Semiconductor manufacturing equipment

Abstract

A semiconductor manufacturing device that simply and appropriately adjusts brightness when coaxial illumination and oblique illumination are performed simultaneously is provided. [Solution] The system includes a coaxial illumination unit (52) that irradiates a workpiece (W) with coaxial illumination light (L1) in a first wavelength range, an oblique illumination unit (ring illumination unit 54) that irradiates the workpiece with oblique illumination light (L2) in a second wavelength range different from the first wavelength range, an imaging unit (color camera 56) that separates a mixed light of specularly reflected light (L1A) of the coaxial illumination light that is specularly reflected by the workpiece and scattered light (L2A) of the oblique illumination light that is scattered by the workpiece into specularly reflected light and scattered light and simultaneously captures the mixed light, and an image processing unit (78) that combines a first captured image (bright-field image D1) of the specularly reflected light and a second captured image (dark-field image D2) of the scattered light captured by the imaging unit to generate a composite image (DC).

Description

[Technical Field]

[0001] The present invention relates to a microscope capable of simultaneously performing coaxial illumination and oblique illumination, and to a semiconductor manufacturing apparatus equipped with this microscope. [Background technology]

[0002] A known semiconductor manufacturing device is a dicing device that uses a disk-shaped blade rotated at high speed by a spindle to dice (cut) a workpiece such as a silicon wafer (see Patent Document 1). This dicing device dices the workpiece along the streets (also called planned division lines) formed in a grid pattern on the workpiece by moving the blade that rotates at high speed along the streets.

[0003] In a dicing machine, the blade is aligned with the workpiece prior to dicing the workpiece. When performing alignment, a microscope (photography device) is used to photograph the workpiece pattern (including structures), the position of each street is detected based on edge information of the pattern etc. contained in the photographed image, and the blade is aligned to the street processing start position based on the detection results.

[0004] Known microscopes for such dicing machines include a coaxial illumination unit that provides coaxial illumination (also referred to as incident illumination or coaxial incident illumination) to the workpiece, and a ring illumination unit that illuminates the workpiece from an oblique direction (oblique illumination) (see Patent Document 2). A bright-field image of the workpiece surface is obtained by imaging the specularly reflected light reflected by the workpiece surface by the coaxial illumination with the microscope. Also, a dark-field image of the workpiece surface is obtained by imaging the scattered light scattered by the workpiece surface with the oblique illumination with the microscope.

[0005] The workpiece surface contains both patterns that are easier to recognize in a bright-field image created by coaxial illumination and patterns that are easier to recognize in a dark-field image created by oblique illumination. For this reason, it is desirable to simultaneously use a microscope to perform coaxial illumination and oblique illumination, and simultaneously capture images of the specularly reflected light and scattered light, thereby obtaining an image of the workpiece surface and performing alignment based on this image. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Patent Publication No. 2021-84201 [Patent Document 2] Japanese Patent Publication No. 2020-4862 Summary of the Invention [Problem to be solved by the invention]

[0007] However, with the recent miniaturization and three-dimensionalization of semiconductor devices, it is necessary to recognize edge information of finer patterns on the workpiece from a captured image of the workpiece surface in the above-mentioned alignment. Therefore, before capturing an image of the workpiece surface using the microscope described in Patent Document 2, it is necessary to appropriately adjust the brightness (dimming) of the coaxial illumination unit and ring illumination unit.

[0008] Fig. 18 is an explanatory diagram showing an example of a bright-field image D1 obtained after adjusting the brightness of a coaxial illumination unit using white light. Fig. 19 is an explanatory diagram showing an example of a dark-field image D2 obtained after adjusting the brightness of a ring illumination unit using white light. Fig. 20 is an explanatory diagram for explaining the problem of brightness adjustment when coaxial illumination by a coaxial illumination unit and oblique illumination by a ring illumination unit are performed simultaneously.

[0009] As shown in Figure 18, when only coaxial illumination is used with a microscope, an optimal bright-field image D1 can be obtained by automatically adjusting the brightness using a known method based on a bright-field image captured by the microscope. Also, as shown in Figure 19, when only oblique illumination is used with a microscope, an optimal dark-field image D2 can be obtained by automatically adjusting the brightness using a known method based on a dark-field image captured by the microscope.

[0010] However, if light in the same wavelength range (e.g., white light) is used for both coaxial and oblique illumination, it is impossible to distinguish whether the light captured by the microscope is specularly reflected light (coaxial illumination light) or scattered light (oblique illumination light). For this reason, the brightness of both the coaxial and oblique illumination cannot be adjusted simultaneously.

[0011] Furthermore, even if the brightness of the coaxial illumination and the oblique illumination are adjusted separately, when both coaxial and oblique illuminations are used simultaneously, the combined illumination light from both the coaxial and oblique illuminations results in an overly bright image D3 captured by the microscope, as shown in Figure 20. This makes it impossible to take advantage of the results of the individual brightness adjustments for the coaxial and oblique illuminations. As a result, to capture an optimal image D3, it is necessary to repeatedly adjust the brightness of the coaxial and oblique illuminations while simultaneously using both the coaxial and oblique illuminations. This is extremely time-consuming and may disrupt the balance of the light intensity between the coaxial and oblique illuminations. Furthermore, there is a risk that the pattern in the image D3 may be partially lost due to overexposure or other reasons.

[0012] The present invention has been made in view of the above circumstances, and relates to a microscope and semiconductor manufacturing apparatus that can easily and appropriately adjust brightness when coaxial illumination and oblique illumination are performed simultaneously. [Means for solving the problem]

[0013] A microscope for achieving the object of the present invention is provided in a semiconductor manufacturing device and comprises a coaxial illumination unit that irradiates a workpiece with coaxial illumination light in a first wavelength range, an oblique illumination unit that irradiates the workpiece with oblique illumination light in a second wavelength range different from the first wavelength range, and an imaging unit into which mixed light of specularly reflected light of the coaxial illumination light that is specularly reflected by the workpiece and scattered light of the oblique illumination light that is scattered by the workpiece is incident, and which separates the mixed light into specularly reflected light and scattered light and images them simultaneously.

[0014] According to this microscope, even when coaxial illumination and oblique illumination are performed simultaneously, it is possible to adjust the brightness of both the coaxial illumination unit and the oblique illumination unit.

[0015] In a microscope according to another aspect of the present invention, the imaging unit is a color imaging element including a plurality of pixels arranged two-dimensionally and a plurality of color filters arranged on the plurality of pixels, the plurality of color filters including a plurality of first color filters that transmit light in a first wavelength range and a plurality of second color filters that transmit light in a second wavelength range, thereby making it possible to separate the mixed light into specularly reflected light and scattered light and capture the images simultaneously and coaxially.

[0016] In a microscope according to another aspect of the present invention, the imaging unit includes a light separation element that separates the mixed light into a first light and a second light, a first filter that is disposed on the optical path of the first light separated by the light separation element and that transmits only light in a first wavelength range, a first imaging element that images the specularly reflected light that has passed through the first filter, a second filter that is disposed on the optical path of the second light separated by the light separation element and that transmits only light in a second wavelength range, and a second imaging element that images the scattered light that has passed through the second filter. This allows the mixed light to be separated into the specularly reflected light and the scattered light and imaged simultaneously, thereby obtaining an image with even higher resolution.

[0017] In a microscope according to another aspect of the present invention, the imaging unit includes a wavelength separation element that separates the mixed light into specularly reflected light in a first wavelength range and scattered light in a second wavelength range, a first imaging element that images the specularly reflected light wavelength-separated by the wavelength separation element, and a second imaging element that images the scattered light wavelength-separated by the wavelength separation element. This allows the mixed light to be separated into specularly reflected light and scattered light and imaged simultaneously, resulting in a higher-resolution image. Furthermore, the number of components in the microscope can be reduced, thereby contributing to a smaller and less expensive microscope.

[0018] In a microscope according to another aspect of the present invention, both the coaxial illumination light and the oblique illumination light are visible light.

[0019] In a microscope according to another aspect of the present invention, one of the coaxial illumination light and the oblique illumination light is visible light and the other is infrared light, and the imaging unit captures light from the specularly reflected light and scattered light that corresponds to visible light to output a surface image of the workpiece, and also captures light from the specularly reflected light and scattered light that corresponds to infrared light to output an infrared transmission image of the interior of the workpiece, thereby simultaneously obtaining information about the interior region and surface of the workpiece.

[0020] To achieve the object of the present invention, a semiconductor manufacturing apparatus includes the microscope described above.

[0021] A semiconductor manufacturing apparatus according to another aspect of the present invention includes an image processing unit that generates a composite image by combining a first image captured by the imaging unit using specular reflection light and a second image captured by the imaging unit using scattered light.

[0022] In another aspect of the semiconductor manufacturing apparatus of the present invention, a brightness adjustment unit is provided that adjusts the brightness of the light source of the coaxial illumination unit based on a first captured image of specularly reflected light captured by the imaging unit, and adjusts the brightness of the light source of the oblique illumination unit based on a second captured image of scattered light captured by the imaging unit. [Effects of the Invention]

[0023] The present invention makes it possible to easily and appropriately adjust the brightness when coaxial illumination and oblique illumination are performed simultaneously. [Brief explanation of the drawings]

[0024] [Figure 1] FIG. 1 is a perspective view of a dicing device according to a first embodiment. [Figure 2] FIG. [Figure 3] FIG. 1 is a side view of a microscope according to a first embodiment. [Figure 4] FIG. 2 is an enlarged view of a color filter array provided on the light receiving surface of the color imaging element. [Figure 5] FIG. 10 is an enlarged view showing a modified example of a color filter array. [Figure 6] FIG. 1 is an explanatory diagram showing an example of a bright-field image captured by a color camera. [Figure 7] FIG. 1 is an explanatory diagram showing an example of a dark-field image captured by a color camera. [Figure 8] FIG. 2 is a functional block diagram of a general control unit of the dicing device of the first embodiment. [Figure 9] FIG. 10 is an explanatory diagram showing an example of a method for generating a composite image by an image processing unit. [Figure 10] 3 is a flowchart showing the flow of a dicing process of a workpiece by the dicing device of the first embodiment. [Figure 11] FIG. 10 is a block diagram of a dicing device according to a second embodiment. [Figure 12] FIG. 10 is a side view of a microscope according to a second embodiment. [Figure 13] FIG. 10 is a side view of a modified example of the microscope of the second embodiment. [Figure 14] FIG. 10 is a side view of a microscope of the dicing apparatus according to the third embodiment. [Figure 15] FIG. 10 is an explanatory diagram for explaining a modified example of the light source of the first and second embodiments. [Figure 16] FIG. 10 is a side view showing an example of a microscope of a dicing device that captures images of specularly reflected light and scattered light in a time-division manner. [Figure 17] 10A and 10B are diagrams showing modified examples of the ring illumination unit of the microscope. [Figure 18] FIG. 10 is an explanatory diagram showing an example of a bright-field image obtained after adjusting the brightness of a coaxial illumination unit using white light. [Figure 19] FIG. 10 is an explanatory diagram showing an example of a dark-field image obtained after adjusting the brightness of a ring illumination unit using white light. [Figure 20] 10 is an explanatory diagram for explaining a problem with brightness adjustment when coaxial illumination by a coaxial illumination unit and oblique illumination by a ring illumination unit are performed simultaneously. FIG. DETAILED DESCRIPTION OF THE INVENTION

[0025] [First embodiment] 1 is a perspective view of a dicing apparatus 10 of the first embodiment. Note that the X, Y, and Z axes in the figure are mutually perpendicular, with the X and Y axes being parallel to the horizontal direction and the Z axis being perpendicular to the horizontal direction.

[0026] 1, a dicing apparatus 10 corresponds to the semiconductor manufacturing apparatus (also referred to as a workpiece processing apparatus) of the present invention, and dices a flat workpiece W such as a silicon wafer (semiconductor wafer). The dicing apparatus 10 includes a load port 12, a transfer mechanism 14, a processing unit 16, and a cleaning unit 18.

[0027] A cassette containing a large number of workpieces W mounted on a frame F is placed on the load port 12. A transport mechanism 14 transports the workpieces W. A processing unit 16 dices the workpieces W. A cleaning unit 18 spin-cleans the diced workpieces W. Inside the housing 10A of the dicing device 10 is provided a general control unit 60 (see FIG. 8) that controls the operation of each unit of the dicing device 10. The general control unit 60 may also be provided outside the housing 10A.

[0028] The unmachined workpiece W stored in a cassette placed on the load port 12 is transported by the transport mechanism 14 to the processing unit 16, where it is subjected to dicing such as cutting or grooving to separate it into individual chips. The workpiece W processed by the processing unit 16 is then transported by the transport mechanism 14 to the cleaning unit 18, where it is cleaned, and then transported by the transport mechanism 14 to the load port 12 and stored in a cassette.

[0029] 2 is a perspective view of the appearance of processing unit 16. As shown in FIG. 2 and the above-mentioned FIG. 1, processing unit 16 is the above-mentioned twin-spindle dicer, and includes a pair of blades 21A, 21B, a blade cover (not shown), a pair of spindles 22A, 22B, a microscope 23, and a table 31.

[0030] Blades 21A and 21B are formed in a disk shape. The tip shapes of blades 21A and 21B, i.e., the cross-sectional shapes of the outer peripheries (cutting edges) of blades 21A and 21B along the radial direction of blades 21A and 21B, are rectangular (other shapes such as V-shapes are also acceptable). Blades 21A and 21B are arranged opposite each other in the Y-axis direction and are held by spindles 22A and 22B so as to be rotatable about blade rotation axes parallel to the Y-axis direction.

[0031] The spindles 22A and 22B incorporate high-frequency motors, which rotate the blades 21A and 21B at high speed around the blade rotation axis. As a result, the workpiece W is diced from its front surface (device forming surface) by the blades 21A and 21B. Kerfs (grooves) are formed in the workpiece W by dicing the workpiece W with the blades 21A and 21B.

[0032] The microscope 23 is provided on a Z carriage 44 integrally with, for example, the spindle 22A (or spindle 22B), and is held by the Y carriage 43 and the Z carriage 44 so as to be movable in the Y and Z axes together with the spindle 22A. The microscope 23 photographs the surface of the workpiece W. The microscope 23 is used for aligning the workpiece W with the blades 21A and 21B.

[0033] The table 31 has a workpiece holding surface 31a formed in a porous state, and this workpiece holding surface 31a adsorbs and holds the workpiece W from its back side. The table 31 is held by an X carriage 36 (described later) so as to be movable in the X-axis direction, and is held by a rotation unit 37 (described later) so as to be rotatable about a rotation axis CA.

[0034] Processing section 16 is provided with X base 32, X guide 34, X drive section 35, X carriage 36, and rotation unit 37. X base 32 has a flat plate shape extending in the X-axis direction, and X guide 34 is provided on its upper surface in the Z-axis direction. X guide 34 has a shape that extends in the X-axis direction, and guides X carriage 36 along the X-axis direction. X drive section 35 uses an actuator such as a linear motor, and moves (drives) X carriage 36 in the X-axis direction along X guide 34.

[0035] Rotation unit 37 is provided on the upper surface of X-carriage 36. Table 31 is provided on the upper surface of rotation unit 37. Rotation unit 37 is driven to rotate by rotation drive section 38 (see FIG. 8) which is composed of a motor, gears, etc. As a result, rotation unit 37 rotates table 31 in the θ direction around its rotation axis CA.

[0036] The workpiece W transferred from the load port 12 by the transfer mechanism 14 is held by suction on the table 31, and moves and rotates integrally with the table 31.

[0037] Processing unit 16 is also provided with a Y base 41, a Y guide 42, a pair of Y carriages 43, and a pair of Z carriages 44. Y base 41 has a gate-like shape that straddles X base 32 in the Y-axis direction. Y guide 42 is provided on a side surface of Y base 41 in the X-axis direction. Y guide 42 has a shape that extends in the Y-axis direction, and guides each of the pair of Y carriages 43 along the Y-axis direction. The pair of Y carriages 43 are driven independently along Y guide 42 by Y drive unit 46 (see FIG. 8), which is an actuator constituted by, for example, a stepping motor, a ball screw, etc.

[0038] A Z carriage 44 is provided on each of the pair of Y carriages 43 so as to be movable in the Z-axis direction via a Z drive unit 48 (see FIG. 8) configured with an actuator such as a stepping motor. Spindle 22A and microscope 23 are provided on one side of Z carriage 44, and spindle 22B is provided on the other side of Z carriage 44.

[0039] By driving the X carriage 36, the rotation unit 37, the Y carriages 43, and the Z carriages 44, the blades 21A, 21B and the microscope 23 can be moved in the X, Y, and Z axis directions and the θ direction relative to the table 31 and the workpiece W. This makes it possible to adjust the position of the microscope 23 with respect to the workpiece W before starting alignment, and to align the blades 21A, 21B with respect to the processing start position of the workpiece W after alignment detection. Furthermore, when dicing the workpiece W with the blades 21A, 21B, cutting feed of the workpiece W in the X direction, and index feed of the blades 21A, 21B in the Y axis direction and cutting feed in the Z axis direction can be performed.

[0040] [microscope] 3 is a side view of the microscope 23 of the first embodiment. The microscope 23 is capable of coaxial illumination and oblique illumination of the workpiece W, and is roughly equipped with a coaxial illumination unit 52, a ring illumination unit 54, and a color camera 56.

[0041] The coaxial illumination unit 52 irradiates the surface of the workpiece W with coaxial illumination light L1. The coaxial illumination unit 52 includes a coaxial illumination light source 52a, a half mirror 52b, and an objective lens 52c. The objective lens 52c is disposed in a position facing the surface of the workpiece W and the workpiece holding surface 31a, the half mirror 52b is disposed above the objective lens 52c in the Z direction, and the coaxial illumination light source 52a is disposed to the side of the half mirror 52b.

[0042] The coaxial illumination light source 52a is, for example, an LED (light emitting diode), and emits coaxial illumination light L1 toward the half mirror 52b. As this coaxial illumination light L1, light in a specific wavelength range λA (corresponding to the first wavelength range of the present invention) of visible light, for example, red light, is used.

[0043] The half mirror 52b reflects the coaxial illumination light L1 incident from the coaxial illumination light source 52a toward the objective lens 52c. The half mirror 52b also transmits the specularly reflected light L1A and scattered light L2A (described later) incident from the objective lens 52c, and emits them toward the color camera 56.

[0044] The objective lens 52c irradiates the coaxial illumination light L1 incident from the half mirror 52b onto the surface of the workpiece W. As a result, specularly reflected light L1A of the coaxial illumination light L1 specularly reflected by the surface of the workpiece W passes through the objective lens 52c and the half mirror 52b in this order and enters the color camera 56.

[0045] The ring illumination unit 54 has, for example, a plurality of light sources 54a arranged at equal angular pitches along the circumferential direction of a circle centered on the optical axis O1 of the objective lens 52c so as to surround the objective lens 52c. Each light source 54a irradiates the surface of the workpiece W and the workpiece holding surface 31a with oblique illumination light L2 from an oblique direction. This oblique illumination light L2 is light in a wavelength range λB (corresponding to the second wavelength range of the present invention) that is different from the wavelength range λA of visible light, for example, blue light. When the oblique illumination light L2 is irradiated obliquely onto the surface of the workpiece W, the oblique illumination light L2 is scattered by the surface, and a portion of the scattered light L2A passes through the objective lens 52c and further passes through the half mirror 52b to enter the color camera 56.

[0046] In this embodiment, the coaxial illumination unit 52 irradiates the surface of the workpiece W with coaxial illumination light L1, and the ring illumination unit 54 irradiates the surface of the workpiece W with oblique illumination light L2, simultaneously under the control of a general control unit 60 (see FIG. 8), which will be described later. As a result, mixed light LM of specularly reflected light L1A and scattered light L2A is incident on the color camera 56. Note that, if necessary, it is also possible to selectively irradiate the surface of the workpiece W with coaxial illumination light L1 by the coaxial illumination unit 52 and irradiate the surface of the workpiece W with oblique illumination light L2 by the ring illumination unit 54.

[0047] The color camera 56 separates the mixed light LM that has passed through the half mirror 52b into specularly reflected light L1A and scattered light L2A, and then simultaneously captures the specularly reflected light L1A and scattered light L2A. The color camera 56 includes an imaging lens 56a and a color image sensor 56b.

[0048] The imaging lens 56a forms an image of the mixed light LM incident from the half mirror 52b on the light receiving surface of the color image sensor 56b.

[0049] Fig. 4 is an enlarged view of a color filter array 58 provided on the light receiving surface of the color image sensor 56b. As shown in Fig. 4, the color image sensor 56b is, for example, a two-dimensional image sensor of a CCD (Charge Coupled Device) type or a CMOS (Complementary Metal Oxide Semiconductor) type in which a plurality of pixels 57 (light receiving elements) indicated by dotted lines in the figure are arranged two-dimensionally in the X and Y directions, and is provided with a color filter array 58 on its light receiving surface side.

[0050] 4, the color filter array 58 includes multiple green color filters 58G indicated by "G" in the figure, multiple blue color filters 58B indicated by "B" in the figure, and multiple red color filters 58R indicated by "R" in the figure. The color filters 58R, 58G, and 58B of each color are arranged on each pixel 57 in, for example, a known Bayer array.

[0051] The color filter 58R corresponds to the first color filter of the present invention and transmits only the specularly reflected light L1A, which is red light, of the mixed light LM. The color filter 58B corresponds to the second color filter of the present invention and transmits only the scattered light L2A, which is blue light, of the mixed light LM. The color filter 58G does not transmit the mixed light LM. As a result, the mixed light LM incident on the color filter array 58 is wavelength-separated into the specularly reflected light L1A that passes through each color filter 58R and the scattered light L2A that passes through each color filter 58B. As a result, the specularly reflected light L1A is captured by the multiple pixels 57 corresponding to each color filter 58R of the color image sensor 56b, and the scattered light L2A is captured by the multiple pixels 57 corresponding to each color filter 58B. In other words, the specularly reflected light L1A and the scattered light L2A are simultaneously captured by the color image sensor 56b while being separated from each other.

[0052] 5 is an enlarged view showing a modified example of the color filter array 58. When the color filter array 58 is configured in a Bayer arrangement as shown in FIG. 4 above, imaging is not performed by a plurality of pixels 57 for each color filter 58G of the color imaging element 56b. In this case, the number of pixels 57 that capture the specularly reflected light L1A and the scattered light L2A is reduced, resulting in a decrease in the resolution of a bright-field image D1 and a dark-field image D2 (see FIGS. 6 and 7) described below.

[0053] Therefore, as shown in Figure 5, by configuring the color filter array 58 with only color filters 58R, 58B corresponding to the specularly reflected light L1A (wavelength range λA) and the scattered light L2A (wavelength range λB), the number of pixels 57 that capture the specularly reflected light L1A and the scattered light L2A can be increased, i.e., the resolution of the bright-field image D1 and the dark-field image D2 (see Figures 6 and 7) can be improved.

[0054] FIG. 6 is an explanatory diagram showing an example of a bright-field image D1 captured by the color camera 56. FIG. 7 is an explanatory diagram showing an example of a dark-field image D2 captured by the color camera 56. As shown in FIG. 6, a bright-field image D1 (corresponding to the first captured image of the present invention) of the surface of the workpiece W is obtained by capturing specularly reflected light L1A using only the plurality of pixels 57 for each color filter 58R of the color imaging element 56b. Also, as shown in FIG. 7, a dark-field image D2 (corresponding to the second captured image of the present invention) of the surface of the workpiece W is obtained by capturing scattered light L2A using only the plurality of pixels 57 for each color filter 58B of the color imaging element 56b. Therefore, when coaxial illumination and oblique illumination are performed simultaneously, the bright-field image D1 and the dark-field image D2 can be captured simultaneously and coaxially by the color camera 56.

[0055] [General Control Unit] Fig. 8 is a functional block diagram of the overall control unit 60 of the dicing apparatus 10 of the first embodiment. As shown in Fig. 8, the overall control unit 60 controls each unit of the dicing apparatus 10 overall, and causes the units to perform various operations including brightness adjustment of the coaxial illumination unit 52 and the ring illumination unit 54 (hereinafter simply referred to as brightness adjustment), alignment detection, alignment, and dicing.

[0056] The overall control unit 60 includes an arithmetic circuit configured with various processors, memories, etc. The various processors include a central processing unit (CPU), a graphics processing unit (GPU), an application specific integrated circuit (ASIC), and a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)). The various functions of the overall control unit 60 may be realized by a single processor, or by multiple processors of the same or different types.

[0057] In addition to the spindles 22A, 22B, microscope 23, X drive unit 35, rotation drive unit 38, Y drive unit 46, and Z drive unit 48, an operation unit 62, a memory unit 64, a display unit 66, etc. are connected to the overall control unit 60.

[0058] The operation unit 62 uses a keyboard, mouse, operation panel, operation buttons, etc., and accepts input of various operations by an operator. The memory unit 64 stores a control program (not shown) for the dicing apparatus 10, various setting information, etc. The display unit 66 uses various known monitors, such as a liquid crystal display. This display unit 66 displays images (bright-field image D1 and dark-field image D2) captured by the microscope 23, a composite image DC generated by an image processing unit 78 (described later), various setting screens for the dicing apparatus 10, etc.

[0059] The overall control unit 60 executes a control program (not shown) stored in the storage unit 64, thereby functioning as a blade drive control unit 70, a movement control unit 72, an imaging control unit 74, a brightness adjustment unit 76, an image processing unit 78, an alignment detection unit 80, and a processing control unit 82. Note that what is described as a "unit" of the overall control unit 60 may also be a "circuit," a "device," or a "device." In other words, what is described as a "unit" may be composed of firmware, software, hardware, or a combination of these.

[0060] The blade drive control unit 70 controls the rotational drive of the blades 21A and 21B by the spindles 22A and 22B.

[0061] The movement control unit 72 drives a relative movement mechanism 49 including an X drive unit 35 (X carriage 36), a rotation drive unit 38 (rotation unit 37), a Y drive unit 46 (Y carriage 43), and a Z drive unit 48 (Z carriage 44), thereby moving the blades 21A, 21B, and the microscope 23 relative to the table 31 and the workpiece W.

[0062] For example, before aligning the workpiece W with the blades 21A and 21B, the movement control unit 72 drives the relative movement mechanism 49 to adjust the position of the microscope 23 to a position where the alignment detection pattern formed on the surface of the workpiece W can be photographed.

[0063] Furthermore, when aligning the workpiece W with the blades 21A, 21B (after detecting the alignment), the movement control unit 72 drives the relative movement mechanism 49 to align the blades 21A, 21B with the processing start position of the workpiece W.

[0064] Furthermore, when cutting the workpiece W with the blades 21A and 21B, the movement control unit 72 drives the relative movement mechanism 49 to perform cutting feed of the workpiece W in the X direction, index feed of the blades 21A and 21B in the Y axis direction, and cutting feed of the blades 21A and 21B in the Z axis direction.

[0065] The photography control unit 74 controls photography of the workpiece W by the microscope 23. After adjusting the position of the microscope 23 as described above, the photography control unit 74 starts coaxial illumination by the coaxial illumination unit 52 and oblique illumination by the ring illumination unit 54, and also causes the color camera 56 to photograph the workpiece W, i.e., to photograph a bright-field image D1 and a dark-field image D2. Here, the photography of the bright-field image D1 and the dark-field image D2 may be performed at least before adjusting the brightness of the coaxial illumination unit 52 and the ring illumination unit 54 of the microscope 23, and after adjusting the brightness and before alignment detection, or may be performed continuously. The color camera 56 outputs the bright-field image D1 and the dark-field image D2 photographed before the brightness adjustment to the brightness adjustment unit 76 described below, and outputs the bright-field image D1 and the dark-field image D2 photographed before alignment detection to the image processing unit 78 described below.

[0066] The brightness adjustment unit 76 operates after the position adjustment of the microscope 23 described above. When a bright-field image D1 and a dark-field image D2 are input from the color camera 56, the brightness adjustment unit 76 adjusts the brightness of the coaxial illumination unit 52 (coaxial illumination light source 52a) based on the bright-field image D1 and adjusts the brightness of the ring illumination unit 54 (each light source 54a) based on the dark-field image D2.

[0067] For example, based on a bright-field image D1 obtained by capturing only the specularly reflected light L1A, the brightness adjustment unit 76 adjusts the amount of coaxial illumination light L1 emitted from the coaxial illumination light source 52a so as to optimize (maximize) the contrast of this bright-field image D1. Also, based on a dark-field image D2 obtained by capturing only the scattered light L2A, the brightness adjustment unit 76 adjusts the amount of oblique illumination light L2 emitted from each light source 54a so as to optimize the contrast of this dark-field image D2. In this case, the brightness adjustment of the coaxial illumination unit 52 and the brightness adjustment of the ring illumination unit 54 can each be performed by a known method, and further, both brightness adjustments can be performed simultaneously (or in a time-division manner).

[0068] In this way, the brightness adjustment unit 76 adjusts the brightness of both the coaxial illumination unit 52 and the ring illumination unit 54, so that the color camera 56 can capture a bright-field image D1 and a dark-field image D2 with high contrast, i.e., a high-precision recognizable pattern on the surface of the workpiece W. The bright-field image D1 and the dark-field image D2 captured by the color camera 56 after brightness adjustment are input from the color camera 56 to the image processing unit 78.

[0069] Fig. 9 is an explanatory diagram showing an example of a method for generating a composite image DC by the image processing unit 78. As shown in Fig. 9 and the above-described Fig. 8, the image processing unit 78 generates a composite image DC by combining a bright-field image D1 (see Fig. 6) and a dark-field image D2 (see Fig. 7) input from the color camera 56 after brightness adjustment by the brightness adjustment unit 76.

[0070] For example, the image processing unit 78 first divides each of the bright-field image D1 and the dark-field image D2 into a plurality of regions as shown by reference numerals 9A and 9B in Fig. 9, and calculates a contrast evaluation value for each divided region (see International Publication No. 2013 / 047482). Note that in Fig. 9, the number of divided regions is shown to be smaller than the actual number in order to avoid cluttering the drawing.

[0071] Next, the image processing unit 78 compares the contrast evaluation value of the bright-field image D1 with the contrast evaluation value of the dark-field image D2 for each divided region, and selects and stitches together the divided region with the higher contrast evaluation value to generate a composite image DC as shown by reference numeral 9C in FIG. 9. This results in a composite image DC in which high-contrast regions are selected and stitched together from the bright-field image D1 and the dark-field image D2. As a result, a composite image DC is obtained that includes both patterns that are easier to recognize in the bright-field image D1 obtained by coaxial illumination and patterns that are easier to recognize in the dark-field image D2 obtained by oblique illumination. The image processing unit 78 then outputs the composite image DC to the alignment detection unit 80.

[0072] The method for generating the composite image DC by the image processing unit 78 is not particularly limited, and one of the bright-field image D1 and the dark-field image D2 may simply be superimposed on the other. Alternatively, after extracting only luminance information from the bright-field image D1 and the dark-field image D2, a monochrome composite image DC may be generated by the method shown in Fig. 9, or one of the monochrome bright-field image D1 and the dark-field image D2 may be superimposed on the other.

[0073] Furthermore, in this embodiment, as described above, the color camera 56 captures the bright-field image D1 and the dark-field image D2 coaxially, so that the image processing unit 78 can omit positional deviation correction when combining the bright-field image D1 and the dark-field image D2.

[0074] 8, based on the composite image DC input from the image processing unit 78, the alignment detection unit 80 performs alignment detection to detect the position of each street (not shown) of the workpiece W by detecting edge information and the like of the alignment detection pattern of the workpiece W included in this composite image DC using a known image recognition method. Then, the alignment detection unit 80 outputs the alignment detection result to the movement control unit 72. As a result, the movement control unit 72 drives the relative movement mechanism 49 to perform alignment between the streets (not shown) of the processing target and the blades 21A, 21B.

[0075] After the alignment is complete, the processing control unit 82 drives the spindles 22A, 22B and the relative movement mechanism 49 via the blade drive control unit 70 and the movement control unit 72 to perform dicing processing with the blades 21A, 21B for each street (not shown) of the workpiece W. Note that since the dicing device 10 of this embodiment is a so-called twin-spindle dicer, the processing control unit 82 selectively executes, for example, a meeting cutting method or a step cutting method for dicing the workpiece W (see Patent Document 1 above).

[0076] [Operation of the first embodiment] 10 is a flowchart showing the flow of dicing processing of the workpiece W by the dicing apparatus 10 of the first embodiment configured as described above. As shown in Fig. 10, when the workpiece W is sucked and held on the table 31, each part of the integrated control unit 60 operates. Then, the movement control unit 72 drives the relative movement mechanism 49 to adjust the position of the microscope 23 to a position where the alignment detection pattern on the surface of the workpiece W can be photographed.

[0077] When the position adjustment of the microscope 23 is completed, the photography control unit 74 starts the emission of coaxial illumination light L1 from the coaxial illumination light source 52a and the emission of oblique illumination light L2 from each light source 54a, and also starts imaging by the color image sensor 56b. As a result, the coaxial illumination unit 52 simultaneously emits the coaxial illumination light L1 onto the workpiece W (coaxial illumination), and the ring illumination unit 54 simultaneously emits the oblique illumination light L2 onto the workpiece W (oblique illumination) (step S1).

[0078] When coaxial illumination and oblique illumination are started simultaneously, a mixed light LM consisting of specularly reflected light L1A of the coaxial illumination light L1 that is specularly reflected by the surface of the workpiece W and a portion of scattered light L2A of the oblique illumination light L2 that is scattered by the surface of the workpiece W passes through the objective lens 52c and the half mirror 52b and enters the color camera 56, and then passes through the imaging lens 56a of this color camera 56 and enters the color filter array 58 of the color image sensor 56b.

[0079] The mixed light LM incident on the color filter array 58 is wavelength-separated into specularly reflected light L1A that passes through each color filter 58R and scattered light L2A that passes through each color filter 58B. The specularly reflected light L1A is then captured by the plurality of pixels 57 corresponding to each color filter 58R of the color image sensor 56b. At the same time, the scattered light L2A is captured by the plurality of pixels 57 corresponding to each color filter 58B of the color image sensor 56b. As a result, the specularly reflected light L1A and the scattered light L2A are simultaneously captured by the color image sensor 56b in a mutually separated state (step S2). As a result, a bright-field image D1 capturing the specularly reflected light L1A and a dark-field image D2 capturing the scattered light L2A are simultaneously output from the color camera 56 to the brightness adjustment unit 76.

[0080] In this embodiment, by making the wavelength range λA of the coaxial illumination light L1 (specularly reflected light L1A) and the wavelength range λB of the oblique illumination light L2 (scattered light L2A) different from each other, the mixed light LM can be easily wavelength-separated into the specularly reflected light L1A and the scattered light L2A in the color camera 56 and imaged individually.

[0081] Next, when the brightness adjustment unit 76 acquires the bright-field image D1 and the dark-field image D2 from the color camera 56 (step S3), it simultaneously adjusts the brightness of the coaxial illumination unit 52 (coaxial illumination light source 52a) based on the bright-field image D1 and the brightness of the ring illumination unit 54 (each light source 54a) based on the dark-field image D2 (step S4). In this case, it is sufficient to adjust the brightness of the coaxial illumination unit 52 based on the bright-field image D1 and the brightness of the ring illumination unit 54 based on the dark-field image D2 individually using known methods, so that the brightness adjustments of both the coaxial illumination unit 52 and the ring illumination unit 54 can be easily performed.

[0082] Furthermore, in this embodiment, since the wavelength ranges λA and λB do not overlap, the problem of the captured image D3 being too bright under illumination by both the coaxial illumination unit 52 and the ring illumination unit 54 as shown in FIG. 20 does not occur, and a bright-field image D1 and a dark-field image D2 with brightness levels that can be used for brightness adjustment are obtained. Therefore, in this embodiment, it is sufficient to adjust the brightness of both the coaxial illumination unit 52 and the ring illumination unit 54 individually, eliminating the need to repeatedly adjust the brightness of the coaxial illumination unit 52 and the ring illumination unit 54 while simultaneously performing coaxial illumination and oblique illumination, as was done in the conventional method. As a result, brightness adjustment of both the coaxial illumination unit 52 and the ring illumination unit 54 can be performed easily and quickly.

[0083] When the brightness adjustments of the coaxial illumination unit 52 and the ring illumination unit 54 are completed, the photography control unit 74 causes the color image sensor 56b to start imaging. As a result, similar to step S2 described above, the color image sensor 56b simultaneously captures the specularly reflected light L1A and the scattered light L2A while separating them from each other (step S5). Then, the color camera 56 simultaneously outputs to the image processing unit 78 a bright-field image D1 capturing the specularly reflected light L1A and a dark-field image D2 capturing the scattered light L2A.

[0084] Next, when the image processing unit 78 acquires the bright-field image D1 and the dark-field image D2 from the color camera 56 (step S6), it generates a composite image DC by selecting and stitching together high-contrast regions from the bright-field image D1 and the dark-field image D2, as described above with reference to FIG. 9 (step S7). This results in a composite image DC that includes both patterns that are easier to recognize in the bright-field image D1 obtained by coaxial illumination and patterns that are easier to recognize in the dark-field image D2 obtained by oblique illumination. In this embodiment, because the bright-field image D1 and the dark-field image D2 are captured coaxially by the color camera 56, it is possible to omit positional deviation correction when combining the bright-field image D1 and the dark-field image D2. As a result, the composite image DC can be generated simply and quickly.

[0085] Furthermore, in this embodiment, by differentiating the wavelength range λA of the coaxial illumination light L1 from the wavelength range λB of the oblique illumination light L2 and capturing images by separating the wavelengths of the specularly reflected light L1A and the scattered light L2A, it is possible to prevent the image from becoming too bright and losing the pattern on the surface of the workpiece W, as in the conventional captured image D3, and to obtain a composite image DC that includes more patterns (edge ​​information). This composite image DC is output from the image processing unit 78 to the alignment detection unit 80.

[0086] Then, the alignment detection unit 80 performs alignment detection to detect the position of each street (not shown) of the work W using a known method based on the composite image DC input from the image processing unit 78, and outputs the alignment detection results to the movement control unit 72 (step S8).

[0087] Next, based on the alignment detection result by the alignment detection unit 80, the movement control unit 72 drives the relative movement mechanism 49 to align the street (not shown) of the processing target with the blades 21A and 21B (step S9). Based on the above-mentioned composite image DC, alignment detection can be performed using more patterns (edge ​​information), or using patterns with less information that would have been lost in the past. As a result, the accuracy of alignment detection and alignment can be improved.

[0088] Once alignment is complete, the processing control unit 82 drives the spindles 22A, 22B and the relative movement mechanism 49 via the blade drive control unit 70 and the movement control unit 72 to perform dicing processing using the blades 21A, 21B for each street (not shown) of the workpiece W (step S10).

[0089] As described above, in the first embodiment, by making the wavelength range λA of the coaxial illumination light L1 and the wavelength range λB of the oblique illumination light L2 different from each other, it is possible to easily separate the specular reflected light L1A and the scattered light L2A in terms of wavelength and capture the images individually using the color camera 56. As a result, even when coaxial illumination and oblique illumination are performed simultaneously, the brightness of the coaxial illumination unit 52 and the ring illumination unit 54 can be easily and appropriately adjusted.

[0090] [Second embodiment] Fig. 11 is a block diagram of a dicing apparatus 10 according to the second embodiment. Fig. 12 is a side view of a microscope 23 according to the second embodiment. In the first embodiment, the specularly reflected light L1A and the scattered light L2A are wavelength-separated and imaged using the color image sensor 56b of the color camera 56, but in the second embodiment, the specularly reflected light L1A and the scattered light L2A are wavelength-separated and imaged without using the color image sensor 56b.

[0091] 11 and 12, the dicing apparatus 10 of the second embodiment has basically the same configuration as the first embodiment, except that the microscope 23 is provided with a half mirror 52d, a transmission filter 52e, and a transmission filter 52f, and the microscope 23 is provided with a first camera 59A and a second camera 59B instead of the color camera 56. Therefore, components that are the same in function or configuration as those of the first embodiment are given the same reference numerals, and their description will be omitted. In the second embodiment, the half mirror 52d, the transmission filter 52e, the transmission filter 52f, the first camera 59A, and the second camera 59B constitute the imaging unit of the present invention.

[0092] The half mirror 52d corresponds to the light separating element of the present invention, and is disposed above the half mirror 52b in the Z direction. The half mirror 52d transmits a portion of the mixed light LM (corresponding to the first light of the present invention) incident from the half mirror 52b upward in the Z direction, and reflects the remainder of the mixed light LM (corresponding to the second light of the present invention) sideways, in the X direction here. Note that various light separating elements other than the half mirror 52d may be used as long as they can split the mixed light LM into two.

[0093] The transmission filter 52e (corresponding to the first filter of the present invention) is disposed above the half mirror 52b in the Z direction and on the optical path of the mixed light LM, and is a bandpass filter that transmits only light in the wavelength range λ A. As a result, when the mixed light LM is incident on the transmission filter 52e from the half mirror 52d, only the specularly reflected light L1A, which is light in the wavelength range λ A, transmits it toward a first camera 59A, which will be described later.

[0094] The transmission filter 52f (corresponding to the second filter of the present invention) is disposed to the side of the half mirror 52b (the reflection direction side of the mixed light LM) and on the optical path of the mixed light LM, and is a bandpass filter that transmits only light in the wavelength range λ B. As a result, when the mixed light LM is incident on the transmission filter 52f from the half mirror 52d, only the scattered light L2A, which is light in the wavelength range λ B, transmits it toward a second camera 59B, which will be described later.

[0095] The first camera 59A is disposed above the transmission filter 52e in the Z direction. The first camera 59A is a so-called monochrome camera, and includes a first imaging lens 90a and a first imaging element 92a that does not have a color filter array 58.

[0096] The first imaging lens 90a forms an image of the specularly reflected light L1A that has passed through the transmission filter 52e on the light-receiving surface of the first image sensor 92a. As a result, the first image sensor 92a captures only the specularly reflected light L1A and outputs a bright-field image D1 to the integrated controller 60. At this time, unlike the color image sensor 56b of the first embodiment, the first image sensor 92a captures the specularly reflected light L1A using all of the pixels 57 within the light-receiving surface (within the effective area), and therefore a bright-field image D1 with higher resolution than that of the first embodiment can be obtained.

[0097] The second camera 59B is disposed opposite the exit surface of the transmission filter 52f and includes a second imaging lens 90b and a second image sensor 92b that does not have a color filter array 58. The second imaging lens 90b forms an image of the scattered light L2A that has passed through the transmission filter 52f on the light receiving surface of the second image sensor 92b. This causes the second image sensor 92b to capture only the scattered light L2A and output a dark-field image D2 to the integrated controller 60. At this time, the second image sensor 92b can also capture the scattered light L2A using all pixels 57 within its light receiving surface (effective area), so that a dark-field image D2 with higher resolution than that of the first embodiment can be obtained.

[0098] The imaging control unit 74 of the second embodiment simultaneously captures the specularly reflected light L1A using the first camera 59A and the scattered light L2A using the second camera 59B at least before adjusting the brightness of the coaxial illumination unit 52 and the ring illumination unit 54 and before alignment detection. As a result, before brightness adjustment, a bright-field image D1 and a dark-field image D2 are input from the first camera 59A and the second camera 59B to the brightness adjustment unit 76. As a result, similar to the first embodiment, the brightness adjustment of the coaxial illumination unit 52 and the ring illumination unit 54 is performed by the brightness adjustment unit 76.

[0099] Furthermore, before alignment detection, the bright-field image D1 and the dark-field image D2 are input from the first camera 59A and the second camera 59B to the image processing unit 78. As a result, similar to the first embodiment, the image processing unit 78 generates a composite image DC and the alignment detection unit 80 performs alignment detection.

[0100] In this way, in the second embodiment, by combining the half mirror 52d, the transmission filters 52e and 52f, the first camera 59A, and the second camera 59B, it is possible to easily separate the specularly reflected light L1A and the scattered light L2A by wavelength and simultaneously capture images of them, as in the first embodiment. As a result, the same effects as in the first embodiment can be obtained. Furthermore, by using the first camera 59A and the second camera 59B (monochrome cameras), it is possible to obtain a bright-field image D1 and a dark-field image D2 with higher resolution than in the first embodiment, thereby further improving the accuracy of alignment detection.

[0101] [Modification of the second embodiment] Fig. 13 is a side view of a modified example of the microscope 23 of the second embodiment. In the second embodiment, the mixed light LM is wavelength-separated into specularly reflected light L1A and scattered light L2A using the half mirror 52d and transmission filters 52e and 52f, but the wavelength separation may also be performed using a dichroic mirror 52g (corresponding to the wavelength separation element of the present invention) as shown in Fig. 13. In this case, the dichroic mirror 52g, the first camera 59A, and the second camera 59B constitute the imaging unit of the present invention.

[0102] Note that, as long as the mixed light LM can be wavelength-separated into the specularly reflected light L1A and the scattered light L2A, various wavelength separation elements other than the dichroic mirror 52g may be used.

[0103] As described above, in the modified example of the second embodiment, by using the dichroic mirror 52g to separate the mixed light LM into specularly reflected light L1A and scattered light L2A, the number of parts of the microscope 23 can be reduced compared to the second embodiment, thereby making the microscope 23 smaller and less expensive.

[0104] [Third embodiment] 14 is a side view of the microscope 23 of the dicing apparatus 10 of the third embodiment. In the above-described embodiments, visible light is used as the coaxial illumination light L1 and the oblique illumination light L2, but in the third embodiment, infrared light (including near-infrared light) is used as either the coaxial illumination light L1 or the oblique illumination light L2. The dicing apparatus 10 of the third embodiment has basically the same configuration as the first embodiment, except that the coaxial illumination light source 52a of the microscope 23 emits coaxial illumination light L1 in the infrared wavelength range λIR. Therefore, components that are the same in function or configuration as those of the first embodiment are designated by the same reference numerals, and their description will be omitted.

[0105] As shown in Fig. 14, in the third embodiment, by using coaxial illumination light L1 (infrared light) for coaxial illumination, when the color camera 56 captures the specularly reflected light L1A, a bright-field image D1, which is an infrared transmission image of the internal region AR of the workpiece W, is obtained. On the other hand, the dark-field image D2 is a surface image of the workpiece W, as in the first embodiment. Therefore, in the third embodiment, by simultaneously performing coaxial illumination and oblique illumination, information on the internal region AR of the workpiece W and information on the surface can be obtained simultaneously. Note that, because the focal positions of the color camera 56 differ between the surface and the internal region AR of the workpiece W, it is possible to photograph twice by changing the focal position of the color camera 56, or to photograph using a color camera 56 with a deep depth of field.

[0106] Furthermore, in the third embodiment, by using coaxial illumination light L1 (infrared light) for the coaxial illumination, even if a pattern or the like on the surface of the workpiece W is covered with resin (which transmits infrared light), the position of this pattern can be detected based on the bright-field image D1. Therefore, in the third embodiment, the position of the kerf formed by dicing and the position of the pattern covered with resin can be detected simultaneously using the microscope 23, and the formation position of the kerf (amount of deviation) can be corrected.

[0107] Furthermore, in the third embodiment, information on the internal region AR of the workpiece W and information on the surface can be obtained simultaneously, so even if the resin is thin, information on the resin surface can be obtained based on the dark-field image D2. That is, when only infrared light is used for both coaxial illumination and oblique illumination, if the resin is thin, the internal pattern becomes visible, making it difficult to observe the resin surface, but in the third embodiment, visible light is used for oblique illumination, making it possible to observe the resin surface.

[0108] In the third embodiment, the coaxial illumination is performed with infrared light and the oblique illumination is performed with visible light, but the coaxial illumination may be performed with visible light and the oblique illumination may be performed with infrared light. Also, in the second embodiment, either the coaxial illumination light L1 or the oblique illumination light L2 may be infrared light.

[0109] [Variations of light source] 15 is an explanatory diagram for explaining a modified example of the light source of the first and second embodiments. In the first and second embodiments, the coaxial illumination unit 52 and the ring illumination unit 54 each have an independent light source (coaxial illumination light source 52a, each light source 54a), but the coaxial illumination unit 52 and the ring illumination unit 54 may share a light source, that is, a common light source 100 may be provided in the microscope 23.

[0110] As shown in FIG. 15, the common light source 100 includes a white LED 101, a half mirror 102, a transmission filter 104, a mirror 106, and a transmission filter 108.

[0111] The white LED 101 emits white light L0 toward the half mirror 102. The half mirror 102 splits the white light L0 into two, and emits one of the two toward a transmission filter 104 and the other toward a mirror 106. The mirror 106 emits the white light L0 incident from the half mirror 102 toward a transmission filter 108.

[0112] The transmission filter 104 is basically the same as the transmission filter 52e of the second embodiment, and transmits light in the wavelength range λA. As a result, when white light L0 is incident on the transmission filter 104 from the half mirror 102, the transmission filter 104 transmits only the coaxial illumination light L1 in the wavelength range λA and supplies it to the coaxial illumination unit 52. This allows the coaxial illumination light L1 to be emitted from the coaxial illumination unit 52.

[0113] The transmission filter 108 is basically the same as the transmission filter 52f in the second embodiment, and transmits light in the wavelength range λB. As a result, when white light L0 is incident on the transmission filter 108 from the mirror 106, the transmission filter 108 transmits only oblique illumination light L2 in the wavelength range λB and supplies it to the ring illumination unit 54. This allows the ring illumination unit 54 to emit oblique illumination light L2. By using the common light source 100 in this manner, light (including visible light, infrared light, and ultraviolet light) of two specific wavelengths (wavelength ranges λA and λB) can be extracted from the white light L0. Note that a halogen lamp that emits white light L0 may be used instead of the white LED 101.

[0114] [Other Variations of Microscopes] 16 is a side view showing an example of a microscope 110 of the dicing apparatus 10 that captures images of the specularly reflected light L1A and the scattered light L2A in a time-division manner. The microscope 23 in each of the above embodiments simultaneously captures images of the specularly reflected light L1A and the scattered light L2A, but the specularly reflected light L1A and the scattered light L2A may also be captured in a time-division manner, as in the microscope 110 shown in FIG.

[0115] The microscope 110 has basically the same configuration as the microscope 23 of the second embodiment (see FIG. 12) except that it includes a filter switching mechanism 111 instead of the half mirror 52d, the transmission filters 52e and 52f, and the second camera 59B. Therefore, components that are the same in function or configuration as those of the second embodiment are given the same reference numerals, and descriptions thereof will be omitted.

[0116] The filter switching mechanism 111 is disposed between the half mirror 52b and the first camera 59A. Under the control of the integrated control unit 60, the filter switching mechanism 111 selectively places a transmission filter 112A or a transmission filter 112B on the optical path of the mixed light LM.

[0117] The transmission filter 112A is basically the same as the transmission filter 52e of the second embodiment, and transmits light in the wavelength range λA. As a result, when the transmission filter 112A is disposed on the optical path of the mixed light LM, only the specularly reflected light L1A enters the first camera 59A. As a result, a bright-field image D1 is output from the first camera 59A to the integrated controller 60.

[0118] The transmission filter 112B is basically the same as the transmission filter 52f of the second embodiment and transmits light in the wavelength range λB. As a result, when the transmission filter 112B is disposed on the optical path of the mixed light LM, only the scattered light L2A enters the first camera 59A. As a result, a dark-field image D2 is output from the first camera 59A to the integrated controller 60.

[0119] As described above, the first camera 59A can capture the specularly reflected light L1A and the scattered light L2A in a time-division manner by driving the filter switching mechanism 111 to sequentially arrange the transmission filter 112A and the transmission filter 112B on the optical path of the mixed light LM. In this case, too, brightness adjustment, alignment detection, and the like can be performed based on the bright-field image D1 and the dark-field image D2 captured by the first camera 59A, as in the above-described embodiments.

[0120] In addition, when the color image sensor 56b of the microscope 23 of the first embodiment is a CMOS type, by controlling the color image sensor 56b, it is possible to perform time-division imaging of the regular reflection light L1A by the multiple pixels 57 for each color filter 58R and imaging of the scattered light L2A by the multiple pixels 57 for each color filter 58B.

[0121] [others] In the first and second embodiments described above, red light is used as the coaxial illumination light L1 and blue light (with a short wavelength) that is easily scattered on the surface of the workpiece W is used as the oblique illumination light L2. However, blue light may be used as the coaxial illumination light L1 and red light as the oblique illumination light L2. Furthermore, the wavelength range λA of the coaxial illumination light L1 and the wavelength range λB of the oblique illumination light L2 are not particularly limited as long as they are different from each other. For example, ultraviolet light may be used as one of the coaxial illumination light L1 and the oblique illumination light L2 and visible light may be used as the other, or ultraviolet light may be used as one and infrared light as the other. Ultraviolet light is particularly effective as the oblique illumination light L2 because of its high scattering intensity.

[0122] The wavelength ranges λA and λB are not limited to single wavelengths and may be central wavelengths of broad spectra. In this case, for example, the entire infrared light range may be used as the coaxial illumination light L1, and the entire visible light range from blue light to red light may be used as the oblique illumination light L2.

[0123] Furthermore, the wavelength ranges λA and λB do not have to be continuous wavelength ranges as long as they are different from each other. For example, if a green wavelength range is used as the coaxial illumination light L1 in the wavelength range λA, light in the red wavelength range and light in the blue wavelength range may be used as the oblique illumination light L2 in the wavelength range λB.

[0124] Furthermore, light of three or more different wavelength bands may be used as the coaxial illumination light L1 and the oblique illumination light L2. In this case, for example, as shown in Fig. 17, a ring illumination unit 54-1 (each light source 54a) may be added to the microscope 23, which irradiates the workpiece holding surface 31a with oblique illumination light L3 of the wavelength band λC from an oblique direction at an incident angle different from that of the ring illumination unit 54. Furthermore, a plurality of ring illumination units may be added, each with different incident angles of oblique illumination light with respect to the workpiece holding surface 31a and different wavelength bands.

[0125] In each of the above embodiments, the workpiece W is illuminated obliquely by the ring illumination unit 54, but various types of oblique illumination units capable of providing oblique illumination may be used instead of the ring illumination unit 54.

[0126] In each of the above embodiments, the dicing device 10 and its microscope 23 that perform dicing processing of the workpiece W using blades 21A and 21B have been used as examples, but the present invention can also be applied to a dicing device 10 and its microscope 23 that perform dicing processing of a wafer using laser light.

[0127] In the above embodiments, the dicing apparatus 10 and its microscope 23 have been described as examples, but the present invention can be applied to various semiconductor manufacturing apparatuses and microscopes thereof, such as a laser processing apparatus that forms a laser processing area inside a wafer, a diagnostic apparatus that diagnoses the blades 21A and 21B, or a prober that inspects a plurality of chips formed on a workpiece W. Furthermore, the present invention can also be applied to microscopes used for purposes other than semiconductor manufacturing apparatuses, which simultaneously perform coaxial illumination and oblique illumination on various objects to be observed. [Explanation of symbols]

[0128] 10 Dicing equipment 10A housing 12 Loading Port 14 Transport mechanism 16 Processing Department 18 Cleaning section 21A, 21B blades 22A, 22B Spindle 23 Microscope 31 Table 31a Work holding surface 32 X Base 34 X Guide 35 X drive unit 36 x Carriage 37 Rotating Unit 38 Rotation drive unit 41 Y Base 42 Y guide 43 Y carriage 44 Z carriage 46 Y drive unit 48 Z drive unit 49 Relative movement mechanism 52 Coaxial lighting section 52a coaxial illumination source 52b Half mirror 52c objective lens 52d Half Mirror 52e Transmission Filter 52f transmission filter 52g Dichroic Mirror 54 Ring lighting unit 54a light source 56 Color Camera 56a Imaging lens 56b color image sensor 57 pixels 58 Color Filter Array 58B Color Filter 58G Color Filter 58R color filter 59A 1st Camera 59B Second Camera 60 General Control Unit 62 Operation section 64 Memory section 66 Display section 70 Blade drive control unit 72 Movement control unit 74 Shooting control unit 76 Brightness adjustment section 78 Image Processing Unit 80 Alignment detection unit 82 Processing control unit 90a First imaging lens 90b Second imaging lens 92a First image sensor 92b Second imaging element 100 common light source 101 White LED 102 Half Mirror 104 Transmission Filter 106 Mirror 108 Transmission Filter 110 Microscope 111 Filter switching mechanism 112A, 112B Transmission Filter AR internal area CA Rotation Axis D1 Brightfield image D2 dark field image D3 captured image DC composite image F Frame L0 white light L1 coaxial illumination light L1A Specular reflection light L2 Oblique illumination light L2A scattered light LM mixed light O1 optical axis double work λA wavelength range λB wavelength range λIR wavelength range

Claims

[Claim 1] a coaxial illumination unit that irradiates a workpiece with coaxial illumination light in a first wavelength range; an oblique illumination unit that irradiates the workpiece with oblique illumination light in a second wavelength range different from the first wavelength range; an imaging unit into which mixed light of specularly reflected light of the coaxial illumination light specularly reflected by the workpiece and scattered light of the oblique illumination light scattered by the workpiece is incident, and which separates the mixed light into the specularly reflected light and the scattered light and simultaneously images them; an image processing unit that generates a composite image by combining the first captured image of the specular reflection light and the second captured image of the scattered light captured by the imaging unit; Equipped with The image processing unit Dividing the first captured image and the second captured image into a plurality of divided regions, calculating a contrast evaluation value of the first captured image and a contrast evaluation value of the second captured image for each of the divided regions; a semiconductor manufacturing device that compares the contrast evaluation value of the first captured image with the contrast evaluation value of the second captured image for each divided area, selects and connects the divided area with the higher contrast evaluation value to generate the composite image.

Images