Inspection device and inspection method

Abstract

An inspection device 1 comprises: a pulse formation unit 3 that generates a pulse train Pb in which a plurality of light pulses Pb1, Pb2 having different center wavelengths are arranged at predetermined time intervals; a first adjustment unit 4 that makes the polarization directions of each of the light pulses Pb1, Pb2 of the pulse train Pb inputted to an photonic integrated circuit 5 coincide with respect to the in-plane direction of a substrate 50; a correlation optical system 7 that has polarization dependency in a predetermined polarization direction and generates correlation light Pc of the pulse train Pb outputted from the photonic integrated circuit 5; a light detection unit 8 that detects a time waveform of the correlation light Pc; and a determination unit 9 that determines the quality of the photonic integrated circuit 5 on the basis of the time waveform. The light detection unit 8 detects the time waveform a plurality of times. The determination unit 9 determines the quality of the photonic integrated circuit 5 on the basis of a plurality of feature amounts of the time waveform thus detected a plurality of times.

Description

Inspection device and inspection method

[0001] The present disclosure relates to an inspection device and an inspection method.

[0002] The dispersion measurement device described in Patent Document 1 includes a pulse forming unit that forms, from a first optical pulse output from an object to be measured, a pulse train including a plurality of second optical pulses that have a time difference from one another and different center wavelengths; a correlation optical system that receives the pulse train output from the pulse forming unit and outputs correlated light that includes the autocorrelation of the pulse train; a light detection unit that detects the time waveform of the correlated light; and a calculation unit that estimates the amount of chromatic dispersion of the object to be measured based on the feature quantities of the time waveform.

[0003] The dispersion measurement device described above is expected to be used as an inspection device for photonic integrated circuits (PICs). An optical integrated circuit is a device that integrates various optical functions and transmits information signals using light. For example, when a pulse train is input into an optical integrated circuit, chromatic dispersion can occur as the pulse train propagates through the waveguide of the optical integrated circuit. Therefore, the quality of an optical integrated circuit can be determined by estimating the chromatic dispersion in the optical integrated circuit using a dispersion measurement device and inspecting whether the chromatic dispersion deviates from the design value.

[0004] Japanese Patent Application Laid-Open No. 2020-169946

[0005] When determining the quality of an optical integrated circuit using a dispersion measurement device, structural factors such as the state of the sidewalls that make up the waveguide may affect the error in repeated measurements of the chromatic dispersion amount. Therefore, there is a need for a technology that takes structural factors into account and can improve the accuracy of determining the quality of an optical integrated circuit.

[0006] The present disclosure provides an inspection device and an inspection method that take structural factors into consideration and can improve the accuracy of determining whether an optical integrated circuit is good or bad.

[0007] The gist of the present disclosure is as follows.

[0008] [1] An inspection device for inspecting an optical integrated circuit having a waveguide on a substrate, the inspection device comprising: a pulse forming unit that generates a pulse train in which a plurality of optical pulses with different center wavelengths are arranged at a predetermined time interval; a first adjustment unit that aligns the polarization direction of each optical pulse of the pulse train input to the optical integrated circuit with an in-plane direction of the substrate; a correlation optical system that has polarization dependency on the predetermined polarization direction and generates correlated light including cross-correlated light or auto-correlated light of the pulse train that is output from the optical integrated circuit through the waveguide; an optical detection unit that detects a time waveform of the correlated light; and a judgment unit that judges the quality of the optical integrated circuit based on the time waveform, wherein the optical detection unit detects the time waveform multiple times, and the judgment unit judges the quality of the optical integrated circuit based on multiple feature quantities in the time waveform detected multiple times.

[0009] In this inspection system, a pulse train is input into the optical integrated circuit so that the polarization direction of each of the multiple optical pulses is aligned with the in-plane direction of the substrate. As a result, the polarization direction of the optical pulses intersects with structures such as the sidewalls that make up the waveguide, and if there is an abnormality in their formation, the error in repeated measurements of the chromatic dispersion amount increases. In this inspection system, the chromatic dispersion amount is considered as a feature value in the time waveform. By using the error in repeated measurements of feature values ​​calculated based on multiple feature values, the accuracy of pass / fail judgment of optical integrated circuits can be improved while taking structural factors into account.

[0010] [2] The inspection device according to [1], wherein the judgment unit judges the quality of the optical integrated circuit based on whether the plurality of feature quantities are within a predetermined tolerance range. In this case, by comparing the plurality of feature quantities with the tolerance range, the accuracy of the quality judgment can be further improved.

[0011] [3] The inspection device according to [1], wherein the judgment unit judges the quality of the optical integrated circuit based on whether or not the amount of measurement error calculated based on the plurality of feature quantities is within a predetermined tolerance range. In this case, the accuracy of the quality judgment of the optical integrated circuit can be further improved by taking into account the statistical trends of the plurality of feature quantities.

[0012] [4] The inspection device according to [2] or [3], further comprising a storage unit for storing the tolerance data, wherein the judgment unit reads the tolerance data from the storage unit and judges the quality of the optical integrated circuit. In this case, by having a storage unit for storing the tolerance data in advance, the time required to read the data can be shortened, thereby shortening the time required to judge the quality.

[0013] [5] The inspection device according to any one of [1] to [4], further comprising a first polarization-maintaining fiber that optically connects the pulse shaping unit to the optical integrated circuit, and the first adjustment unit is configured with a rotation mechanism that rotates the first polarization-maintaining fiber about its axis. In this case, by rotating the first polarization-maintaining fiber about its axis, the polarization direction of each optical pulse in the pulse train input to the optical integrated circuit can be easily adjusted.

[0014] [6] The inspection device according to any one of [1] to [5], further comprising a second adjustment unit that extracts a TE mode component of the pulse train output from the optical integrated circuit. In this case, the second adjustment unit inputs the pulse train into the correlation optical system so that the polarization direction of the pulse train coincides with the polarization direction in which the correlation optical system has polarization dependency. Therefore, correlation light can be efficiently generated in the correlation optical system.

[0015] [7] The inspection device according to [6], further comprising a second polarization-maintaining fiber that optically connects the optical integrated circuit to the correlation optical system, and the second adjustment unit is configured by a rotation mechanism that rotates the second polarization-maintaining fiber about its axis. In this case, by rotating the in-plane direction of the substrate of the optical integrated circuit about the axis of the second polarization-maintaining fiber, the polarization mode of each optical pulse of the pulse train input to the second polarization-maintaining fiber can be easily selected.

[0016] [8] The inspection device according to [7], wherein the rotation mechanism rotates the second polarization-maintaining fiber around its axis so as to extract the TE mode component of the pulse train output from the optical integrated circuit. In this case, it is possible to prevent the pulse train Pb from being input to the correlation optical system 7 in a polarization direction that is undesirable for measurement.

[0017] [9] An inspection method for inspecting an optical integrated circuit having a waveguide provided on a substrate, the inspection method comprising: a pulse forming step of generating a pulse train in which a plurality of optical pulses having different center wavelengths are arranged at a predetermined time interval; a first adjustment step of aligning the polarization direction of each optical pulse of the pulse train input to the optical integrated circuit with an in-plane direction of the substrate; a correlation light generation step of generating correlation light including cross-correlated light or auto-correlated light of the pulse train output from the optical integrated circuit using a correlation optical system having polarization dependency in the predetermined polarization direction; an optical detection step of detecting a time waveform of the correlation light; and a judgment step of judging the quality of the optical integrated circuit based on feature quantities of the time waveform, wherein the optical detection step detects the time waveform multiple times, and the judgment step judges the quality of the optical integrated circuit based on multiple feature quantities in the time waveform detected multiple times.

[0018] In this inspection method, a pulse train is input into an optical integrated circuit so that the polarization direction of each of the multiple optical pulses is aligned with the in-plane direction of the substrate. As a result, the polarization direction of the optical pulses intersects with structures such as sidewalls that make up the waveguide, and if there is an abnormality in their formation, the error in repeated measurements of the chromatic dispersion amount increases. In this inspection method, the chromatic dispersion amount is considered as a feature value in the time waveform. By using the error in repeated measurements of feature values ​​calculated based on multiple feature values, the accuracy of pass / fail judgment of optical integrated circuits can be improved while taking structural factors into account.

[0019]

[10] The inspection method according to [9], wherein the determining step determines whether the optical integrated circuit is good or bad based on whether the plurality of feature quantities are within a predetermined tolerance range. In this case, by comparing the plurality of feature quantities with the tolerance range, the accuracy of the pass / fail determination can be further improved.

[0020]

[11] The inspection method according to [9], wherein the determining step determines whether the optical integrated circuit is good or bad based on whether a measurement error calculated based on the plurality of feature quantities is within a predetermined tolerance. In this case, the accuracy of determining whether the optical integrated circuit is good or bad can be further improved by taking into account the statistical trends of the plurality of feature quantities.

[0021] According to the present disclosure, structural factors can be taken into consideration to improve the accuracy of determining whether an optical integrated circuit is good or bad.

[0022] 10(a), 10(b), 10(c), and 10(d) are diagrams showing examples of delay times of pulse trains. A diagram showing an example of the hardware configuration of a determination unit. A flowchart showing an inspection method using an inspection device. A flowchart showing an example of a method for setting a tolerance range. A graph showing an example of an ideal characteristic and a tolerance range. A graph showing an example of a pass / fail judgment of an optical integrated circuit. A graph showing another example of a pass / fail judgment of an optical integrated circuit.

[0023] Hereinafter, a preferred embodiment of an inspection device according to one aspect of the present disclosure will be described in detail with reference to the drawings.

[0024] [Configuration of Inspection Apparatus] Figure 1 is a diagram schematically showing the configuration of an inspection apparatus according to one embodiment of the present invention. This inspection apparatus 1 is an apparatus for inspecting an optical integrated circuit 5. The optical integrated circuit 5 includes a waveguide 53 (see Figure 4) provided on a substrate 50. In this embodiment, the inspection apparatus 1 is an apparatus for measuring the chromatic dispersion of light to be measured that has propagated through the waveguide 53 of the optical integrated circuit 5. The inspection apparatus 1 includes a pulsed laser light source 2, a pulse forming unit 3, a first adjusting unit 4, a second adjusting unit 6, a correlation optical system 7, a light detecting unit 8, and a determining unit 9.

[0025] The optical input end 3a of the pulse forming unit 3 is optically coupled to the pulse laser light source 2 either spatially or via an optical waveguide such as an optical fiber. The optical input end 4a of the first adjusting unit 4 is optically coupled to the optical output end 3b of the pulse forming unit 3 either spatially or via an optical waveguide such as an optical fiber. The optical input end 5a of the optical integrated circuit 5 to be inspected is optically coupled to the optical output end 4b of the first adjusting unit 4 either spatially or via an optical waveguide such as an optical fiber. The optical input end 6a of the second adjusting unit 6 is optically coupled to the optical output end 5b of the optical integrated circuit 5 either spatially or via an optical waveguide such as an optical fiber.

[0026] The optical input end 7a of the correlation optical system 7 is optically coupled to the optical output end 6b of the second adjusting unit 6, either spatially or via an optical waveguide such as an optical fiber. The optical detecting unit 8 is optically coupled to the optical output end 7b of the correlation optical system 7, either spatially or via an optical waveguide such as an optical fiber. The determining unit 9 is electrically connected to each of the pulse forming unit 3, the optical integrated circuit 5, and the optical detecting unit 8. The determining unit 9 may also be electrically connected to each of the first adjusting unit 4 and the second adjusting unit 6.

[0027] The pulsed laser light source 2 outputs a coherent optical pulse Pa to be measured. The pulsed laser light source 2 is, for example, a femtosecond laser, and in one embodiment, is a solid-state laser light source such as an LD directly pumped Yb:YAG pulsed laser. The time waveform of the optical pulse Pa to be measured is, for example, a Gaussian function. The optical pulse Pa to be measured is an optical pulse having a certain bandwidth and includes multiple continuous wavelength components. The pulsed laser light source 2 may be a light source that outputs an attosecond laser, a femtosecond laser, or a picosecond laser, or may be a titanium sapphire laser light source, a semiconductor laser light source, or an F2 laser light source.

[0028] The pulse shaping unit 3 forms a pulse train Pb including a plurality of optical pulses from the measured optical pulse Pa. The pulse train Pb is a group of single pulses generated using each of the wavelength bands that are obtained by dividing the spectrum constituting the measured optical pulse Pa into a plurality of wavelength bands. Note that there may be overlapping portions at the boundaries between the plurality of wavelength bands. In the following example, the pulse train Pb includes two single-pulse optical pulses Pb1 and Pb2. The central wavelength and wavelength band of the optical pulse Pb1 are different from the central wavelength and wavelength band of the optical pulse Pb2.

[0029] The pulse train Pb comprises a plurality of optical pulses Pb1 and Pb2 with different center wavelengths arranged at a predetermined time interval. The peaks of the two optical pulses Pb1 and Pb2 are separated in time, and the propagation timings of the two optical pulses Pb1 and Pb2 are shifted from each other. In other words, one optical pulse Pb2 is delayed in time with respect to the other optical pulse Pb1. However, the tail portions of adjacent optical pulses Pb1 and Pb2 may overlap each other.

[0030] FIG. 2 is a diagram showing an example configuration of the pulse shaping unit 3. The pulse shaping unit 3 includes a diffraction grating 31, a lens 32, a spatial light modulator (SLM) 33, a lens 34, and a diffraction grating 35. The diffraction grating 31 is an example of a spectroscopic element. The diffraction grating 31 is optically coupled to the pulse laser light source 2. The spatial light modulator 33 is optically coupled to the diffraction grating 31 via a lens 32. The diffraction grating 31 spatially separates multiple wavelength components contained in the measured optical pulse Pa into individual wavelengths. Note that other optical components such as a prism may be used as the spectroscopic element instead of the diffraction grating 31. The measured optical pulse Pa is incident obliquely on the diffraction grating 31 and is split into multiple wavelength components. The light P1 containing the multiple wavelength components is focused by the lens 32 into each wavelength component and formed into an image on the modulation surface of the spatial light modulator 33. The lens 32 may be a convex lens made of a light-transmitting member or a concave mirror having a concave light-reflecting surface.

[0031] The spatial light modulator 33 converts the light P1 into modulated light P2. The spatial light modulator 33 may be a reflective spatial light modulator or a transmissive spatial light modulator. Furthermore, the spatial light modulator 33 may be a liquid crystal spatial light modulator or a digital mirror device. The modulated light P2 includes light pulses Pb1 and Pb2. The spatial light modulator 33 selects any two wavelength components from the light P1, which includes multiple wavelength components, as the light pulses Pb1 and Pb2. For example, the spatial light modulator 33 may intensity-modulate the wavelength components of the light pulses Pb1 and Pb2 to a level that does not affect the other wavelength components. The wavelength difference, which is the difference between the wavelength of the light pulse Pb1 and the wavelength of the light pulse Pb2, changes depending on the value of the selected wavelength component. This allows the spatial light modulator 33 to set the wavelength difference between the two light pulses Pb1 and Pb2 to any desired value. The wavelength difference that can be set is, for example, 1 nm to 50 nm.

[0032] The spatial light modulator 33 shifts the phases of the two light pulses Pb1 and Pb2 from each other. To do this, the spatial light modulator 33 performs phase modulation on the light P1. As a result, the light pulse Pb2 has a time delay with respect to the light pulse Pb1. In other words, the spatial light modulator 33 sets the time difference between the two light pulses Pb1 and Pb2 to an arbitrary magnitude. The spatial light modulator 33 performs intensity modulation and phase modulation on the light P1 to generate modulated light P2.

[0033] The modulation of light P1 by the spatial light modulator 33 will be described in more detail. Fig. 3 is a diagram showing the modulation surface 33a of the spatial light modulator 33. As shown in Fig. 3, on the modulation surface 33a, a plurality of modulation regions 33b are arranged along a certain direction A, and each modulation region 33b extends in a direction B intersecting with direction A. Direction A is the direction of light dispersion by the diffraction grating 31. This modulation surface 33a functions as a Fourier transform plane, and each corresponding wavelength component after dispersion is incident on each of the plurality of modulation regions 33b.

[0034] The spatial light modulator 33 modulates the intensity and phase of each incident wavelength component in each modulation region 33b independently of other wavelength components. Each modulation region 33b includes, for example, an intensity modulation pattern and a phase modulation pattern. The intensity modulation pattern and phase modulation pattern are, for example, different for each modulation region 33b. The intensity modulation pattern and phase modulation pattern in each modulation region 33b vary, for example, along direction A and are constant in direction B. The modulation amount of the intensity pattern of the modulation region 33b onto which the wavelength components of the light pulses Pb1 and Pb2 are incident may be greater than the modulation amount of the intensity pattern of the modulation region 33b onto which the other wavelength components are incident. The modulation amount of the phase pattern of the modulation region 33b onto which the wavelength component of the light pulse Pb1 is incident may be greater than the modulation amount of the phase pattern of the modulation region 33b onto which the wavelength component of the light pulse Pb2 is incident.

[0035] The optical pulses Pb1 and Pb2 contained in the modulated light P2 modulated by the spatial light modulator 33 are focused by the lens 34 to a single point on the diffraction grating 35. In this case, the lens 34 functions as a focusing optical system that focuses the modulated light P2. The lens 34 may be a convex lens made of a light-transmitting member, or a concave mirror having a concave light-reflecting surface. The diffraction grating 35 also functions as a combining optical system that combines the optical pulses Pb1 and Pb2. That is, the optical pulses Pb1 and Pb2 of the modulated light P2 are focused and combined by the lens 34 and the diffraction grating 35 to form a bandwidth-controlled multi-pulse (pulse train Pb).

[0036] Referring again to FIG. 1 , the first adjustment unit 4 adjusts the polarization direction of the pulse train Pb output from the pulse shaping unit 3. The first adjustment unit 4 aligns the polarization direction of each optical pulse Pb1, Pb2 of the pulse train Pb input to the optical integrated circuit 5 with the in-plane direction of the substrate 50. In the example of FIG. 1 , for convenience of explanation, the direction in which the pulse train Pb travels is the X-axis direction. The in-plane direction of the substrate 50 is defined by the Y-axis direction, which is perpendicular to the X-axis direction, and the X-axis direction. The thickness direction of the substrate 50 is defined by the Z-axis direction, which is perpendicular to the X-axis direction and the Y-axis direction.

[0037] The in-plane direction of the substrate 50 coincides with the polarization direction in the TE (Transverse Electric) mode (horizontal polarization mode). In the TE mode, the direction of the main component of the electric field that constitutes the optical pulses Pb1 and Pb2 of the pulse train Pb is along the Y-axis direction, which coincides with the in-plane direction of the substrate 50.

[0038] The direction perpendicular to the in-plane direction of the substrate 50 coincides with the polarization direction in the TM (Transverse Magnetic) mode (longitudinal polarization mode). In the TM mode, the direction of the main component of the electric field that constitutes the optical pulses Pb1 and Pb2 is along the Z-axis direction, which is perpendicular to the in-plane direction of the substrate 50.

[0039] In one example, the inspection device 1 further includes a first polarization-maintaining fiber 11 optically connecting the optical output end 3b of the pulse forming unit 3 and the optical input end 5a of the optical integrated circuit 5. The first polarization-maintaining fiber 11 is, for example, a fiber that maintains the polarization direction of the pulse train Pb. The polarization direction of the pulse train Pb propagating through the first polarization-maintaining fiber 11 is maintained in a constant direction.

[0040] The first adjustment unit 4 is configured with a rotation mechanism that rotates the first polarization-maintaining fiber 11 around its axis. The axis of the first polarization-maintaining fiber 11 coincides with the propagation direction (X-axis direction) of the pulse train Pb. The first adjustment unit 4 is configured with, for example, a fiber rotator. The first adjustment unit 4 receives, for example, from the determination unit 9, an instruction regarding the angle by which to rotate the first polarization-maintaining fiber 11. Based on this instruction, the first adjustment unit 4 may rotate the first polarization-maintaining fiber 11 around its axis. For example, the first adjustment unit 4 rotates the first polarization-maintaining fiber 11 around its axis by a predetermined angle. In this case, the polarization directions of the optical pulses Pb1 and Pb2 rotate by the predetermined angle. The predetermined angle is, for example, greater than 0 degrees and less than 90 degrees.

[0041] In the example of FIG. 1, the polarization directions of the optical pulses Pb1 and Pb2 are at a first angle θ from the in-plane direction of the substrate 50 when they are output from the pulse forming unit 3. １The polarization directions of the optical pulses Pb1 and Pb2 are inclined at a first angle θ from the Y axis in a plane formed by the Y axis and the Z axis, for example. １ The first adjustment unit 4 adjusts the first polarization-maintaining fiber 11 around its axis at a first angle θ １ , the polarization mode of the optical pulses Pb1 and Pb2 is converted to the TE mode. As a result, the polarization direction of the optical pulses Pb1 and Pb2 input to the optical integrated circuit 5 coincides with the in-plane direction of the substrate 50.

[0042] FIG. 4 is a plan view schematically illustrating an optical integrated circuit 5. The optical integrated circuit 5 is a device that integrates various optical functions and transmits information signals by light. The optical integrated circuit 5 includes a substrate 50. As shown in FIG. 4, the optical integrated circuit 5 is configured to include various elements, and these elements are mounted on the substrate 50. In the example shown in FIG. 4, the optical integrated circuit 5 includes an optical input section 51, an optical output section 52, a waveguide 53, photodiodes 54 and 57, an optical ring resonator circuit 55, a photonic crystal 56, and the like.

[0043] The optical input section 51 is a section where light is input from the first adjustment section 4 through the first polarization-maintaining fiber 11 into the optical integrated circuit 5. The optical input section 51 is formed at the optical input end 5a, which is one side of the optical integrated circuit 5. Although only one optical input section 51 is shown in Fig. 4, a plurality of optical input sections 51 are formed at the optical input end 5a. For example, in a practical state after testing, an optical cable is connected to the optical input section 51.

[0044] The optical output unit 52 is a part that outputs the pulse train Pb from the optical integrated circuit 5 toward the second adjustment unit 6. The optical output unit 52 is formed at the optical output end 5b, which is the other end face of the optical integrated circuit 5. Although a single optical output unit 52 is shown in FIG. 4, a plurality of optical output units 52 may be formed at the optical output end 5b. For example, in a practical state after inspection, an optical cable may be connected to each of the plurality of optical output units 52. The optical output unit 52 may output light guided by the waveguide 53 to the outside of the optical integrated circuit 5.

[0045] The photodiodes 54 and 57 are photodetectors provided inside the optical integrated circuit 5. The optical ring resonator circuit 55 is a circuit that sets the resonant wavelength. The photonic crystal 56 has a structure in which the refractive index changes periodically, and is a structure that confines light in a small region and manipulates the light. The optical integrated circuit 5 may include various elements in addition to the components (elements) described above. For example, the optical integrated circuit 5 may have a microheater (not shown) mounted on each element, such as the optical ring resonator circuit 55 or the optical modulator. The microheater changes the refractive index of the waveguide that constitutes each element, such as the optical ring resonator circuit 55 or the optical modulator, thereby controlling the resonant wavelength of the guided light.

[0046] 4, the waveguide 53 is provided linearly from the optical input portion 51 to the optical output portion 52, but the waveguide 53 may be formed, for example, in a curved or spiral shape. When the waveguide 53 is curved or spiral, the direction of curvature of the waveguide 53 coincides with the in-plane direction of the substrate 50. The direction of curvature of the waveguide 53 refers, for example, to the direction in which the waveguide 53 is curved, that is, the direction in which the curve of the waveguide 53 is described.

[0047] Chromatic dispersion occurs in the pulse train Pb input to the optical integrated circuit 5 as it propagates through the waveguide 53. Since the optical pulses Pb1 and Pb2 have different wavelengths, they propagate at different speeds through the waveguide 53. This causes a difference in the propagation time between the optical pulse Pb1 and the optical pulse Pb2, and the time interval between the optical pulse Pb1 and the optical pulse Pb2 varies.

[0048] In this embodiment, the optical integrated circuit 5 is optically designed so that the polarization mode of the input optical pulses is maintained when they are output. However, the polarization mode of the optical pulses may change slightly while propagating through the waveguide 53. In this case, as illustrated in FIG. 1 , the polarization modes of the optical pulses Pb1 and Pb2 output from the optical integrated circuit 5 are slightly changed when they are rotated at a second angle θ from the in-plane direction of the substrate 50. ２ The second angle θ ２ is the first angle θ １ It may be smaller than

[0049] The likelihood of polarization mixing occurring in the optical pulses Pb1 and Pb2 output from the optical integrated circuit 5 varies depending on the polarization direction of the optical pulses Pb1 and Pb2 when they are input to the optical integrated circuit 5. Polarization mixing is a phenomenon in which the TE mode and TM mode are mixed in the polarization modes of the optical pulses Pb1 and Pb2. In this case, when the ratio of the TE mode component to the TM mode component is expressed as a percentage, the closer the ratio is to 50%, the greater the influence of polarization mixing. If the influence of polarization mixing is large, errors in repeated measurements of the amount of chromatic dispersion may be more likely to occur in the determination unit described below.

[0050] The magnitude of the effect of polarization mixing varies depending on the polarization direction of the optical pulses Pb1 and Pb2 input to the optical integrated circuit 5. When the polarization direction of the optical pulses Pb1 and Pb2 input to the optical integrated circuit 5 is perpendicular to the in-plane direction of the substrate 50 (when input in TM mode), the ratio of TE mode mixed into TM mode in the optical pulses Pb1 and Pb2 output from the optical integrated circuit 5 is about several percent. In contrast, when the polarization direction of the optical pulses Pb1 and Pb2 input to the optical integrated circuit 5 coincides with the in-plane direction of the substrate 50 (when input in TE mode), the ratio of TM mode mixed into TE mode in the optical pulses Pb1 and Pb2 output from the optical integrated circuit 5 increases to about 20%. In other words, when input in TE mode, the effect of polarization mixing is greater than when input in TM mode.

[0051] The reason why the influence of polarization mixing is greater when input in TE mode than when input in TM mode is presumably due to structural factors of the waveguide, but is not limited to this. FIG. 5 is a cross-sectional view showing an example of a cross section of a waveguide 53 of an optical integrated circuit 5. The waveguide 53 includes a core layer 531 and a cladding layer 532. The core layer 531 passes through the center of the waveguide 53 and propagates the pulse train Pb. The cladding layer 532 is formed to cover the periphery of the core layer 531. The core layer 531 includes a pair of side walls 531a, an upper wall 531b, and a lower wall 531c. The pair of side walls 531a of the core layer 531 are preferably formed parallel to the Z-axis direction, which is perpendicular to the in-plane direction of the substrate 50.

[0052] However, for example, the core layer 531 and the cladding layer 532 may be manufactured by laminating thin films in layers in the Z-axis direction. In this case, the pair of sidewalls 531a may be formed at a slight inclination relative to a direction parallel to the Z-axis direction. When the optical pulses Pb1 and Pb2 are input in TE mode, evanescent waves are likely to leak in the in-plane direction of the substrate 50. Because the in-plane direction of the substrate 50 intersects with the direction in which the pair of sidewalls 531a are formed, the TE mode is susceptible to the influence of the inclination of the pair of sidewalls 531a, and the influence of polarization mixing may be significant. That is, structural factors of the waveguide 53, such as the inclination of the pair of sidewalls 531a, affect the magnitude of polarization mixing, which in turn affects the error in repeated measurements of the chromatic dispersion amount. The inspection device 1 considers the chromatic dispersion amount as a feature value in the time waveform and uses the error in repeated measurements of the feature value calculated based on multiple feature values ​​to accurately determine the pass / fail of the optical integrated circuit 5 while taking the above-mentioned structural factors into consideration.

[0053] Referring again to FIG. 1 , the second adjustment unit 6 adjusts the polarization direction of the pulse train Pb that is output from the optical integrated circuit 5 and that has undergone chromatic dispersion. The second adjustment unit 6 adjusts the polarization direction of the pulse train Pb so that the TE mode components of each optical pulse Pb1, Pb2 of the pulse train Pb can be extracted. Although details will be described later, in this embodiment, the polarization direction of the correlation optical system 7 matches the polarization direction in TE mode, so the second adjustment unit 6 can also match the polarization direction of each optical pulse Pb1, Pb2 of the pulse train Pb that is input to the subsequent correlation optical system 7 with the polarization direction of the correlation optical system 7.

[0054] In one example, the inspection device 1 further includes a second polarization-maintaining fiber 12 that optically connects the optical output end 5b of the optical integrated circuit 5 and the optical input end 7a of the correlation optical system 7. The second polarization-maintaining fiber 12 is, for example, a polarization-maintaining fiber that maintains the polarization direction of the pulse train Pb, similar to the first polarization-maintaining fiber 11.

[0055] The second adjustment unit 6 is configured with a rotation mechanism that rotates the second polarization-maintaining fiber 12 around its axis. For example, the second adjustment unit 6 rotates the second polarization-maintaining fiber 12 around its axis so as to extract the TE mode components of the optical pulses Pb1 and Pb2 of the pulse train Pb. The axis of the second polarization-maintaining fiber 12 coincides with the propagation direction (X-axis direction) of the pulse train Pb. The second adjustment unit 6 is configured with, for example, a fiber rotator. The second adjustment unit 6 receives, for example, from the determination unit 9, instructions regarding the angle by which the second polarization-maintaining fiber 12 should be rotated. Based on this instruction, the second adjustment unit 6 may rotate the second polarization-maintaining fiber 12 around its axis. For example, the second adjustment unit 6 rotates the second polarization-maintaining fiber 12 around its axis by a predetermined angle. In this case, the polarization directions of the optical pulses Pb1 and Pb2 rotate by the predetermined angle.

[0056] In the example of FIG. 1, the polarization directions of the optical pulses Pb1 and Pb2 are, as described above, at the time when they are output from the optical integrated circuit 5, at the second angle θ from the in-plane direction of the substrate 50. ２ Therefore, the second adjustment unit 6 adjusts the polarization axis of the second polarization-maintaining fiber 12 so that it is parallel to the Y axis (for example, by rotating the second polarization-maintaining fiber 12 around its axis at a second angle θ ２ TE mode components of the optical pulses Pb1 and Pb2 are extracted by rotating the optical pulses Pb1 and Pb2 at the polarization direction (rotating the optical pulses Pb1 and Pb2 at the polarization direction) in the optical integrated circuit 5. As a result, the polarization direction of the optical pulses Pb1 and Pb2 output from the optical integrated circuit 5 coincides with the polarization direction of the correlation optical system 7.

[0057] The correlation optical system 7 generates correlation light including cross-correlated light or auto-correlated light of the pulse train Pb output from the optical integrated circuit 5 through the waveguide 53. In this embodiment, the correlation optical system 7 includes a lens 71, an optical element 72, and a lens 73. The lens 71 is provided on the optical path between the pulse forming unit 3 and the optical element 72, and focuses the pulse train Pb output from the second adjustment unit 6 onto the optical element 72.

[0058] The optical element 72 is a light emitter that includes at least one of a nonlinear optical crystal that generates second harmonic waves (SHG) and a phosphor. Examples of nonlinear optical crystals include KTP (KTiOPO4) crystal, LBO (LiB3O5) crystal, and BBO (β-BaB2O4) crystal. Examples of phosphors include coumarin, stilbene, and rhodamine. The optical element 72 receives a pulse train Pb and generates correlation light Pc that includes cross-correlation or auto-correlation of the pulse train Pb.

[0059] The optical element 72 has polarization dependence in a predetermined direction. For example, the optical element 72 is an element that has polarization dependence on the polarization direction of the optical pulses Pb1 and Pb2 of the pulse train Pb. In this embodiment, the polarization direction in which the optical element 72 has polarization dependence matches the polarization direction in TE mode. The second adjustment unit 6 inputs the pulse train Pb to the correlation optical system 7 so that the polarization direction of the optical pulses Pb1 and Pb2 matches the polarization direction in which the optical element 72 has polarization dependence, allowing the correlation optical system 7 to efficiently generate the correlated light Pc. The lens 71 collimates or focuses the correlated light Pc output from the optical element 72.

[0060] An example of the configuration of the correlation optical system 7 will now be described in detail. Fig. 6 is a diagram schematically illustrating an example of the configuration of the correlation optical system 7. This correlation optical system 7A has a beam splitter 74 as an optical branching component that branches the pulse train Pb into two. The beam splitter 74 is optically coupled to the second adjustment unit 6 shown in Fig. 1, and transmits a portion of the pulse train Pb input from the second adjustment unit 6 and reflects the remainder. The branching ratio of the beam splitter 74 is, for example, 1:1.

[0061] One pulse train Pba split by the beam splitter 74 passes through an optical path 7c including a plurality of mirrors 75 and reaches the lens 71. The other pulse train Pbb split by the beam splitter 74 passes through an optical path 7d including a plurality of mirrors 76 and reaches the lens 71. The optical lengths of the optical paths 7c and 7d are different. Therefore, the plurality of mirrors 75 and the plurality of mirrors 76 form a delay optical system that imparts a time difference between the one pulse train Pba and the other pulse train Pbb split by the beam splitter 74. Furthermore, at least a portion of the plurality of mirrors 76 are mounted on a moving stage 77, making the optical length of the optical path 7d variable. Therefore, with this configuration, the time difference between the pulse trains Pba and Pbb can be varied.

[0062] In this example, the optical element 72 includes a nonlinear optical crystal. The lens 71 focuses each of the pulse trains Pba and Pbb toward the optical element 72, and causes the optical axes of the pulse trains Pba and Pbb to intersect with each other at a predetermined angle in the optical element 72. As a result, in the optical element 72, which is a nonlinear optical crystal, a second harmonic is generated starting from the intersection of the pulse trains Pba and Pbb. This second harmonic is correlated light Pc, which includes the autocorrelation of the pulse train Pb. This correlated light Pc is collimated or focused by the lens 73, and then input to the photodetector 8.

[0063] 7 is a diagram schematically illustrating another exemplary configuration of the correlation optical system 7B. In this correlation optical system 7B, the pulse train Pb passes through an optical path 7e to reach the lens 71, and the reference light pulse Pr, which is a single pulse, passes through an optical path 7f to reach the lens 71. The optical path 7f includes multiple mirrors 78 and is bent in a U-shape. Furthermore, at least some of the multiple mirrors 78 are mounted on a moving stage 79, making the optical length of the optical path 7f variable. Therefore, with this configuration, the time difference between the pulse train Pb and the reference light pulse Pr (the difference in timing at which they reach the lens 71) can be varied.

[0064] In this example, the optical element 72 also includes a nonlinear optical crystal. The lens 71 focuses the pulse train Pb and the reference optical pulse Pr toward the optical element 72, and causes the optical axis of the pulse train Pb and the optical axis of the reference optical pulse Pr to intersect with each other at a predetermined angle in the optical element 72. As a result, in the optical element 72, which is a nonlinear optical crystal, a second harmonic is generated starting from the intersection of the pulse train Pb and the reference optical pulse Pr. This second harmonic is correlated light Pc, which includes the cross-correlation of the pulse train Pb. This correlated light Pc is collimated or focused by the lens 73, and then input to the photodetector 8.

[0065] 8 is a diagram conceptually illustrating the characteristic quantities of the correlated light Pc. FIG. 8( a) shows an example of the time waveform of the correlated light Pc when no chromatic dispersion occurs in the pulse train Pb in the waveguide 53 of the optical integrated circuit 5 (chromatic dispersion is zero). FIG. 8( b) shows an example of the time waveform of the correlated light Pc when chromatic dispersion occurs in the pulse train Pb in the waveguide 53 of the optical integrated circuit 5 (chromatic dispersion is not zero). In the example of FIG. 8, the correlated light Pc includes two optical pulses Pc1 and Pc2 corresponding to the optical pulses Pb1 and Pb2, respectively. Here, the peak intensity of the optical pulse Pc1 is denoted as PE1, and the peak intensity of the optical pulse Pc2 is denoted as PE2. The full width at half maximum (FWHM) of the optical pulse Pc1 is denoted as W1, and the full width at half maximum (FWHM) of the optical pulse Pc2 is denoted as W2. The peak time interval (pulse interval) between the peaks of the optical pulses Pc1 and Pc2 is denoted as G.

[0066] 8(a) and 8(b), when chromatic dispersion occurs in the pulse train Pb, the peak intensities PE1 and PE2 of the optical pulses Pc1 and Pc2 are lower, and the full widths at half maximum W1 and W2 of the optical pulses Pc1 and Pc2 are larger, compared to when chromatic dispersion does not occur. Furthermore, the peak time interval G is longer.

[0067] Thus, when chromatic dispersion occurs in the pulse train Pb, the feature quantities (peak intensities PE1, PE2, full widths at half maximum W1, W2, and peak time interval G) of the time waveform of the correlated light Pc change significantly compared to when chromatic dispersion does not occur. The amount of change depends on the amount of chromatic dispersion of the pulse train Pb in the waveguide 53. Therefore, there is a correlation between the feature quantities of the time waveform of the correlated light Pc and the amount of chromatic dispersion. Based on this correlation, the inspection device 1 disclosed herein regards the amount of chromatic dispersion as a feature quantity in the time waveform and judges the quality of the optical integrated circuit 5.

[0068] Referring again to FIG. 1 , the light detection unit 8 receives the correlated light Pc output from the correlation optical system 7 and detects the time waveform of the correlated light Pc. The light detection unit 8 is configured to include a photodetector such as a photodiode. The light detection unit 8 detects the time waveform of the correlated light Pc by converting the intensity of the correlated light Pc into an electrical signal. The electrical signal that is the detection result is provided to the determination unit 9.

[0069] The determination unit 9 determines the quality of the optical integrated circuit 5 based on the feature quantities of the time waveform of the correlated light Pc provided by the light detection unit 8. As described above, when the correlated light Pc including the cross-correlation or autocorrelation of the pulse train Pb is generated, various feature quantities (e.g., pulse interval, peak intensity, pulse width, etc.) in the time waveform of the correlated light Pc have a significant correlation with the amount of chromatic dispersion of the object to be measured. Therefore, the determination unit 9 can accurately evaluate the amount of chromatic dispersion of the pulse train Pb in the optical integrated circuit 5 that is the object to be measured by evaluating the feature quantities of the time waveform of the correlated light Pc.

[0070] The time waveform of the correlated light Pc will be described in more detail. Fig. 9 is a graph showing an example of the time waveform of the correlated light Pc. The horizontal axis of the graph of the time waveform of the correlated light Pc represents the delay time, and the vertical axis of the graph of the time waveform of the correlated light Pc represents the intensity of the correlated light Pc.

[0071] In the correlation optical system 7A shown in FIG. 6 , the pulse train Pb is separated into pulse trains Pba and Pbb by a beam splitter 74. The pulse trains Pba and Pbb are then input to an optical element 72. The optical element 72 generates correlated light Pc, which is a second harmonic, and the correlated light Pc is input to a photodetector 8. The photodetector 8 detects the intensity of the correlated light Pc. The intensity of the correlated light Pc varies depending on the time difference between the pulse trains Pba and Pbb—in other words, the delay time of the pulse train Pba relative to the pulse train Pbb, or the delay time of the pulse train Pbb relative to the pulse train Pba. The delay time is adjusted by a moving stage 77. As shown in FIGS. 10( a ) to 10 ( d ), delay times T1, T2, and T3 are the times from the midpoint of two pulses included in the pulse train Pba to the midpoint of two pulses included in the pulse train Pbb.

[0072] The light detection unit 8 may include, for example, a control unit in addition to the photodetector, and the delay time may be changed by gradually changing the position of the moving stage 77 using the control unit. The control unit may also include, for example, a counter, and measure the delay time each time the pulse train Pb is input. The control unit may then generate a time waveform of the correlated light Pc by plotting the intensity of the electrical signal of the correlated light Pc detected by the photodetector for each measured delay time.

[0073] The example of Figure 9 illustrates the intensity of the correlated light Pc when the delay times are T1, T2, 0, and T3. As shown in Figure 10(a), delay time T1 is the delay time of pulse train Pba relative to pulse train Pbb. At delay time T1, pulse train Pba lags pulse train Pbb, and pulse train Pbb and pulse train Pba do not overlap at all. As shown in Figure 10(b), delay time T2 is the delay time of pulse train Pba relative to pulse train Pbb. At delay time T2, pulse train Pba lags pulse train Pbb, and part of pulse train Pbb and part of pulse train Pba overlap. When delay time is 0, as shown in Figure 10(c), there is no time difference between pulse train Pba and pulse train Pbb, and in this case, pulse train Pbb and pulse train Pba completely overlap. As shown in FIG. 10(d), delay time T3 is the delay time of pulse train Pbb relative to pulse train Pba, and during delay time T3, pulse train Pbb lags behind pulse train Pba, and part of pulse train Pbb overlaps part of pulse train Pba.

[0074] 9 and 10 , the smaller the time difference between the pulse trains Pba and Pbb and the larger the overlapping area between the pulse trains Pbb and Pba, the stronger the optical intensity of the correlated light Pc. The time waveform of the correlated light Pc includes a first peak PK1, a second peak PK2, and a third peak PK3. The first peak PK1 occurs when the delay time is zero. The second peak PK2 occurs when the pulse train Pba lags behind the pulse train Pbb, and one of the two pulses in the pulse train Pbb completely overlaps with one of the two pulses in the pulse train Pba. The third peak PK3 occurs when the pulse train Pbb lags behind the pulse train Pba, and the other of the two pulses in the pulse train Pbb completely overlaps with the other of the two pulses in the pulse train Pba.

[0075] The light detection unit 8 detects, for example, the interval between the first peak PK1 and the second peak PK2 or the interval between the first peak PK1 and the third peak PK3 as a peak time interval G, which is a type of feature quantity of the time waveform. The light detection unit 8 may provide the detected peak time interval G to the determination unit 9.

[0076] In the above example, the controller included in the light detection unit 8 generates the time waveform of the correlated light Pc. However, the determination unit 9 may generate the time waveform of the correlated light Pc. That is, the determination unit 9 may function as the light detection unit. In this case, the light detection unit 8 may include a photodetector and output an electrical signal of the detected correlated light Pc to the determination unit 9. Alternatively, the determination unit 9 may include a photodetector instead of the light detection unit 8. In this case, the correlated light Pc may be input to the determination unit 9 from the correlation optical system 7A, and the photodetector of the determination unit 9 may detect the time waveform of the correlated light Pc and convert the intensity of the correlated light Pc into an electrical signal. The determination unit 9 may generate the time waveform of the correlated light Pc by plotting the intensity of the electrical signal of the correlated light Pc output from the light detection unit 8 or the electrical signal converted by the photodetector of the determination unit 9 for each measured delay time. The determination unit 9 may change the delay time by gradually changing the position of the moving stage 77.

[0077] In the above example, the correlation optical system 7A has been described as an example, but the above example may also be applied to the correlation optical system 7B. In this case, the light detection unit 8 or the determination unit 9 may change the delay time of the pulse train Pb with respect to the reference light pulse Pr, or the delay time of the reference light pulse Pr with respect to the pulse train Pb, by gradually changing the position of the moving stage 79. The light detection unit 8 or the determination unit 9 may then generate the time waveform of the correlation light Pc by plotting the intensity of the electrical signal of the correlation light Pc for each measured delay time. Even when applied to the correlation optical system 7B, the determination unit 9 may function as a light detection unit, as in the case when applied to the correlation optical system 7A.

[0078] Fig. 11 is a diagram schematically illustrating an example of the hardware configuration of the determination unit 9. As illustrated in Fig. 11, the determination unit 9 can be physically configured as a typical computer including a processor (CPU) 91, main storage devices such as a ROM 92 and a RAM 93, input devices 94 such as a keyboard, mouse, and touch screen, output devices 95 such as a display (including a touch screen), a communication module 96 such as a network card for transmitting and receiving data to and from other devices, an auxiliary storage device 97 such as a hard disk, etc.

[0079] The processor 91 of the computer can realize the function of the determination unit 9 by means of a chromatic dispersion amount calculation program. In other words, the chromatic dispersion amount calculation program causes the processor 91 of the computer to operate as the determination unit 9. The chromatic dispersion amount calculation program is stored in a storage device (storage medium) inside or outside the computer, such as the auxiliary storage device 97. The storage device may be a non-transitory recording medium. Examples of the recording medium include recording media such as a flexible disk, CD, DVD, recording media such as ROM, semiconductor memory, cloud server, etc.

[0080] [Inspection Method Using Inspection Apparatus] Fig. 12 is a flowchart showing an inspection method using the inspection apparatus 1 having the above configuration. In this method, first, in a pulse forming step S1, a pulse train Pb is generated in which multiple optical pulses Pb1, Pb2 having different center wavelengths are arranged at a predetermined time interval. For example, multiple wavelength components contained in the measured optical pulse Pa are spatially separated by wavelength, and the multiple wavelength components are intensity-modulated and phase-modulated using a spatial light modulator 33, and then the multiple wavelength components are focused. This makes it possible to easily generate the pulse train Pb.

[0081] Subsequently, in a first adjustment step S2, the polarization directions of the optical pulses Pb1, Pb2 of the pulse train Pb input to the optical integrated circuit 5 are aligned with the in-plane direction of the substrate 50. For example, a rotation mechanism is used to rotate the first polarization-maintaining fiber 11 around its axis, which optically connects the pulse forming unit 3 and the optical integrated circuit 5. The first polarization-maintaining fiber 11 is rotated so that the polarization directions of the optical pulses Pb1, Pb2 are aligned with the in-plane direction of the substrate 50.

[0082] Subsequently, in a propagation step S3, the pulse train Pb is propagated in the waveguide 53 of the optical integrated circuit 5. The optical pulses Pb1 and Pb2, whose polarization directions have been rotated in the first adjustment step S2, are input from the optical input unit 51, guided by the waveguide 53, and then output from the optical output unit 52. In the propagation step S3, chromatic dispersion occurs in the pulse train Pb as it propagates through the waveguide 53. For example, a difference occurs between the propagation time of the optical pulse Pb1 and the propagation time of the optical pulse Pb2, and the time interval between the optical pulse Pb1 and the optical pulse Pb2 varies.

[0083] Subsequently, in a second adjustment step S4, the TE mode components of the optical pulses Pb1 and Pb2 of the pulse train Pb output from the optical integrated circuit 5 are extracted. For example, a rotation mechanism is used to rotate the second polarization-maintaining fiber 12 optically connecting the optical integrated circuit 5 and the correlation optical system 7 around its axis so as to extract the TE mode components of the optical pulses Pb1 and Pb2 of the pulse train Pb output from the optical integrated circuit 5. This also makes it possible to rotate the second polarization-maintaining fiber 12 so that the polarization direction of the optical pulses Pb1 and Pb2 coincides with the polarization direction of the correlation optical system 7. Furthermore, the rotation mechanism may be used to adjust the polarization axis of the second polarization-maintaining fiber 12 so that the TE mode components of the optical pulses Pb1 and Pb2 of the pulse train Pb are extracted. This makes it possible to selectively extract the TE mode components of the optical pulses Pb1 and Pb2 of the pulse train Pb.

[0084] Subsequently, in correlation light generating step S5, correlation light Pc including cross-correlation or auto-correlation of the pulse train Pb is generated using an optical element 72 including at least one of a nonlinear optical crystal and a phosphor. For example, as shown in Fig. 6, the pulse train Pb is branched into two, one branched pulse train Pbb is time-delayed with respect to the other pulse train Pba, and correlation light Pc including auto-correlation of the pulse train Pb is generated based on the time-delayed one pulse train Pbb and the other pulse train Pba.

[0085] Subsequently, the time waveform of the correlated light Pc is detected in the light detection step S6. The time waveform of the correlated light Pc includes a feature amount.

[0086] Subsequently, in a repeating step S7, the pulse forming step S1 to the light detecting step S6 are repeated multiple times. In the repeating step S7, for example, the light detecting unit 8 detects the time waveform multiple times. The number of repetitions may be, for example, two or more times and may be ten or less times. Each of the time waveforms detected multiple times may include a different feature amount.

[0087] Next, in a determination step S8, the quality of the optical integrated circuit 5 is determined based on a plurality of feature quantities in the time waveform detected a plurality of times. Here, the quality is determined by comparing a predetermined tolerance with the plurality of feature quantities. FIG. 13 is a flowchart showing an example of a method for setting the tolerance. The tolerance may be set at any timing in the flowchart showing the inspection method shown in FIG. 12. The tolerance may be set before the pulse forming step S1 or before the determination step S8.

[0088] First, in an ideal characteristic acquisition step S11, an ideal characteristic indicating the relationship between the feature amount and the waveguide length is acquired. In the following description, the peak time interval G (pulse interval) between the peak of the optical pulse Pc1 and the peak of the optical pulse Pc2 will be described as an example of the feature amount. FIG. 14 is a diagram showing an example of the ideal characteristic GR and the allowable range MG. In the ideal characteristic GR, the pulse interval tends to decrease as the length of the waveguide 53 increases. The length of the waveguide 53 and the pulse interval exhibit a linear relationship.

[0089] The ideal characteristic GR varies depending on the cross-sectional area, structure, or material of the core layer 531 and clad layer 532 that constitute the waveguide 53. Referring again to FIG. 5 , for example, the ideal characteristic GR changes when the height h or width w of the core layer 531 is changed. Alternatively, for example, the ideal characteristic GR changes when the cross-sectional shape of the core layer 531 is changed from a rectangular shape to a trapezoidal shape.

[0090] In the ideal characteristic acquisition step S11, the ideal characteristic GR may be acquired for each cross-sectional area, structure, or material of the core layer 531 and the cladding layer 532. In the ideal characteristic acquisition step S11, for example, the ideal characteristic GR may be statistically derived from characteristics indicating the relationship between previously measured feature quantities and the waveguide length. Alternatively, in the ideal characteristic acquisition step S11, for example, the ideal characteristic GR may be theoretically derived using techniques such as numerical analysis and machine learning based on data regarding the cross-sectional area, structure, or material of the core layer 531 and the cladding layer 532.

[0091] Next, in an allowable range setting step S12, an allowable range MG is set as a comparison target with the multiple feature quantities. As shown in FIG. 14 , the allowable range MG may be set based on the ideal characteristic GR. For example, the allowable range MG may be set as a predetermined numerical range based on the ideal characteristic GR. The predetermined numerical range may be, for example, ±5%. The predetermined numerical range may be, for example, ±2% to ±10%. Alternatively, the allowable range MG may be set by the user to any range based on the ideal characteristic GR based on the time waveform detected multiple times in the repeat step S7. The set allowable range MG is stored, for example, in the auxiliary storage device 97 (storage unit).

[0092] 12 again, in judgment step S8, the set allowable range MG is compared with the plurality of feature quantities (pulse intervals) to judge the acceptability of the optical integrated circuit 5. In judgment step S8, for example, the judgment unit 9 reads the data of the allowable range MG from the auxiliary storage device 97 and judges the acceptability of the optical integrated circuit 5.

[0093] Fig. 15 is a graph showing an example of determining whether an optical integrated circuit 5 is good or bad. The example of Fig. 15 shows pulse interval values ​​measured when the lengths of the waveguide 53 are set to 20 mm, 40 mm, 60 mm, and 100 mm in a waveguide 53 in which the cross-sectional area, structure, or material of the core layer 531 and the cladding layer 532 are the same. In the example of Fig. 15, the pulse interval is measured three times for each waveguide length, and the measured pulse interval value is compared with the allowable range MG.

[0094] In the example of Figure 15, if any one of the three pulse interval values ​​falls outside the allowable range MG, the optical integrated circuit 5 to be measured may be judged as "failed." On the other hand, if all three pulse interval values ​​are within the allowable range MG, the optical integrated circuit 5 to be measured may be judged as "good." For example, if the waveguide length is 20 mm, all three pulse interval values ​​fall within the allowable range MG, so an optical integrated circuit 5 with a waveguide length of 20 mm is judged as "good." For example, if the waveguide length is 40 mm, one of the three pulse interval values ​​falls outside the allowable range MG, so an optical integrated circuit 5 with a waveguide length of 40 mm is judged as "failed."

[0095] Fig. 16 is a graph showing another example of pass / fail judgment of the optical integrated circuit 5. The example of Fig. 16 differs from the example of Fig. 15 in that the amount of measurement error of the values ​​of three pulse intervals is calculated and the amount of measurement error is compared with the allowable range MG. The amount of measurement error is a value calculated by statistically processing the values ​​of the three pulse intervals, such as a standard deviation.

[0096] In the example of Figure 16, if the amount of measurement error falls outside the allowable range MG, the optical integrated circuit 5 to be measured may be judged as "defective." On the other hand, if the amount of measurement error falls within the allowable range MG, the optical integrated circuit 5 to be measured may be judged as "good." For example, if the waveguide length is 20 mm, the amount of measurement error falls within the allowable range MG, so the optical integrated circuit 5 with a waveguide length of 20 mm is judged as "good." For example, if the waveguide length is 40 mm, the amount of measurement error falls outside the allowable range MG, so the optical integrated circuit 5 with a waveguide length of 40 mm is judged as "defective."

[0097] [Operations and Effects] As described above, in the inspection device 1, the pulse train Pb is input to the optical integrated circuit 5 so that the polarization direction of each of the multiple optical pulses Pb1 and Pb2 is aligned with the in-plane direction of the substrate 50. As a result, the polarization directions of the multiple optical pulses Pb1 and Pb2 intersect with structures such as the sidewalls 531a that constitute the waveguide 53, and if there is an abnormality in the formation state of these structures, the error in repeated measurements of the chromatic dispersion amount will increase. In this inspection device 1, the chromatic dispersion amount is regarded as a feature amount in the time waveform. By using the error in repeated measurements of feature amounts calculated based on multiple feature amounts, it is possible to improve the accuracy of determining whether the optical integrated circuit 5 is good or bad, while taking structural factors into consideration.

[0098] On the other hand, since repeated measurement errors of the feature quantities are large, for example, if the quality of the optical integrated circuit 5 is judged based only on the feature quantities in a time waveform detected once, an erroneous judgment may be made. In contrast, in the inspection device 1, the judgment unit 9 judges the quality of the optical integrated circuit 5 based on multiple feature quantities in time waveforms detected multiple times. This prevents erroneous judgments of the quality of the optical integrated circuit 5, and enables accurate quality judgments to be made.

[0099] Furthermore, in the inspection device 1, the pulse train Pb is input to the optical integrated circuit 5 so that the polarization directions of the multiple optical pulses Pb1 and Pb2 are aligned with the in-plane direction of the substrate 50. This reduces the influence of at least one of the upper wall 531b and the lower wall 531c of the core layer 531 on the amount of chromatic dispersion. The influence of at least one of the upper wall 531b and the lower wall 531c on the amount of chromatic dispersion is, for example, due to the surface roughness of at least one of the upper wall 531b and the lower wall 531c. Alternatively, the amount of chromatic dispersion may be affected even if at least one of the upper wall 531b and the lower wall 531c is coated with a special coating material, or if at least one of the upper wall 531b and the lower wall 531c is not covered by the cladding layer 532 and is exposed to air. Because the in-plane direction of the substrate 50 is parallel to the direction in which the upper wall 531b and the lower wall 531c are formed, the pulse train Pb is less likely to be influenced by the upper wall 531b and the lower wall 531c. This allows the quality of the optical integrated circuit 5 to be judged while accurately taking into consideration the inclination of the pair of side walls 531a.

[0100] The determination unit 9 may determine the quality of the optical integrated circuit 5 based on whether the plurality of feature quantities are within a predetermined tolerance range MG. In this case, by comparing the plurality of feature quantities with the tolerance range, the accuracy of the quality determination can be further improved.

[0101] The determining unit 9 may determine the quality of the optical integrated circuit 5 based on whether the amount of measurement error calculated based on the plurality of feature quantities is within a predetermined tolerance range MG. In this case, the accuracy of determining the quality of the optical integrated circuit 5 can be further improved based on the statistical trends of the plurality of feature quantities.

[0102] The inspection device 1 may further have a storage unit (auxiliary storage device 97) that stores data on the tolerance range MG, and the judgment unit 9 may read the data on the tolerance range MG from the storage unit to judge the acceptability of the optical integrated circuit 5. In this case, having a storage unit that stores the tolerance range data in advance can reduce the time required to read the data, thereby shortening the time required to judge the acceptability.

[0103] The inspection device 1 may further include a first polarization-maintaining fiber 11 that optically connects the pulse forming unit 3 to the optical integrated circuit 5, and the first adjustment unit 4 may be configured with a rotation mechanism that rotates the first polarization-maintaining fiber 11 about its axis. In this case, by rotating the first polarization-maintaining fiber 11 about its axis, the polarization directions of the optical pulses Pb1, Pb2 of the pulse train Pb input to the optical integrated circuit 5 can be easily adjusted.

[0104] The inspection device 1 may further include a second adjustment unit 6 that extracts TE mode components of each of the optical pulses Pb1, Pb2 of the pulse train Pb emitted from the optical integrated circuit 5. In this case, the second adjustment unit 6 inputs the pulse train Pb to the correlation optical system 7 so that the polarization direction of the pulse train Pb (for example, a direction that coincides with the TE mode) coincides with the polarization direction in which the correlation optical system 7 has polarization dependency. Therefore, the correlation light Pc can be efficiently generated in the correlation optical system 7.

[0105] The inspection device 1 may further include a second polarization-maintaining fiber 12 that optically connects the optical integrated circuit 5 to the correlation optical system 7, and the second adjustment unit 6 may be configured with a rotation mechanism that rotates the second polarization-maintaining fiber 12 about its axis. In this case, by rotating the in-plane direction of the substrate 50 of the optical integrated circuit 5 about the axis of the second polarization-maintaining fiber 12, the polarization mode of each of the optical pulses Pb1, Pb2 of the pulse train Pb input to the second polarization-maintaining fiber 12 can be easily selected.

[0106] The rotation mechanism that rotates the second polarization-maintaining fiber 12 may rotate the second polarization-maintaining fiber 12 about its axis so as to extract the TE mode component of the pulse train Pb output from the optical integrated circuit 5. In this case, it is possible to prevent the pulse train Pb from being input to the correlation optical system 7 in a polarization direction that is undesirable for measurement (for example, a polarization direction deviating from the polarization direction of the correlation optical system 7).

[0107] In the inspection method described above, the pulse train Pb is input to the optical integrated circuit 5 so that the polarization direction of each of the multiple optical pulses Pb1 and Pb2 is aligned with the in-plane direction of the substrate 50. As a result, the polarization directions of the multiple optical pulses Pb1 and Pb2 intersect with structures such as the sidewalls 531a that constitute the waveguide 53, and if there is an abnormality in the formation state of these structures, the error in repeated measurements of the chromatic dispersion amount will increase. In this inspection method, the chromatic dispersion amount is regarded as a feature amount in the time waveform. By using the error in repeated measurements of feature amounts calculated based on multiple feature amounts, it is possible to improve the accuracy of determining whether the optical integrated circuit 5 is good or bad, while taking structural factors into consideration.

[0108] In the judgment step S8, the quality of the optical integrated circuit 5 may be judged based on whether or not the plurality of feature quantities are within a predetermined tolerance range MG. In this case, by comparing the plurality of feature quantities with the tolerance range, the accuracy of the quality judgment can be further improved.

[0109] In the judgment step S8, the quality of the optical integrated circuit 5 may be judged based on whether the measurement error amounts for the plurality of feature quantities are within a predetermined tolerance range MG. In this case, the accuracy of the quality judgment of the optical integrated circuit 5 can be further improved based on the statistical trends of the plurality of feature quantities.

[0110] [Modifications] Although the embodiments of the present disclosure have been described above, the present disclosure is not necessarily limited to the above-described embodiments, and various modifications are possible without departing from the spirit of the present disclosure.

[0111] In the tolerance setting step S12, the tolerance MG set may be stored as a list according to the cross-sectional area, structure, or material of the core layer 531 and the cladding layer 532. For example, the tolerance MG is stored as a look-up table. The set tolerance MG may be stored in, for example, a storage device external to the inspection device 1.

[0112] If the polarization directions of the optical pulses Pb1 and Pb2 coincide with the in-plane direction of the substrate 50 when they are output from the pulse forming unit 3, the first adjusting unit 4 may maintain the polarization directions without rotating the first polarization-maintaining fiber 11. In this case, the first adjusting step S2 may be unnecessary.

[0113] If the polarization directions of the optical pulses Pb1 and Pb2 coincide with the in-plane direction of the substrate 50 when they are output from the optical integrated circuit 5, the second adjustment unit 6 may maintain the polarization directions without rotating the second polarization-maintaining fiber 12. Furthermore, if the polarization axis of the second polarization-maintaining fiber 12 coincides with the in-plane direction of the substrate 50, the second adjustment unit 6 may maintain the polarization direction without rotating the second polarization-maintaining fiber 12. In these cases, the second adjustment step S4 may be unnecessary.

[0114] In the decision step S8, the average value of the three pulse interval values ​​may be calculated and compared with the allowable range MG.

[0115] 1...inspection device, 3...pulse forming unit, 4...first adjustment unit, 6...second adjustment unit, 5...optical integrated circuit, 7, 7A, 7B...correlation optical system, 8...light detection unit, 9...judgment unit, 11...first polarization-maintaining fiber, 12...second polarization-maintaining fiber, 33...spatial light modulator, 50...substrate, 53...waveguide, 97...auxiliary memory device (memory unit), MG...tolerance range, Pb, Pba, Pbb...pulse train, Pb1, Pb2, Pc1, Pc2...optical pulse, Pc...correlated light, S1...pulse forming step, S2...first adjustment step, S4...second adjustment step, S5...correlated light generation step, S6...light detection step, S8...judgment step.

Claims

1. An inspection device for inspecting an optical integrated circuit having a waveguide provided on a substrate, comprising: a pulse forming unit that generates a pulse train in which a plurality of optical pulses having different center wavelengths are arranged at predetermined time intervals; a first adjustment unit that aligns the polarization direction of each optical pulse of the pulse train input to the optical integrated circuit with the in-plane direction of the substrate; a correlation optical system that has polarization dependency on the predetermined polarization direction and generates correlation light including cross-correlated light or auto-correlated light of the pulse train that is output from the optical integrated circuit through the waveguide; an optical detection unit that detects the time waveform of the correlation light; and a judgment unit that judges the quality of the optical integrated circuit based on the time waveform, wherein the optical detection unit detects the time waveform multiple times, and the judgment unit judges the quality of the optical integrated circuit based on multiple feature quantities in the time waveform detected multiple times.

2. The inspection device according to claim 1, wherein said determining section determines whether said optical integrated circuit is good or bad depending on whether said plurality of feature amounts are within a predetermined tolerance range.

3. An inspection device according to claim 1, wherein the judgment unit judges whether the optical integrated circuit is good or bad depending on whether the amount of measurement error calculated based on the plurality of feature quantities is within a predetermined allowable range.

4. An inspection device according to claim 2 or 3, further comprising a storage unit for storing data on the tolerance range, wherein the judgment unit reads out the data on the tolerance range from the storage unit and judges whether the optical integrated circuit is good or bad.

5. An inspection device according to any one of claims 1 to 4, further comprising a first polarization-maintaining fiber that optically connects the pulse forming section to the optical integrated circuit, and the first adjustment section is configured by a rotation mechanism that rotates the first polarization-maintaining fiber around its axis.

6. An inspection device according to any one of claims 1 to 5, further comprising a second adjustment section for extracting a TE mode component of the pulse train output from the optical integrated circuit.

7. An inspection device according to claim 6, further comprising a second polarization-maintaining fiber that optically connects the optical integrated circuit to the correlation optical system, and wherein the second adjustment unit is configured by a rotation mechanism that rotates the second polarization-maintaining fiber around its axis.

8. An inspection device according to claim 7, wherein said rotation mechanism rotates said second polarization-maintaining fiber around its axis so as to extract the TE mode component of said pulse train output from said optical integrated circuit.

9. An inspection method for inspecting an optical integrated circuit having a waveguide provided on a substrate, comprising: a pulse forming step of generating a pulse train in which a plurality of optical pulses having different center wavelengths are arranged at a predetermined time interval; a first adjustment step of orthogonalizing the polarization direction of each optical pulse of the pulse train input to the optical integrated circuit to be orthogonal to the in-plane direction of the substrate; a correlation light generation step of generating correlation light including cross-correlated light or auto-correlated light of the pulse train output from the optical integrated circuit using a correlation optical system having polarization dependency in the predetermined polarization direction; an optical detection step of detecting a time waveform of the correlation light; and a judgment step of judging the quality of the optical integrated circuit based on the time waveform, wherein the optical detection step detects the time waveform a plurality of times, and the judgment step judges the quality of the optical integrated circuit based on a plurality of feature quantities in the time waveform detected a plurality of times.

10. An inspection method according to claim 9, wherein said determining step determines whether said optical integrated circuit is good or bad depending on whether said plurality of feature quantities are within a predetermined tolerance range.

11. An inspection method according to claim 9, wherein in the judgment step, the quality of the optical integrated circuit is judged based on whether or not the amount of measurement error calculated based on the plurality of feature quantities is within a predetermined tolerance range.

Images