, VR is the difference in elevation between the placement surface 40 and the horizontal movement line L0, and θ is the inclination angle formed by the inclined movement line LN and the horizontal movement line L0. Also, Δ is an appropriate correction value determined according to the thickness and length of the semiconductor chip 10.

[0028] In this embodiment, the correction value Δ is a positive value, and is set so that the first reference position Po is located in an outer region of the second reference position Ps, which is the final stop position on the horizontal movement line L0, with respect to the position O directly above the inspection target. However, the present invention is not limited to this configuration, and the correction value Δ may be a negative value, and the first reference position Po may be set so as to be located in an inner region of the second reference position Ps.

[0029] In either case, the end face imaging devices SLa and SLb, which have been moved to the first reference position Po on the horizontal movement line L0 under the control of the central control device CTL, will capture an image of the semiconductor chip 10 at or outside the upper part of the camera screen based on the above-mentioned positive correction value Δ. Note that FIG. 1(b) shows a state in which the bottom edge of the semiconductor chip 10 is captured at the top of the camera screen. However, since the camera is completely out of focus, the reflected light from the semiconductor chip 10 is actually captured dimly as a "bright captured area."

[0030] Next, the central control device CTL starts the photographing operation of the end surface photographing devices SLa and SLb, and moves them further forward on the horizontal movement line L0 toward the plane photographing device HR while photographing the semiconductor chip 10 that is the inspection target. Then, the "bright capture portion" in the photographed image descends from the state shown in FIG. 1(b) toward the state shown in FIG. 1(b) in response to the forward movement of the end surface photographing device SL.

[0031] Therefore, the central control unit CTL stops the forward movement on the horizontal movement line L0 when the "bright capture unit" reaches the center of the captured image, and sets the stop position as the second reference position Ps. Here, the state in which the "bright capture unit" reaches the center of the captured image means a state in which the "bright capture unit" is surrounded by a "dark background" that is clearly inferior in brightness, as shown in Figure 1(c), and does not take into account the chip end face 10 as in Prior Art 1.

[0032] This is because, since the horizontal movement line L0 is set as an out-of-focus line, it is practically impossible to accurately recognize the chip end face 10 on the horizontal movement line L0 in terms of computer processing.

[0033] Furthermore, in this embodiment, contrary to the operation of Prior Art Document 1, the end surface image capturing devices SLa and SLb are then lowered on the inclined movement line LN, so that the image captured at the second reference position Ps will not be overlooked in the subsequent downward movement as long as the line of sight of the camera and the inclined movement line LN are aligned. Therefore, in this embodiment, the condition for stopping the horizontal movement of the end surface image capturing devices SLa and SLb is that a "dark area (dark background)" of at least 15% or more of the vertical and horizontal size of the camera screen exists around the "bright capturing portion."

[0034] Thereafter, the tilt movement unit functions under the control of the central control unit CTL, and the illumination unit and microscope, which are essential parts of the end face imaging device SL, move downward on the tilt movement line LN. Note that the tilt movement on the tilt movement line LN that descends from the second reference position Ps and the imaging operation on the tilt movement line LN will be described in detail later with reference to Figures 5 and 6.

[0035] Based on the above, first, in relation to the photographing operation, an incident-light microscope, the focal depth, and the depth of field δ will be explained based on Figures 3(a) and 3(b). As shown in Figure 3(a), the incident-light microscope is composed of an objective lens 1 facing the sample 6, an imaging lens 2 that forms an image of the photographing light 30 on the image sensor 7, an illumination lens 3 that receives the illumination light 20, a splitter 5 that directs the illumination light 20 that has passed through the illumination lens 3 to the sample 6, and a light source 4 that generates the illumination light 20.

[0036] The imaging light 30, which is a wave reflected from the chip end faces 10a and 10b (sample 6), passes through the splitter 5 and is guided to the image sensor 7. Furthermore, illumination light 20 is introduced into the imaging main body 42, and passes through the objective lens 1 to illuminate the end face 10a of the semiconductor chip 10. The waves reflected from the chip end faces 10a and 10b are then focused as imaging light on the image sensor 7 of the imaging camera 41.

[0037] Next, the depth of field δ (see Figure 3(b)) means the allowable range of the distance between the sample surface 6 and the lens, in other words, the depth of field δ means the range in which the image is in focus. This depth of field depends on the focal length of the lens and the aperture, but the image of an object that exists within the depth of field will exist within the imaging range (depth of focus) of the image sensor 7, such as a CCD (Charge Coupled Device).

[0038] In this embodiment, to evaluate the quality of the chip end faces 10a and 10b of the semiconductor chip 10, a total magnification of 10 to 20 times is suitable, and correspondingly, the depth of field δ is about δ=10 μm to 2 μm.

[0039] Furthermore, it is preferable that the multiple images to be acquired during the inclined movement of the end face imaging device SL are acquired intermittently every time the end face imaging device SL moves a specified distance (imaging pitch) Pi, and the imaging pitch Pi is determined taking into consideration the height dimension H of the inspection range that includes the judgment area that is the target of the pass / fail judgment. More specifically, as described in Prior Art Document 1, the imaging pitch Pi on the inclined movement line LN is set appropriately under the condition Pi≦δ-H*SIN(θ).

[0040] 4 is a diagram explaining this relationship, with the top surface of the semiconductor chip 10 being the origin line ORG (see FIG. 2(a)), and the range H downward from the origin line ORG being the inspection range H. FIG. 4(b) shows the case where the focal position F1 is located at -X1 on the front side of the origin line ORG, and FIG. 4(c) shows the case where the focal position F2 is located at +X2 on the back side of the origin line ORG.

[0041] Next, the inclined movement and imaging operation on the inclined movement line LN will be described with reference to Fig. 5. Although not particularly limited, in this embodiment, the inclined movement of the end surface imaging device SLa and the end surface imaging device SLb is managed by the rotation angle of each stepping motor Mz that drives each inclined movement unit (slide table 43). Specifically, each time the stepping motor Mz of the end surface imaging device SLa and the stepping motor Mz of the end surface imaging device SLb rotates by a predetermined angle Φ, a pulse signal is output from the rotary encoder, and this pulse signal is transmitted to the central control device CTL as an interrupt signal (encoder interrupt) (see Figs. 5(c) and 5(d)).

[0042] The imaging pitches Pia, Pib of the end surface imaging devices SLa and SLb are set to the same value determined appropriately under the condition Pi≦δ-H*SIN(θ), and this imaging pitch Pi (Pia, Pib) is managed by the number of encoder interrupts N. Here, the number of interrupts N means that the stepping motor Mz has rotated by a rotation angle N*Φ and the end surface imaging devices SLa, SLb have moved by the imaging pitch Pi. Therefore, each time the number of interrupts N corresponding to the imaging pitch Pi is reached, the end surface imaging devices SLa and SLb image the end surfaces 10a, 10b of the semiconductor chip 10.

[0043] 5(c) and 5(d) illustrate this relationship, showing the imaging timings Pia1, Pia2, ... Pian and the imaging timings Pib1, Pib2, ... Pibn, which are defined by the number of interrupts N. Note that, to prevent the imaging timings Pia and Pib of the end surface imaging device SLa and the end surface imaging device SLb from overlapping, the rotation start timings of the stepping motors Mz and Mz are appropriately shifted by approximately N / 2 of the number of interrupts, which corresponds to approximately 1 / 2 of the imaging pitch Pi when converted into distance, as shown in the figure.

[0044] 1(b) to 1(d), in order to improve the accuracy of computer-based judgment, the image capture screen of the end face imaging device SL must clearly show a dark background with poor brightness, and the semiconductor chip 10 must be captured brightly. Therefore, in this embodiment, while one of the end face imaging devices SLa and SLb is performing an imaging operation, the other end face imaging device has its illumination unit turned off, i.e., does not emit inspection light, based on its respective mechanical shutter.

[0045] 5(c) and 5(d), irradiation of the inspection light begins at the encoder interruption one timing before the imaging timings Pia and Pib, and irradiation of the inspection light ends at the encoder interruption one timing after the imaging timings Pia and Pib. As explained above, the rotation start timings of the stepping motors Mz of the end surface imaging devices SLa and SLb are appropriately shifted by about 1 / 2 the imaging pitch Pi, so that when one end surface imaging device is irradiated with inspection light, the other end surface imaging device will not perform an imaging operation.

[0046] As described above, in this embodiment, for any end face imaging device SL, the imaging screen shows the semiconductor chip 10 brightly against a dark background with clearly poor brightness, thereby optimizing the computer-based process for determining the semiconductor chip end face.

[0047] Incidentally, the need to photograph the semiconductor chip 10 brightly against a clearly dark background with poor brightness does not depend on whether the end face of the semiconductor chip 10 is in focus. Therefore, the positioning operation for detecting the second reference position Ps (start position of the tilt movement) on the horizontal movement line L0 is not performed simultaneously by the two end face photographing devices SLa and SLb, but is performed sequentially.

[0048] Next, the inclined movement and photographing operation on the inclined movement line LN will be explained based on a flowchart. Fig. 5(a) shows the encoder interrupt process (SLa interrupt) of the end surface photographing device SLa, and Fig. 5(b) shows the encoder interrupt process (SLb interrupt) of the end surface photographing device SLa. Both processes are started when the stepping motors Mz, Mz rotate by a predetermined angle Φ.

[0049] When any of the encoder interrupt processes is started, it is determined whether or not the irradiation timing has arrived (ST1), and if it has arrived, the irradiation of the inspection light is started (ST2). Next, it is determined whether or not the photographing timing (Piaj / Pibj) has arrived (ST3) for j=1 to n, and if it has not arrived, the interrupt process is ended.

[0050] On the other hand, if it is the photographing timing (Piaj / Pibj), the chip end face 10a / 10b is photographed, the image is stored (ST4), and the irradiation of the inspection light is stopped (ST5). Next, it is determined whether the end face photographing device SL has reached the limit position where the photographing process should end (ST6). If it has not reached the limit position, the interrupt process ends.

[0051] On the other hand, when the end surface photographing device SL reaches the limit position where the photographing process should end, it executes a final process to determine the quality of the end surface of the inspection target based on the acquired series of image data (ST7), and clears the control flags Ia and Ib to 0 to prohibit subsequent encoder interrupts.

[0052] Here, the control flags Ia and Ib are flags corresponding to whether or not an encoder interrupt is permitted, and when the positioning operation for detecting the second reference position Ps (start position of the inclined movement) is completed and the inclined movement on the inclined movement line LN is started, the control flags Ia and Ib are set to 1 in response to the encoder interrupt being permitted (see ST13 in Figure 6).

[0053] 6(a) is a flowchart illustrating the overall inspection operation achieved by positioning movement on the horizontal movement line L0 and inclined movement on the inclined movement line LN. First, the number of interrupts N corresponding to the photographing timing Pi is determined, and characteristic parameters are identified for a reference image of a reference chip that is a typical non-defective product of the semiconductor chip 10 (ST11). Furthermore, for the reference range, which is the pass / fail judgment area of ​​the reference chip, a horizontal range H1 / HMAX to H2 / HMAX from the left end of the reference chip and a vertical range V1 / VMAX to V2 / VMAX from the top end are identified.

[0054] Next, the central control unit CTL moves the holding device BS horizontally to position the semiconductor chip 10, which is the inspection target on the mounting surface 40, directly below the plane imaging device HR, and then positions the end surface imaging device SLa and the end surface imaging device SLb at their respective inclined movement start positions (second reference position Ps) on the horizontal movement line L0 (ST12).

[0055] The positioning operation is performed, for example, in the order of end surface imaging device SLa and end surface imaging device SLb. On the horizontal movement line L0, end surface imaging device SLa is first moved to a first reference position Po and then to a second reference position Ps. Next, the same positioning movement is performed for end surface imaging device SLb. Note that, as shown in FIG. 1, end surface imaging device SLa is positioned at a second reference position Ps corresponding to chip end surface 10a, and end surface imaging device SLb is positioned at another second reference position Ps corresponding to chip end surface 10b.

[0056] As explained above, when the end surface photographing device SLa moves horizontally while photographing the semiconductor chip 10 and is being positioned, no inspection light is emitted from the end surface photographing device SLb, and conversely, when the end surface photographing device SLb moves horizontally while photographing the semiconductor chip 10, no inspection light is emitted from the end surface photographing device SLa.

[0057] When the above positioning operation is completed, the central control device CTL enables the SLa encoder interrupt and the SLb encoder interrupt, sets both control flags Ia and Ib to 1 (ST13), and starts the rotation of the stepping motors Mz, Mz of the end surface imaging device SLa and the end surface imaging device SLb. As shown in Figures 5(c) and 5(d), first, the stepping motor Mz of the end surface imaging device SLa starts to rotate, and after this stepping motor Mz has rotated about 1 / 2 the imaging pitch Pi, the stepping motor Mz of the end surface imaging device SLb starts to rotate.

[0058] By the processing of step ST13, an encoder interrupt is started every time the stepping motor Mz rotates by a predetermined angle Φ, and the operation of Fig. 6(b) or Fig. 6(c) is executed. The processing contents of the SLa encoder interrupt and the SLb encoder interrupt are as shown in Fig. 5(a) and Fig. 5(b) described above.

[0059] Fig. 6(d) is a diagram illustrating the operation of the image acquisition timing in step ST4 of Fig. 5. When the image acquisition timing Pia / Pib is reached, first, the end faces 10a and 10b of the semiconductor chip 10 are photographed and stored (ST41). Next, the area of ​​the semiconductor chip 10 is determined, and an evaluation line extending from the top surface 10c of the semiconductor chip 10 to the end faces 10a and 10b of the semiconductor chip 10 is identified (ST42).

[0060] As shown in Figure 1(d), when the image is in focus, the top surface 10c and the end surfaces 10a / 10b of the semiconductor chip are brightly photographed, and an even brighter band portion, indicated by a double line, is detected between them. Note that for convenience of illustration, the band portion is shown by a double line, but this band portion is the curved portion connecting the top surface 10c and the end surfaces 10a / 10b, and in reality, it is brighter than the flat surfaces 10a / 10b.

[0061] The evaluation line is then defined to detect this bright band. Note that when the image is out of focus, the band cannot be clearly detected. However, even in this case, as shown in Figure 1(c), the "bright capture area" appears in the "dark background," so the evaluation line is the line that crosses the "bright capture area" vertically, including the "dark background."

[0062] Following the processing of step ST42, the brightness data on the evaluation line is subjected to averaging processing such as calculating a moving average value to remove noise (ST43), and the noise-processed data is then differentiated to extract positive and negative edges (ST44).Furthermore, focusing on the positive or negative edges, the positive edges are binarized based on a predetermined threshold value TH, and the pulse width PL of the binarized pulse wave is calculated as a first evaluation value (ST45).

[0063] If the calculated first evaluation value (pulse width PL) is narrower than the specified value, the brightness data on the evaluation line is evaluated from the viewpoint of whether the portion indicated by the double line in Fig. 1(d) (i.e., the band portion that shines brighter than the flat surface 10a / 10b) is clearly detected, and a second evaluation value is calculated (ST45). This second evaluation value is a parameter that indicates whether the band portion indicated by the double line in Fig. 1(d) is clear or not.

[0064] Next, this second evaluation value is compared with the previous evaluation value ANS, and if it is better than the previous evaluation value, the evaluation value ANS is rewritten to the current second evaluation value, and the image number i corresponding to this second evaluation value is stored (ST47).

[0065] As described above, for each image acquisition timing Pia / Pib, the pulse width PL that has the potential to clearly identify the band portion that is the boundary between the chip top surface and the chip end surface is calculated, and the best second evaluation value is sequentially updated to the variable ANS.

[0066] Based on the above, the final processing (ST7) shown in Figures 5(a) and 5(b) will be described. As shown in Figure 6(d), in the final processing, an optimal image is identified based on the image number i corresponding to the second evaluation value ANS stored at that time (ST71). Then, for the identified optimal image i, the background is removed and the necessary feature parameters are calculated.

[0067] Next, the similarity is calculated based on a comparison with the characteristic parameters of the reference image, and the chip end faces 10a, 10b corresponding to the optimal image i are judged to be good or bad (ST72).For semiconductor chips 10 that are judged to be defective, their placement position on the mounting surface, etc. is stored and they are then subjected to the necessary subsequent processing (ST13).

[0068] Incidentally, in the examples described so far, an incident-light microscope has been used, but this is not limitative and a transmission microscope is also suitable. Figure 7 shows an example in which a transmission microscope is used, and shows a holding device BS in which a transparent glass table TL and a transparent film material 40 that serves as a mounting surface 40 are laminated together. A large number of semiconductor chips 10...10 are attached to the transparent film material 40 in an orderly manner with a predetermined arrangement gap W between them.

[0069] In this embodiment, an illumination unit LH is disposed below a transparent glass table TL, and illumination light 20 from the illumination unit LH illuminates the entire back surface of a transparent film material 40. This illumination light 20 is blocked by the back surfaces of the semiconductor chips 10, but passes through the gaps between the semiconductor chips 10 to reach both left and right end surfaces 10a, 10b of the semiconductor chip 10. Then, reflected light 30 from both left and right end surfaces 10a, 10b of the semiconductor chip 10 reaches end surface imaging devices SLa, SLb equipped with transmission microscopes.

[0070] 8(a) and 8(b) show images captured by the end face imaging device SLa. The blue to purple bands indicate the semiconductor chip end face 10a, which is the reflecting surface of the illumination light 20, and above that, the semiconductor chip top face 10c, which blocks the illumination light 20, appears black. The white areas are due to the illumination light that has passed through the gaps between the semiconductor chips 10, 10.

[0071] Figure 8(b) is a captured image in which the arrangement gap W of the semiconductor chips is narrow in the left-right direction of Figure 7. Therefore, the top surface 10c of the semiconductor chip 10 on the left side of the semiconductor chip 10 that is the inspection target in Figure 7 appears black at the bottom of Figure 8(b). Note that not only the arrangement gap W but also the tilt angle θ and microscope magnification are not the same between Figure 8(a) and Figure 8(b), and the thickness of the semiconductor chip end surface 10a is not the same.

[0072] According to this embodiment, not only can the arrangement gap W be narrowed, but the boundary line between the chip top surface 10c and the chip end surfaces 10a / 10b can be clearly grasped, making it easy to determine the pass / fail judgment area and enabling high-precision pass / fail judgment. Furthermore, there is no need to operate the end surface imaging device SLa and the end surface imaging device SLb sequentially, and there is no need to control the illumination light on / off; it is sufficient to leave the device lit at all times. Note that the end surfaces 10a and 10b of the semiconductor chip 10 are imaged at each image acquisition timing Pia / Pib, and the optimal image is selected based on the data on the evaluation line. Other aspects are the same as those of the other embodiments.

[0073] Although the above describes the specific embodiments, the specific configuration can be modified as appropriate without departing from the spirit of the present invention. For example, in the embodiments, the chip end faces are photographed while the end face imaging devices SLa and SLb are moving, but the chip end faces may be photographed while they are temporarily stopped. Furthermore, the method of detecting the minimum pulse width PL is not limited to the sequential operation shown in FIG. 6 . For example, batch processing is also suitable, in which all the required number of images are acquired and saved, and then edge extraction is performed on each image to select the optimal image.

[0074] In the embodiment of Fig. 5, for the sake of convenience, the rotation angle and number of rotations of the stepping motor Mz are determined based on the output pulses of a rotary encoder, but this configuration is not essential. For example, it is of course possible to determine the rotation angle and number of rotations of the stepping motor Mz by counting the drive pulses that rotate the stepping motor Mz without using a rotary encoder. Furthermore, for the sake of convenience, the above embodiment employs the interrupt processing shown in Fig. 5, but this is not limiting. Similar processing can be achieved without using interrupt processing by having the central control device CTL sample the drive pulses of the stepping motors Mz, Mz of the end surface imaging devices SLa, SLb.

[0075] In the embodiment of Figure 8, it goes without saying that the two end surface imaging devices SLa and SLb may start horizontal or tilted movement at any timing, including simultaneous timing, and each may perform imaging operations at any timing. [Explanation of symbols]

[0076] HR planar imaging device SLa,SLb end surface imaging device Po 1st reference position Ps 2nd reference position CTL Motion control device (central control device)

Claims

1. a plane imaging device capable of imaging the upper surface of a semiconductor chip to be imaged, the semiconductor chip being placed on a light-reflective or light-transmitting mounting surface; an end face photographing device capable of photographing the chip end face of the semiconductor chip while reciprocating in an oblique direction between a position obliquely above the semiconductor chip and the semiconductor chip; an operation control device that positions the end surface imaging device at a first reference position based on an image captured by the planar imaging device, moves the end surface imaging device in a horizontal direction from the first reference position, evaluates an image captured of the semiconductor chip during movement, stops the end surface imaging device at a second reference position, and then tilts and moves the end surface imaging device toward the semiconductor chip, the first reference position is a position including a second reference position, located farther or closer to the first reference position in a planar positional relationship with the semiconductor chip, and the second reference position is a position at which an entire image of the semiconductor chip is captured within an imaging screen of the end face imaging device; The operation control device includes: an imaging means for imaging a plurality of images of the chip end face while illuminating the chip end face and tilting the end face imaging device from a second reference position; and a determination means for extracting data on an evaluation line extending from the top surface of the chip to the end surface of the chip from the multiple images taken by the photographing means, evaluating the change in the data on the evaluation line, and determining the optimal image.

2. 2. The inspection device according to claim 1, further comprising an illumination unit for illuminating the underside of the semiconductor chip placed on a light-transmitting mounting surface.

3. the semiconductor chip is configured to include a first end face imaging device capable of capturing a first image including a first chip end face and a top face of the semiconductor chip, and a second end face imaging device capable of capturing a second image including a second chip end face and a top face of the semiconductor chip, The inspection device according to claim 1 , wherein the first end surface imaging device and the second end surface imaging device are configured to perform imaging operations in time sequence or at any timing.

4. The first end surface imaging device and the second end surface imaging device perform imaging operations in time sequence, 4. The inspection device according to claim 3, further comprising an illumination unit that illuminates the first chip end face when illumination onto the second chip end face is stopped, and illuminates the first chip end face when illumination onto the second chip end face is stopped.

5. The first end surface imaging device and the second end surface imaging device perform imaging operations at any timing, 4. The inspection device according to claim 3, further comprising an illumination unit for illuminating the underside of the semiconductor chip placed on a light-transmitting mounting surface.

6. a plane imaging device capable of imaging the upper surface of a semiconductor chip to be imaged, the semiconductor chip being placed on a light-reflective or light-transmitting mounting surface; an end face photographing device capable of photographing the chip end face of the semiconductor chip while reciprocating in an oblique direction between a position obliquely above the semiconductor chip and the semiconductor chip; a first step of positioning the end surface imaging device at a first reference position based on an image captured by the planar imaging device; a second step of moving the end surface photographing device in a horizontal direction from a first reference position, evaluating the photographed image of the semiconductor chip during the movement, and stopping the end surface photographing device at a second reference position; and an operation control device that executes a third step of tilting the end surface imaging device toward the semiconductor chip, the first reference position is a position including a second reference position, located farther or closer to the first reference position in a planar positional relationship with the semiconductor chip, and the second reference position is a position at which an entire image of the semiconductor chip is captured within an imaging screen of the end face imaging device; The operation control device includes: an imaging means for imaging a plurality of images of the chip end face while illuminating the chip end face and tilting the end face imaging device from a second reference position; A semiconductor element inspection method comprising: a specifying means for extracting data on an evaluation line extending from the top surface of the chip to the end surface of the chip from the multiple images taken by the photographing means, evaluating the change in the data on the evaluation line, and specifying the optimal image.

Citation Information

Patent Citations

  • Method and device for determining quality of semiconductor chip

    WO2023012966A1