JP6762834B2 - Measuring device - Google Patents

Description

The present invention relates to a measuring device capable of measuring the thickness of a thick wafer with high accuracy.

A wafer in which a plurality of devices such as ICs and LSIs are partitioned by a scheduled division line and formed on the upper surface is divided into individual devices by a dicing device and a laser processing device after the lower surface is ground and thinned by a grinding device and carried. Used for electrical equipment such as telephones and personal computers.

The grinding device for grinding the lower surface of the wafer includes a chuck table for holding the wafer, a wafer held on the chuck table, and a grinding means rotatably provided with a grinding wheel for grinding the wafer held on the chuck table. It is generally composed of a measuring means for measuring the thickness of the wafer held on the chuck table, and the wafer can be processed to a predetermined thickness.

As a measuring means for measuring the thickness of a work piece such as a wafer, a contact type measuring means for measuring the thickness of a wafer by bringing the tip of a prober into contact with the ground surface of the wafer has been conventionally known. However, if the contact type measuring means is used, the ground surface may be scratched, and in order to avoid this, the light reflected from the ground surface of the wafer and the light transmitted through the wafer and reflected from the opposite surface It has been proposed to use a non-contact type measuring means for measuring the thickness by a spectral interference waveform (see, for example, Patent Document 1).

Further, when the focusing point of the laser beam having a wavelength that is transparent to the wafer is positioned and irradiated from the lower surface of the wafer to the inside corresponding to the planned division line to form the modified layer along the planned division line. It is also known that a non-contact type measuring means for detecting the height of the lower surface of a wafer by a spectral interference waveform is used to position a condensing point at an appropriate internal position (see, for example, Patent Document 2).

Japanese Unexamined Patent Publication No. 2012-021916 Japanese Unexamined Patent Publication No. 2011-122894

By using the non-contact type measuring device described above, the thickness and height can be measured without damaging the wafer. Here, in the non-contact type measuring device, an image sensor that generates interference light by a diffractive lattice and detects the intensity of the interference light for each wavelength is required, but the thickness of the wafer to be measured is As the thickness increases, the number of interfering lights increases, and as a result, the unevenness of the waves constituting the spectral interference waveform becomes finer, and it becomes necessary to increase the number of pixels of the image sensor and increase the resolution in order to improve the measurement accuracy. However, since the image sensor becomes expensive when the number of pixels of the image sensor is increased, the reality is that the number of pixels of the image sensor that can be adopted is limited, and when an image sensor with a small number of pixels is used. There is a problem that sufficient measurement accuracy cannot be obtained when measuring a thick wafer.

The present invention has been made in view of the above facts, and its main technical problem is to have a configuration capable of measuring the thickness with high accuracy even when an image sensor having a small number of pixels is used. The purpose is to provide a measuring device.

In order to solve the above-mentioned main technical problem, according to the present invention, the holding means for holding the waha and the waha held by the holding means are irradiated with light in a predetermined wavelength range having transparency to the waha. A measuring device for measuring thickness, a broadband light source that emits light in a predetermined wavelength range that is transparent to a wafer to a first optical path, and a broadband light source that is arranged in the first optical path and emits light. A condenser that collects light and irradiates a wafer held by the holding means, and a light emitted by the broadband light source that is arranged in the first optical path between the broadband light source and the condenser. An optical branch portion that branches the return light that is guided to the condenser and reflected from the upper surface and the lower surface of the wafer held by the holding means into a second optical path, and the return light that is arranged in the second optical path. A diffractive lattice that disperses light in a predetermined wavelength range, an image sensor that detects the intensity of light for each wavelength dispersed by the diffractive lattice, and a spectrum based on the intensity of light for each wavelength detected by the image sensor. It is composed of at least a calculation means that generates an interference waveform, analyzes the generated spectral interference waveform to calculate the thickness of the wafer, and is located between the optical branching portion and the image sensor on the second optical path. A spectroscopic selection means is provided which expands the irradiation region of the dispersed light and guides the end portion of the wavelength region of the light to the outside of the image sensor and guides the central portion of the dispersed light to the image sensor. Measuring equipment is provided.

The spectroscopic selection means is a zoom lens disposed between the image sensor and the diffractive lattice, and the end of the wavelength region of the spectroscopic light magnified by the zoom lens is outside the image sensor. It can be configured so that the central portion is guided by the image sensor. Further, the spectroscopic selection means is a diffraction lattice, and since the dispersion angle of the light after spectroscopy is set to be expanded, the end of the wavelength region of the dispersed light is outside the image sensor. It may be configured to be guided by.

The diffraction lattice includes a high-density region in which the dispersion angle of light after spectroscopy is large and a low-density region in which the dispersion angle is small with respect to the high-density region, and is guided to the second optical path. When the high density region and the low density region are selectively positioned with respect to the return light, and the high density region is positioned with respect to the return light guided to the second optical path, the region is positioned as a selection. The edge of the wavelength region of the dispersed light is guided to the outside of the image sensor, the central part is guided to the image sensor, and the region having a low density with respect to the return light guided to the second optical path is formed. When positioned, it may be configured so that all of the dispersed light is guided to the image sensor.

The measuring device further includes a beam splitter arranged in the first optical path between the optical branch and the condenser, and a beam splitter that guides a part of the light emitted by the broadband light source to the third optical path. The calculation means includes a reflecting unit that reflects the light guided to the third optical path and returns it to the first optical path via the beam splitter to generate a reference light that serves as a reference for the spectral interference waveform. In addition to the thickness of, the height of the upper surface is calculated based on the spectral interference waveform of the return light reflected from the upper surface of the waiha and the reference light, and the spectral interference waveform of the return light reflected from the lower surface of the waiha and the reference light. The height of the lower surface can be calculated based on this.

The present invention is a holding means for holding a waha and a measuring device for measuring the thickness of the waha by irradiating the waha held by the holding means with light in a predetermined wavelength range having transparency. A broadband light source that emits light in a predetermined wavelength range that is transparent to the first optical path, and light that is arranged in the first optical path and emitted by the broadband light source is collected and held by the holding means. The concentrator that irradiates the wavelength and the light that is arranged in the first optical path between the broadband light source and the condenser is guided to the concentrator and the holding means. An optical branch portion that branches the return light reflected from the upper surface and the lower surface of the held wafer into a second optical path, and a diffraction lattice that is arranged in the second optical path and disperses the return light in a predetermined wavelength range. An image sensor that detects the intensity of light for each wavelength dispersed by the diffraction lattice, and a spectral interference waveform generated by generating a spectral interference waveform based on the intensity of light for each wavelength detected by the image sensor. It is composed of at least a calculation means for calculating the thickness of the wafer by analyzing the wavelength of the wavelength, and the irradiation region of the dispersed light is expanded between the optical branching portion and the image sensor on the second optical path. The number of pixels constituting the image sensor is provided by providing a spectroscopic selection means in which the end portion of the wavelength region of the light is guided to the outside of the image sensor and the central portion of the dispersed light is guided to the image sensor. It is possible to improve the accuracy when forming the spectral interference waveform, and improve the focusing system when measuring the height and thickness of the wafer.

It is an overall perspective view of the laser processing apparatus to which the measuring apparatus based on this invention was applied, and the drawing which shows the wafer which is the object to be measured. It is a block diagram for demonstrating the measuring apparatus configured based on this invention. It is a figure which shows the spectroscopic interference waveform measured by the measuring apparatus of this invention. It is a drawing which shows the optical path length difference and the signal intensity obtained by waveform analysis of the spectral interference waveform shown in FIG. It is a drawing which shows another Example of the spectroscopic selection means which comprises the measuring apparatus of this invention.

Hereinafter, the measuring device according to the present invention will be described in detail with reference to the accompanying drawings.
FIG. 1 shows an overall perspective view of a laser processing device 40 to which a measuring device 70 configured based on the present invention is applied, and a perspective view of a wafer 10 as a work piece to be measured. .. The wafer 10 is made of, for example, a sapphire (Al ₂ O ₃ ) substrate, and a plurality of scheduled division lines 12 are formed in a grid pattern on the upper surface of the wafer 10 having a substantially disk shape. The upper surface of the wafer 10 is divided into a plurality of regions. A device 14 (for example, an optical device such as an LED) is formed in each of the plurality of regions, and the wafer 10 is held by attaching the lower surface side to an adhesive tape T having an outer circumference attached to an annular frame F. ..

The laser processing apparatus 40 shown in FIG. 1 includes a base 41, a holding mechanism 42 for holding the work piece, a processing feed means 43 for moving the holding mechanism 42, and a work piece held by the holding mechanism 42. It is composed of a laser beam irradiating means 44 for irradiating a laser beam, an imaging means 45, a frame body 46 incorporating the laser beam irradiating means 44 extending upward from the upper surface of the base 41 and then extending substantially horizontally, and a computer described later. The control means is provided, and each means is controlled by the control means. Further, a measuring means 70 configured based on the present invention is also built in the frame body 46.

The holding mechanism 42 has a rectangular X-axis direction movable plate 51 movably mounted on the base 41 in the X-axis direction and a rectangular shape mounted on the X-axis direction movable plate 51 movably in the Y-axis direction. It includes a Y-axis direction movable plate 53, a cylindrical support plate 50 fixed to the upper surface of the Y-axis direction movable plate 53, and a rectangular cover plate 52 fixed to the upper end of the support column 50. The cover plate 52 is formed with an elongated hole 52a extending in the Y-axis direction. On the upper surface of the chuck table 54 as a holding means for directly holding the circular workpiece extending upward through the elongated hole 52a, a circular adsorption formed of a porous material and extending substantially horizontally. The chuck 56 is arranged. The suction chuck 56 is connected to a suction means (not shown) by a flow path passing through the support column 50. A plurality of clamps 58 are arranged on the peripheral edge of the chuck table 54 at intervals in the circumferential direction. The X-axis direction is the direction indicated by the arrow X in FIG. 1, and the Y-axis direction is the direction indicated by the arrow Y in FIG. 1 and is orthogonal to the X-axis direction. The planes defined in the X-axis direction and the Y-axis direction are substantially horizontal.

The machining feed means 43 includes an X-axis direction moving means 60, a Y-axis direction moving means 65, and a rotating means (not shown). The X-axis direction moving means 60 has a ball screw 60b extending in the X-axis direction on the base 41 and a motor 60a connected to one end of the ball screw 60b. A nut portion (not shown) of the ball screw 60b is fixed to the lower surface of the movable plate 51 in the X-axis direction. Then, the X-axis direction moving means 60 converts the rotational motion of the motor 60a into a linear motion by the ball screw 60b and transmits it to the X-axis direction movable plate 51, and moves in the X-axis direction along the guide rail 43a on the base 41. The plate 51 is advanced and retracted in the X-axis direction. The Y-axis direction moving means 65 has a ball screw 65b extending in the Y-axis direction on the X-axis direction movable plate 51, and a motor 65a connected to one end of the ball screw 65b. A nut portion (not shown) of the ball screw 65b is fixed to the lower surface of the Y-axis direction movable plate 53. Then, the Y-axis direction moving means 65 converts the rotational motion of the motor 65a into a linear motion by the ball screw 65b, transmits it to the Y-axis direction movable plate 53, and follows the guide rail 51a on the X-axis direction movable plate 51. The movable plate 53 in the Y-axis direction is moved forward and backward in the Y-axis direction. The rotating means is built in the support column 50 and rotates the suction chuck 56 with respect to the support column 50.

The laser beam irradiating means 44 is built in a frame body 46 extending upward from the upper surface of the base 41 and then extending substantially horizontally. The laser beam irradiating means 44 irradiates along the upper surface of the planned division line 12 of the wafer 10 to be processed, and is a laser beam having a wavelength that is transparent to the wafer 10 held on the chuck table 54. This is for carrying out a trait layer forming step in which a condensing point is positioned at a predetermined depth position from the upper surface of the wafer 10 and irradiated to form a modified layer along a scheduled division line 12 and used as a division starting point. .. As the laser beam irradiating means 44, a well-known one can be adopted and does not constitute a main part of the present invention, and therefore details thereof will be omitted.

As described above, the measuring means 70 is built in the frame body 46, and the concentrator 75 forming a part of the measuring means 70 is the laser beam described above on the lower surface side of the tip end portion of the frame body 46. It is arranged at a position adjacent to the condenser of the irradiation means 44 in the X-axis direction.

The measuring means 70 will be described in more detail with reference to FIG. The measuring device 70 in the illustrated embodiment has a broadband light source 71 that emits light having a predetermined wavelength region that is transparent to the wafer 10 to be measured, and a first optical path 70a that emits light from the broadband light source 71. It is provided with an optical branching portion 72 that guides the return light that reverses the first optical path 70a to the second optical path 70b, and a collimation lens 731 that forms the light guided to the first optical path 70a into parallel light. An optical branching means 74 including a collimation lens mechanism 73, a beam splitter 741 for branching light formed in parallel light by the collimation lens mechanism 73 from a first optical path 70a to a third path 70c, and a first optical path. It is provided with a condenser 75 having a built-in condenser lens 751 which is positioned at the terminal end of 70a and concentrates and irradiates the light from the broadband light source 71 on the wafer 10.

As the broadband light source 71, for example, a light source capable of emitting light having a wavelength in the region of 1100 to 1900 nm can be adopted, and for example, an LED, SLD (Superluminescent division), ASE (Amplified Spontaneous Mission), SC (Supercontinum) can be adopted. ), Halogen light source, etc. can be selected

A polarization-retaining fiber circulator can be adopted as the optical branching portion 72, but the present invention is not limited to this, and a polarization-retaining fiber coupler or the like may be used. In the illustrated embodiment, the optical branching means 74 is composed of a beam splitter 741 and a direction changing mirror 742 that constitutes a reflection unit. Further, the path from the broadband light source 71 to the optical branching portion 72, and the optical branching portion 72 forming a part of the first optical path 70a to the collimation lens mechanism 73 are composed of an optical fiber, and the collimation lens mechanism 73 is formed. The transmitted light is collected by the condensing lens 751 of the condensing device 75 and irradiated to the wafer 10.

Although not shown, the condenser 75 is provided with a position adjusting means that moves up and down with respect to the upper surface of the suction chuck 56 of the chuck table 54, and is configured to be able to adjust the position of the focusing point with respect to the wafer 10. There is.

At the end of the third optical path 70c described above, a reflection mirror 76 for reflecting parallel light guided to the third optical path 70c and returning to the third optical path 70c to reverse the light is arranged. .. In the illustrated embodiment, the reflection mirror 76 is attached to the above-mentioned condenser 75.

In the second optical path 70b, a collimation lens 771 and a diffractive lattice 772, and a low-magnification zoom are used to appropriately disperse the light reflected by the workpiece 10 and the reflection mirror 76 and reversed. A spectroscopic selection means 77 configured from the lens 773 is provided. The collimation lens 771 is reflected by the reflection mirror 76, is returned to the first optical path 70a via the third optical path 70c and the beam splitter 741, and travels back through the collimation lens 73 from the optical branch portion 72 to the second optical path 70b. The guided return light (hereinafter referred to as "reference light") is reflected by the upper and lower surfaces of the wafer 10 held on the suction chuck 56 of the chuck table 56, and is reflected on the condenser 75, the beam splitter 741, and the collimation. The lens mechanism 73 is reversed to form each of the return light guided from the optical branching portion 72 to the second optical path 70b into parallel light. The optical path length L1 of the reference light constituting the beam splitter 741 to the reflection mirror 76 is set to be shorter than the optical path length L constituting the beam splitter 741 to the surface of the suction chuck 56 of the chuck table 54. For example, the optical path length is set to be 2000 μm shorter than the optical path length L.

The diffraction lattice 772 is configured to diffract the interference between the reference light and the return light formed in parallel light by the collimation lens 771 and disperse the interference in a predetermined wavelength range, and further expand the dispersion angle. Is set to. Further, the diffraction signal corresponding to each wavelength of the light dispersed by the diffraction grid 772 is configured to be sent to the line image sensor 78 side via the low magnification zoom lens 773. In the line image sensor 78, a large number of light receiving elements are continuously arranged in a straight line, and the light intensity at each wavelength of the light dispersed by the diffraction lattice 772 corresponding to the position of each light receiving element is detected. The light intensity signal detected as described above is sent to the control means 20 described later.

The control means 20 is composed of a computer, and temporarily stores a central processing unit (CPU) that performs arithmetic processing according to a control program, a read-only memory (ROM) that stores a control program and the like, detected detection values, arithmetic results, and the like. It has a readable and writable random access memory (RAM) for storing in, an input interface, and an output interface (details are not shown).

The low-magnification zoom lens 773 arranged in the spectroscopic selection means 77 described above is configured to be capable of enlarging / reducing the light dispersed by the diffraction grid 772 and emitting it as parallel light. In the embodiment shown in FIG. 2, the return light reflected by the upper surface, the lower surface, and the reflection mirror 76 of the weight 10 is dispersed for each wavelength region by the diffraction lattice 772, and the wavelength of the dispersed light is 1100. While it is composed of a wavelength of ~ 1900 nm, by enlarging the return light with the zoom lens 773, the end of the wavelength region constituting the return light, that is, the wavelength region of 1100 to 1300 nm, and 1700 A wavelength region of ~ 1900 nm is guided out of the image sensor, and only the central portion (1300 to 1700 nm) of the wavelength of the dispersed light is guided to the line image sensor 78.

The control means 20 obtains a spectral interference waveform as shown in FIG. 3 based on the detection signal from the line image sensor 78, executes waveform analysis based on the spectral interference waveform and the theoretical waveform function, and performs beam splitter 741. The optical path length L2 from the beam splitter 741 to the upper surface of the waiha 10 held by the chuck table 54, the optical path length L3 from the beam splitter 741 to the lower surface of the waha 10 held by the chuck table 54, and the optical path length L1 of the reference light. The length difference and the optical path length difference between the optical path length L2 and the optical path length L3 are obtained, and the heights of the upper surface and the lower surface of the waiha 10 held on the upper surface of the chuck table 54 and the waha 10 based on the optical path length difference. Find the thickness of. In FIG. 3, the horizontal axis represents the wavelength of the return light, and the vertical axis represents the light intensity.

Hereinafter, an example of waveform analysis performed by the control means 20 based on the spectral interference waveform as shown in FIG. 3 will be described in more detail.
The difference between the optical path length L1 and the optical path length L2 is the first optical path length difference (d1 = L2-L1), and the difference between the optical path length L1 and the optical path length L3 is the second optical path length difference (d2 = L3-L1). The difference between the optical path length L3 and the optical path length L2 is defined as a third optical path length difference (d3 = L3-L2). As described above, the optical path length L1 itself is the optical path length of the reference light that does not change, and is set to be shorter than the optical path lengths L2 and L3 from the beam splitter 741 to the upper and lower surfaces of the wafer 10.

The control means 20 executes waveform analysis based on the spectral interference waveform and the theoretical waveform function. This waveform analysis can be executed based on, for example, the Fourier transform theory or the wavelet transformation theory, but in the embodiment described below, an example using the Fourier transform equations shown in the following equations 1, 2, and 3. Will be described.

In the above formula, λ is the wavelength, d is the first optical path length difference (d1 = L2-L1), the second optical path length difference (d2 = L3-L1), and the third optical path length difference (d3 = L3). −L2) and W (λn) are window functions. In the above formula 1, the wave period is the closest (high correlation) in comparison between the theoretical waveform of cos and the spectral interference waveform (I (λn)), that is, the phase between the spectral interference waveform and the theoretical waveform function. Find the optical path length difference (d) with a high number of relationships. Further, in the above equation 2, the wave period is the closest (high correlation) in comparison between the theoretical waveform of sin and the spectral interference waveform (I (λn)), that is, the spectral interference waveform and the theoretical waveform function. The first optical path length difference (d1 = L2-L1), the second optical path length difference (d2 = L3-L1), and the third optical path length difference (d3 = L3-L2), which have high correlation coefficients of .. Then, the above formula 3 obtains the average value of the result of the formula 1 and the result of the formula 2.

The control means 20 is provided with calculation means for executing calculations based on the above formulas 1, 2, and 3, so that the signal intensity shown in FIG. 4 is based on the spectral interference caused by the difference in the optical path lengths of the return light. A waveform can be obtained. In FIG. 4, the horizontal axis represents the optical path length difference (d), and the vertical axis represents the signal intensity. In the example shown in FIG. 4, the signal intensity is high at the position where the optical path length difference (d) is 1800 μm, 1000 μm, and 800 μm. That is, the signal intensity (S1) at the position where the optical path length difference (d) is 1800 μm represents the first optical path length difference (d1 = L3-L1) which is the optical path length difference between the reference light and the lower surface of the wafer 10. Is. The signal intensity (S2) at the position where the optical path length difference (d) is 1000 μm represents the second optical path length difference (d2 = L2-L1) which is the optical path length difference between the reference light and the upper surface of the wafer 10. Further, the signal intensity (S3) at the position where the optical path length difference (d) is 800 μm represents the third optical path length difference (d3 = L3-L2). The heights (h) of the upper surface and the lower surface of the wafer 10 with respect to the upper surface of the adsorption check 56 of the chuck table 54 are the optical path length L from the beam splitter 741 to the upper surface of the chuck table 54 and the optical path length of the reference light. It is easily calculated from the optical path length difference from L1 (2000 μm in this embodiment). By operating the X-axis direction moving means 60 and the Y-axis direction moving means 65 described above, such measurement is performed along all the planned division lines 12 while moving the wafer 10 in the X-axis direction and the Y-axis direction. And execute.

Then, the heights of the upper surface and the lower surface of the waiha 10 at the coordinates (X coordinate, Y coordinate) of the measurement position specified by the relative X-axis direction and Y-axis direction of the concentrator 75 and the chuck table 54. (H) and the thickness are stored in the random access memory (RAM) of the control means 20 in association with the X coordinate and the Y coordinate. By performing such measurement, it is possible to obtain information on the height and thickness of the upper surface and the lower surface of the entire surface of the wafer 10. Therefore, based on the respective information, for example, the position of the cutting blade when performing laser processing for forming a modified layer inside a predetermined position from the upper surface of the wafer 10 or performing cutting processing using a cutting blade is determined. , It is possible to accurately position the wafer 10 in the height and thickness directions.

The measuring device 70 configured based on the present invention can exert the following effects by being configured as described above. For example, it is assumed that the number of pixels of the line image sensor 78 in this embodiment is 2048 pixels. When the thickness of the wafer 10 is thin (for example, less than 500 μm), the number of irregularities of the waveform shown in the spectral interference waveform shown in FIG. 3 is small, and the shape of the waveform can be substantially faithfully expressed by the number of 512 pixels. .. Therefore, the line image sensor 78 is irradiated with all the light diffracted by the diffractive grid 772 without passing through the low-magnification zoom lens 773 to acquire the spectral interference waveform, and the first 1st to the same are obtained by the above-mentioned calculation means. The third optical path length difference may be calculated. However, for example, when the thickness is 800 μm or more, the number of interference lights increases, the unevenness of the spectral interference waveform increases, and the line image sensor 78 is irradiated with the entire wavelength region of 1100 to 1900 nm constituting the spectrum. In this case, the waveform cannot be faithfully expressed due to the number of pixels, and the measurement accuracy is lowered. In that respect, according to the measuring means 70 based on the present invention, the irradiation region of the light dispersed by the diffraction lattice is enlarged by a low-magnification zoom lens, and the end portion of the light is guided to the outside of the image sensor and separated. A spectroscopic selection means 77 is provided that functions so that only the central portion of the light is guided to the image sensor. As a result, all the light dispersed by applying the zoom lens 773 of the spectroscopic selection means 77 is not guided to the line image sensor 78, and only the central portion of the wavelength region (for example, the region having a wavelength of 1300 to 1700 nm) is lined. It can be guided to the image sensor 78 to obtain a spectral interference waveform, and even if the number of interference lights increases, the waveform shape can be faithfully expressed by the limited number of pixels, and the waver 10 can be used. It is possible to ensure the accuracy of measuring height and thickness.

The present invention is not limited to the above embodiment, and various modifications can be assumed. For example, in the above embodiment, the spectroscopic selection means 77 is composed of a diffraction grid 772 configured to expand the dispersion angle and a zoom lens 773 to further expand the irradiation region, but is not limited thereto. For example, the return light may be formed into parallel light by the collimated lens 771 and directly incident on the line image sensor 78 from the incident diffraction grid 772. In that case, by adjusting the dispersion angle formed by the diffraction lattice 772 and the position of the line image sensor 78, only the central part of the wavelength region (for example, the wavelength 1300) is not guided without guiding all the dispersed light. A region of ~ 1700 nm) is guided to the line image sensor 78, and a spectral interference waveform can be obtained.

Further, as another embodiment of the above-mentioned spectroscopic selection means 77, another spectroscopic selection means 80 as shown in FIG. 5 can be adopted. The spectroscopic selection means 80 shown in the figure includes a collimation lens 81 that forms the return light branched from the optical branching portion 72 into parallel light, and a variable diffraction lattice that diffracts the return light formed into parallel light by the collimation lens 81. The line image sensor 78 is provided with the condensing lens 84 that collects the return light dispersed by the variable diffraction lattice 82, and the return light that has passed through the condensing lens 84 and is condensed. The variable diffraction lattice 82 is composed of a high-density region 82a having a large dispersion angle and a low-density region 82b having a small dispersion angle, and is irradiated with parallel light formed by a collimation lens 81. It is composed of a drive mechanism 83 that switches the region to be formed between the high-density region 82a and the low-density region 82b.

When measuring the height and thickness of a relatively thick wafer 10, an instruction signal is sent to the drive mechanism 83 via the control device 20 to perform variable diffraction in the direction indicated by the arrow in FIG. 5A. The lattice 82 is moved, and the region irradiated with the return light is designated as the high-density region 82a. As a result, the dispersion angle is expanded, and when the return light diffracted by the variable diffraction lattice is irradiated to the line image sensor 78 through the condensing lens 84, all of the dispersed light is the line image sensor. The line image sensor 78 is irradiated with only the central portion of the wavelength region (for example, the region having a wavelength of 1300 to 1700 nm) without being guided by the 78.

On the other hand, when measuring the height and thickness of the relatively thin wafer 10, the variable diffraction grid 82 is moved in the direction indicated by the arrow in FIG. 5 (b) to reduce the density of the region irradiated with the return light. Area 82b. As a result, the dispersion angle is reduced, and all of the return light diffracted by the variable diffraction lattice is irradiated to the line image sensor 78 via the condensing lens 84. According to such a configuration, it is possible to accurately measure the height and thickness of the wafer 10 regardless of whether the wafer 10 is thin or thick.

In the above-described embodiment, the line image sensor 78 is adopted as a means for detecting the light intensity of the return light for each wavelength, but the sensor is not limited to this as long as it can detect the light intensity, and is general. Image sensors such as photodetectors and CCDs can be used.

Further, in the above-described embodiment, not only the thickness of the wafer 10 but also the heights of the upper surface and the lower surface of the wafer 10 held by the chuck table 54 are measured by using the reference light reflected by the reflection mirror 76. However, the present invention is not limited to this, and it is not an essential requirement to generate a reference light. According to the present invention, even when only the thickness of the wafer 10 is measured without using the reference light, the thickness of the wafer 10 can be measured accurately by providing the above-mentioned spectroscopic selection means 77 or 80. It can produce a unique action effect. In that case, only S3 is displayed as the peak of the signal strength shown in FIG.

Further, in the present embodiment, an example in which the measuring device 70 based on the present invention is applied to the laser processing device 40 is disclosed, but the present invention is not limited to this, and is applied to other processing devices such as a grinding device. Is not often specified by the type of processing equipment. Further, the measuring device of the present invention is not limited to the one built in the processing device, and can be configured as an independent measuring device for measuring the thickness and height of the wafer 10.

10: Waha 12: Scheduled division line 14: Device 20: Control means 40: Laser processing device 42: Holding means 43: Processing feeding means 44: Laser beam irradiation means 47: Display means 54: Chuck table 56: Adsorption chuck 70: Measuring means 70a: First optical path 70b: Second optical path 70c: Third optical path 71: Broadband light source 72: Optical branch 73: Collimation lens mechanism 74: Optical branch means 75: Condenser 76: Reflective mirror 77, 80: Spectral selection means 78: Line image sensor 81: Collimation lens 82: Variable diffraction lattice 82a: High density region 82b: Low density region 84: Condensing lens

Claims (5)

A holding means for holding a wafer and a measuring device for measuring the thickness of the wafer by irradiating the wafer held by the holding means with light in a predetermined wavelength range having transparency.
A broadband light source that emits light in a predetermined wavelength range that is transparent to the wafer to the first optical path, and light that is arranged in the first optical path and emitted by the broadband light source is collected and held by the holding means. The concentrator that irradiates the waiha and the light emitted by the broadband light source, which is arranged in the first optical path between the broadband light source and the concentrator, is guided to the concentrator and the holding means. An optical branching portion that branches the return light reflected from the upper surface and the lower surface of the wafer held in the second optical path into a second optical path, and diffraction that is arranged in the second optical path and disperses the return light in a predetermined wavelength range. A lattice, an image sensor that detects the intensity of light for each wavelength dispersed by the diffraction lattice, and a spectral interference waveform generated based on the intensity of light for each wavelength detected by the image sensor, and the generated spectral interference It is composed of at least a calculation means that analyzes the waveform and calculates the thickness of the wafer.
Between the optical branch and the image sensor on the second optical path, the irradiation region of the dispersed light is expanded, and the end of the wavelength region of the light is guided to the outside of the image sensor and is separated. A measuring device provided with a spectroscopic selection means in which a central portion of the emitted light is guided to the image sensor. The spectroscopic selection means is a zoom lens disposed between the image sensor and the diffraction lattice, and the end of the wavelength region of the spectroscopic light magnified by the zoom lens is outside the image sensor. The measuring device according to claim 1, wherein the central portion is guided to the image sensor. The spectroscopic selection means is a diffraction lattice, and by setting the dispersion angle of the separated light to be expanded, the end of the wavelength region of the separated light is guided to the outside of the image sensor. The measuring device according to claim 1, wherein the central portion is guided to the image sensor at the same time. The diffraction lattice includes a high-density region in which the dispersion angle of light after spectroscopy is large and a low-density region in which the dispersion angle is small with respect to the high-density region.
The high-density region and the low-density region are selectively positioned with respect to the return light guided to the second optical path.
When the dense region is positioned with respect to the return light guided to the second optical path, the end of the wavelength region of the dispersed light is guided to the outside of the image sensor and the central portion is the center. Guided by the image sensor
The measuring device according to claim 3, wherein when the low-density region is positioned with respect to the return light guided to the second optical path, all of the dispersed light is guided to the image sensor. The measuring device further includes a beam splitter arranged in the first optical path between the optical branch and the condenser, and a beam splitter that guides a part of the light emitted by the broadband light source to the third optical path. The calculation means includes a reflecting unit that reflects the light guided to the third optical path and returns it to the first optical path via the beam splitter to generate a reference light that serves as a reference for the spectral interference waveform. In addition to the thickness of, the height of the upper surface is calculated based on the spectral interference waveform of the return light reflected from the upper surface of the waiha and the reference light, and the spectral interference waveform of the return light reflected from the lower surface of the waiha and the reference light. The measuring device according to claim 1, wherein the height of the lower surface is calculated based on the above.