WO2024062790A1 - Optical measurement device - Google Patents

Abstract

Provided is an optical measurement device with which it is possible to solve at least one of the abovementioned problems.　A light source device 200 generates wavelength-swept light L1. An irradiation optical system 310 irradiates an object with the wavelength-swept light L1. A pulse light source 210 generates pulse light including a continuous spectrum. A splitter 222 spatially splits the pulse light into a plurality of beams, i.e., n beams (n≥2), according to wavelength. A plurality of fibers FB, i.e., n fibers FB, impart different delays to the n beams. A coupler 226 is a bundle fiber FB or a multi-core fiber that combines the n beams outputted from the n fibers FB. The incidence end of a rod integrator 228 is coupled to the emission end of the coupler 226. A critical illumination system 312 projects a light source image of the emission end of the rod integrator 228 onto an object 2.

Description

light measurement device

The present disclosure relates to an optical measurement device.

Spectroscopic analysis is widely used for component analysis and inspection of objects. In spectroscopic analysis, an object is irradiated with irradiation light, and the spectrum of the object light obtained as a result of the irradiation is measured. Based on the relationship between the spectrum of the object light and the spectrum of the irradiation light, optical properties such as reflection properties (wavelength dependence) or transmission properties can be obtained.

Wavelength sweep spectroscopy is known as one of the methods for measuring optical properties. A wavelength-sweeping spectrometer generates wavelength-swept light whose wavelength changes over time, and irradiates the object to be inspected. The wavelength swept light is a pulse or pulse train in which time and wavelength have a one-to-one relationship. Then, the wavelength-swept light is irradiated onto the inspection target, and the temporal waveform of the light obtained is detected by the light receiver. The output waveform of the optical receiver represents a spectrum whose time axis corresponds to wavelength.

FIG. 1 is a diagram showing a wavelength sweeping type spectrometer 10. The spectroscopic device 10 includes a light source device 20, a spectroscopic head 30, and an arithmetic processing device 40.

The light source device 20 generates wavelength swept light L1. The wavelength swept light L1 is guided to the spectroscopic head 30. The irradiation optical system 31 of the spectroscopic head 30 irradiates the sample 2 with the wavelength swept light L1. The first light receiver 32 detects light (object light) L2 obtained as a result of irradiating the sample 2 with the wavelength swept light L1. The object light L2 may be reflected light or transmitted light of the sample 2.

In the irradiation optical system 31, a portion of the wavelength swept light L1 is branched off as reference light L3. The second photodetector 33 measures the reference light L3.

The first detection signal S1 generated by the first light receiver 32 and the second detection signal S2 generated by the second light receiver 33 are supplied to the arithmetic processing device 40. The object light L2 and the reference light L3 inherit the one-to-one time-wavelength correspondence of the wavelength swept light L1. Therefore, the time waveform of the first detection signal S1 can be converted into the spectrum of the object light L2 by converting the time axis into a wavelength. Similarly, the time waveform of the second detection signal S2 can be converted into the spectrum of the reference light L3 by converting the time axis into a wavelength. The processing unit 40 calculates the ratio of each corresponding wavelength of the object light L2 to the reference light L3, and measures the spectral characteristics (reflectance and transmittance) of the sample 2.

FIG. 2 is a diagram showing a light source device 20 that generates wavelength swept light. The light source device 20 includes a pulse light source 21, a wavelength selection filter 22, a divider 23, a delay line 24, and a coupler 25.

The pulsed light source 21 generates pulsed light having a continuous spectrum. The wavelength selection filter 22 selects a wavelength band to be used for spectroscopy from among the spectral components included in the pulsed light.
The splitter 23 is an arrayed waveguide grating (AWG), and splits the pulsed light into a plurality of n paths according to the wavelength. The delay line 24 gives different delays to the light (divided light) on a plurality of paths. For example, the delay line 24 includes a plurality of fibers FB1 to FBn having different lengths. The coupler 25 is a multi-core fiber or a bundle fiber, and spatially recombines the lights output from the plurality of fibers FB1 to FBn. The recombined light is emitted as wavelength swept light L1.

Patent Document 2 discloses an optical system that irradiates an object with light emitted from a coupler 25, which is a multi-core fiber or a bundle fiber. Each beam emitted from a multi-core fiber or a bundle fiber has a Gaussian intensity distribution, and the irradiation optical system of Patent Document 2 irradiates the target object with the beam while maintaining the Gaussian intensity distribution.

Japanese Patent Application Publication No. 2020-159973 JP2022-42444A US Patent Application Publication No. 2017/0122806

As a result of studying the wavelength swept light light source device 20 shown in FIG. 2, the present inventors came to recognize the following problems.

(Assignment 1)
FIG. 3 is a diagram illustrating one of the problems (referred to as problem 1) of the wavelength sweeping type spectrometer.
The beam of the wavelength swept light L1 is a pulse, and is repeatedly irradiated to a predetermined position, and the object (sample) OBJ is transported by a transport means so as to cross the irradiation position of the wavelength swept light L1. FIG. 3 shows the relative positional relationship between the object OBJ and the beam of the wavelength swept light L1.

When the wavelength swept light L1 has a Gaussian-shaped illuminance distribution, the tail portion of the Gaussian distribution (for example, outside the 1/e ² width) has insufficient intensity and cannot be used for spectroscopic measurements. Therefore, when inspecting the entire object OBJ including the edges, as shown in the upper part of ^FIG . It protrudes outside of the object OBJ. If the protruding light becomes stray light and enters the light receiver, measurement accuracy will decrease. Furthermore, since the stray light does not pass through the object OBJ and is not attenuated, it has high intensity, and if it enters the light receiver, it may adversely affect the reliability of the light receiver.

As shown in the lower part of FIG. 3, stray light can be reduced by offsetting the relative positions of the beams irradiated to both ends of the object OBJ inward so that the tails of the beams are included in the object OBJ. However, in this case, since both ends of the object OBJ cannot be inspected, the measurable range becomes narrow. Furthermore, by integrating the spectra for the number of irradiations N, the S/N ratio is improved by a factor of √N. The number of pulses of the wavelength swept light L1 is reduced, and the S/N ratio is reduced.

(Assignment 2)
FIG. 4 is a diagram illustrating another problem (referred to as problem 2) of the wavelength sweeping type spectrometer. FIG. 4 shows the intensity distribution of a plurality of beams having different wavelengths on the surface (sample surface) of the object OBJ. The thickness d of the object OBJ may vary depending on the irradiation position. As shown in FIG. 4, when beams of wavelengths λ ₁ and λ ₂ are irradiated to different positions, the perceived thicknesses d ₁ and d ₂ will be different for each wavelength. Generally, the Beer-Lambert law shown below holds between the intensity _I0 of the incident light and the intensity I of the transmitted light.
-log(I/I ₀ )=ε・c・d
Here, -log(I/I ₀ ) is absorbance, ε is extinction coefficient, and d is optical path length. If the optical path length, that is, the thickness of the object is constant, the absorbance is proportional to the concentration, so by measuring the absorbance, the concentration of the target substance can be determined. However, as shown in FIG. 4, when the thickness of the object OBJ is not uniform and the beam is irradiated to different positions depending on the wavelength, the absorbance is not included in the object OBJ because the thickness of the object at each wavelength is different. This is affected not only by the absorption coefficient ε of the substance being used, but also by the optical path length d, which causes the problem that the absorbance spectrum cannot be measured correctly.

In order to solve this problem, it is necessary to disclose an irradiation optical system that irradiates multiple beams with different wavelengths to the same position, and a special optical system as disclosed in Patent Document 2 is required.

Note that these problems should not be taken as common knowledge among those skilled in the art, but were independently recognized by the present inventors.

The present disclosure has been made in such a situation, and one exemplary objective of a certain aspect thereof is to provide an optical measurement device that can solve at least one of the above-mentioned problems.

A certain aspect of the present disclosure relates to an optical measurement device. The optical measurement device includes a light source device that generates wavelength swept light and an optical system that irradiates a target object with the wavelength swept light. The light source device includes a pulsed light source that generates pulsed light including a continuous spectrum, a splitter that spatially splits the pulsed light into a plurality of n beams (n≧2) according to the wavelength, and a splitter that spatially splits the pulsed light into a plurality of n beams (n≧2) according to the wavelength. A plurality of n fibers that provide different delays, a coupler that is a bundle fiber or multi-core fiber that combines n beams output from the n fibers, and a rod whose input end is coupled to the output end of the coupler. Including an integrator. The optical system includes a critical illumination system that projects a light source image at the output end of the rod integrator onto the object.

Note that arbitrary combinations of the above components, and mutual substitution of components and expressions among methods, devices, systems, etc., are also effective as aspects of the present invention or the present disclosure. Furthermore, the description in this section (Means for Solving the Problems) does not describe all essential features of the present invention, and therefore, subcombinations of the described features may also constitute the present invention. .

According to a certain aspect of the present disclosure, at least one of the above-mentioned problems can be solved.

FIG. 1 is a diagram showing a wavelength sweep type spectrometer. FIG. 2 is a diagram showing a light source device that generates wavelength swept light. FIG. 2 is a diagram illustrating one of the problems of a wavelength sweeping type spectrometer. FIG. 3 is a diagram illustrating another problem with the wavelength sweep type spectrometer. FIG. 1 is a diagram showing an optical measurement device according to an embodiment. It is a perspective view showing a coupler and a rod integrator. FIG. 3 is a cross-sectional view of the coupler and rod integrator. It is a figure explaining profile conversion by a rod integrator. FIG. 3 is a diagram showing wavelength swept light. FIG. 6 is a diagram illustrating spectroscopy by the optical measurement device of FIG. 5; FIG. 3 is a diagram illustrating the intensity distribution of wavelength swept light irradiated onto the surface of a sample. It is a figure explaining the 1st effect by an optical measuring device. It is a figure explaining the 2nd effect by a light measuring device. FIG. 3 is a diagram illustrating a top-hat type beam profile. It is a figure showing the measurement result of the beam profile of irradiation light generated by the light measurement device concerning an embodiment. FIG. 3 is a diagram illustrating beam profile smoothing processing. FIG. 3 is a diagram showing profiles of a plurality of beams having different wavelengths. FIG. 3 is a diagram showing a beam profile obtained by a rod integrator with a hexagonal cross section. FIG. 3 is a cross-sectional view of the rod integrator. FIG. 3 is a diagram showing a beam profile obtained by a rod integrator with a circular cross section. FIG. 3 is a diagram showing a beam profile obtained by a rod integrator with a triangular cross section. FIG. 3 is a diagram showing a beam profile obtained by a rod integrator with a square cross section. FIG. 3 is a diagram showing a beam profile obtained by a rod integrator with a pentagonal cross section.

(Summary of embodiment)
1 provides an overview of some example embodiments of the present disclosure. This Summary is intended to provide a simplified description of some concepts of one or more embodiments in order to provide a basic understanding of the embodiments and as a prelude to the more detailed description that is presented later. It does not limit the size. Moreover, this summary is not an exhaustive overview of all possible embodiments, and is not limited to essential components of the embodiments. For convenience, "one embodiment" may be used to refer to one embodiment (example or modification) or multiple embodiments (examples or modifications) disclosed in this specification.

An optical measuring device according to one embodiment includes a light source device that generates wavelength swept light and an optical system that irradiates a target object with the wavelength swept light. The light source device includes a pulsed light source that generates pulsed light including a continuous spectrum, a splitter that spatially splits the pulsed light into a plurality of n beams (n≧2) according to the wavelength, and a splitter that spatially splits the pulsed light into a plurality of n beams (n≧2) according to the wavelength. A plurality of n fibers that provide different delays, a coupler that is a bundle fiber or multi-core fiber that combines n beams output from the n fibers, and a rod whose input end is coupled to the output end of the coupler. Including an integrator. The optical system includes a critical illumination system that projects a light source image at the output end of the rod integrator onto the object.

Multiple beams with Gaussian intensity distribution are emitted from the coupler. As the plurality of beams pass through the rod integrator, each beam has a top-hat beam profile that is flatter than the Gaussian distribution at the output end of the rod integrator. Therefore, a light source image in which a plurality of beams having different wavelengths are superimposed with a uniform intensity distribution is generated at the output end of the rod integrator. By directly irradiating this light source image onto the surface of the object (sample surface) using the critical illumination system, a plurality of wavelengths are irradiated to the same location and the same range with a uniform illuminance distribution. A top-hat beam has less energy at the bottom than a Gaussian beam, so even if the beam irradiation position is moved to the edge of the object, stray light can be suppressed.

Note that in this specification, a "top-hat beam profile" does not mean a complete top-hat shape, but may include a profile similar to a top-hat shape. A top-hat beam profile has two characteristics: a steep edge and a flat peak.Related to issue 1, the steepness of the edge is important, and issue 2 Relatedly, peak flatness is important.

For example, if the object is flat, only issue 1 can be focused on, and in that case, the intensity distribution at the output end of the rod integrator only needs to have at least a steep tail. is also included in the beam profile similar to the top hat type.

Alternatively, in the case of a device with sufficient measures against stray light, only issue 2 can be focused on, and the intensity distribution at the output end of the rod integrator only needs to have at least flatness of the peak. A beam profile similar to a top hat type also includes a beam profile similar to that of a top hat type.

In one embodiment, a top-hat beam profile may refer to an edge steepness of 0.2 or less and a flatness of 25% or less.

The steepness is calculated as follows using the beam width corresponding to 40% intensity (40% width) and the beam width corresponding to 10% intensity (10% width) when the peak of the intensity distribution is normalized to 100%. It shall be defined as follows.
Steepness = (10% width - 40% width) / 10% width

Further, flatness is defined by the CV value (Coefficient of Variation) in a range where the intensity is 40% or more (within a 40% width).

In one embodiment, the cross section of the rod integrator may be polygonal. The closer the rod integrator is to a circle, the longer the rod length required to obtain a top-hat beam profile. Conversely, a triangular or rectangular cross section can shorten the rod length required to obtain the same top-hat beam profile, but requires high accuracy in alignment with the coupler. Considering these balances, a pentagonal to octagonal cross section is preferable.

In one embodiment, the rod integrator may have a hexagonal cross section. In this case, it is possible to reduce the required rod length to obtain a top-hat beam profile while easing the required alignment accuracy with the coupler.

In one embodiment, the length of the rod integrator may be greater than 100 mm. This makes it possible to obtain a sufficiently flat beam profile.

In one embodiment, the length of the rod integrator may be less than 500 mm. For example, if you try to achieve a top hat beam profile using a circular rod integrator, you will need a rod length of over 500mm, but such a long rod integrator poses a problem of bending loss due to its own weight, and the length of the rod is also long. It is not easy to position and support rod integrators with high precision. If the length of the rod integrator is 500 mm, preferably 200 mm or less, bending due to its own weight can be suppressed, and it becomes easy to position and support with high precision.

(Embodiment)
Hereinafter, the present disclosure will be described based on preferred embodiments with reference to the drawings. Identical or equivalent components, members, and processes shown in each drawing are designated by the same reference numerals, and redundant explanations will be omitted as appropriate. Further, the embodiments are illustrative rather than limiting the disclosure, and all features and combinations thereof described in the embodiments are not necessarily essential to the disclosure.

The dimensions (thickness, length, width, etc.) of each member described in the drawings may be scaled up or down as appropriate for ease of understanding. Furthermore, the dimensions of multiple members do not necessarily represent their size relationship, and even if a member A is drawn thicker than another member B on a drawing, member A may be drawn thicker than member B. It may be thinner than that.

FIG. 5 is a diagram showing the optical measurement device 100 according to the embodiment. The optical measurement device 100 includes a light source device 200, an irradiation optical system 310, a light receiving device 320, and an arithmetic processing device 400.

The light source device 200 generates wavelength swept light L1 whose wavelength changes over time. In the wavelength swept light L1, time and wavelength are associated in a one-to-one relationship. This means that the wavelength swept light L1 "has wavelength uniqueness."

The light source device 200 includes a pulse light source 210 and a pulse stretcher 220.

The pulsed light source 210 emits broadband pulsed light L1a having a broadband continuous spectrum. The spectrum of the broadband pulsed light L1a is continuous over a wavelength range of at least 10 nm, preferably 50 nm, and more preferably 100 nm, for example in the range of 900 nm to 1300 nm. The width of the wavelength range of the broadband pulsed light L1a should just cover the wavelength range necessary for spectroscopy.

For example, the pulsed light source 210 may include an ultrashort pulse laser and a nonlinear element. Examples of ultrashort pulse lasers include gain switch lasers, microchip lasers, fiber lasers, and the like.

The nonlinear element further widens the spectral width of the ultrashort pulses generated by the ultrashort pulse laser through nonlinear phenomena. A fiber is suitable as the nonlinear element, and for example, a photonic crystal fiber or other nonlinear fiber can be used. Although it is preferable to use a single mode as the fiber mode, a multimode fiber can also be used as long as it exhibits sufficient nonlinearity.

Other broadband pulsed light sources, such as a superluminescent diode (SLD) light source, may also be used as the pulsed light source 210.

The broadband pulsed light L1a output from the nonlinear element has a pulse width on the order of femtoseconds to nanoseconds.

The pulse stretcher 220 stretches the pulse width of the broadband pulsed light L1a and outputs the wavelength swept light L1. Pulse stretcher 220 includes a divider 222, a delay line 224, a coupler 226, and a rod integrator 228.

The splitter 222 splits the broadband pulsed light L1a into a plurality of n beams (n≧2) according to the wavelength. Although the configuration of the divider 222 is not particularly limited, it may be configured with an arrayed waveguide grating (AWG), for example.

The delay line 224 includes a plurality of n fibers FB1 to FBn. Fibers FB1-FBn have different lengths and provide different delays to the beams split by splitter 222.

It is assumed that the broadband pulsed light L1a before division is a positive chirp pulse (up-chirp pulse) whose frequency increases (wavelength decreases) with time. In this case, the leading edge of the pulse contains a component with the longest wavelength λ ₁ and the trailing edge of the pulse contains a component with the shortest wavelength λ _n .

The plurality of fibers FB1 to FBn have different lengths l ₁ to l _n . Assuming that λ ₁ is the longest wavelength and λ _n is the shortest wavelength, in order to make the wavelength swept light L1 the same positive chirped pulse as the broadband pulsed light L1a, the relationship 1 ₁ <l ₂ <...<l _n is established. It is sufficient if it satisfies the following.
As an example, if n=20, the lengths l ₁ to l ₂₀ of the 20 fibers FB may increase in 1 m increments from 1 m to 20 m.

Fibers FB1 to FBn do not need to have different group delay characteristics for each wavelength, and the same fiber (fiber with the same core/clad material) can be used. In this sense, fiber FB can be a multimode fiber, which is advantageous in that unintended nonlinear optical effects can be prevented.

The coupler 226 recombines the plurality of beams given different delays by the delay line 224. For example, coupler 226 is a bundle fiber or multi-core fiber that combines n beams.

The rod integrator 228 has its input end optically coupled to the output end of the coupler 226, and functions as a homogenizer to smooth the beam profile.

FIG. 6 is a perspective view of coupler 226 and rod integrator 228. The coupler 226 includes a plurality of fibers 227, each of which carries beams of different wavelengths λ ₁ , λ ₂ , . . . λ _n . Within each fiber 227, the beam propagates in a single transverse mode, and at the exit end of the fiber 227, the profile of each beam has a Gaussian distribution. The beams of each wavelength are incident on the input end of rod integrator 228 at different positions.

FIG. 7 is a cross-sectional view of the coupler 226 and rod integrator 228. In this example, coupler 226 includes sixty fibers 227 that are distributed across the cross section of rod integrator 228.

FIG. 8 is a diagram illustrating profile conversion by the rod integrator 228. In FIG. 8, only the beams BMi and BMj of the i-th wavelength λi and the j-th wavelength λj are shown. The Gaussian beam emitted from the fiber 227 propagates within the rod integrator 228. When propagating through the rod integrator 228, the beam becomes multimode, and the size of the beam increases to the size of the cross section of the rod integrator 228 at the output end 229 of the rod integrator 228. The profile at the exit end of each beam BMi, BMj approaches a top hat shape.

The above is the configuration of the light source device 200.

FIG. 9 is a diagram showing the wavelength swept light L1. The upper part of FIG. 9 shows the intensity (temporal waveform) I _WS (t) of the wavelength swept light L1, and the lower part shows the temporal change in the wavelength λ of the wavelength swept light L1. In this example, the wavelength swept light L1 is one pulsed light, and the dominant wavelength is λ ₁ at the leading edge, and the dominant wavelength is λ _n at the trailing edge, and the wavelength changes from λ ₁ to λ within one pulse. _n changes over time. In this example, the wavelength swept light L1 is a positive chirped pulse (λ ₁ >λ _n ) whose frequency increases with time, in other words, whose wavelength decreases with time. Note that the wavelength swept light L1 may be a negative chirped pulse whose wavelength becomes longer with time (λ ₁ <λ _n ). As described later, the wavelength swept light L1 may be a pulse train consisting of temporally isolated pulses (wave packets) for each wavelength.

Return to Figure 5. The irradiation optical system 310 includes a critical illumination system 312, a beam splitter 314, and a mirror 316.

The critical illumination system 312 projects the light source image of the output end 229 of the rod integrator 228 onto the surface of the sample 2 (sample surface).

The beam splitter 314 directs a part of the wavelength swept light L1 toward the sample 2. The beam splitter 314 also extracts a part of the wavelength swept light L1 as a reference light L3. Mirror 316 directs reference light L3 toward light receiving device 320.

The light receiving device 320 includes a first light receiver 322, a second light receiver 324, and A/D converters 326 and 328. The first light receiver 322 detects object light L2 obtained by irradiating the sample 2 with the wavelength swept light L1. The object light L2 may be reflected light or transmitted light.

The A/D converter 326 converts the output signal S1 of the first light receiver 322 into a digital signal D1. The second light receiver 324 detects the reference light L3. The A/D converter 328 converts the output signal S2 of the second light receiver 324 into a digital signal D2. The time waveform I _OBJ (t) of the object light L2 indicated by the digital signal D1 and the time waveform I _REF (t) of the reference light L3 indicated by the digital signal D2 are taken into the arithmetic processing device 400.

In the wavelength-swept spectroscopy, there is a one-to-one correspondence between time and wavelength in the wavelength-swept light L1. Naturally, this correspondence relationship also applies to the reference light L3 and also to the object light L2. Using this correspondence between time and wavelength, the processing unit 400 converts the time waveform I _OBJ (t) of the object light L2 into a frequency domain spectrum I _OBJ (λ). The arithmetic processing unit 400 also calculates the reference spectrum I _REF (λ) by converting the temporal waveform I _REF (t) of the reference light L3 into a spectrum and appropriately scaling the spectrum.

Although the processing of the arithmetic processing device 400 is not particularly limited, as an _example , the arithmetic processing device ₄₀₀ calculates the transmittance T( λ) or the reflectance R(λ) can be calculated.
T(λ)=I _OBJ (λ)/I _REF (λ)
R(λ)=I _OBJ (λ)/I _REF (λ)

FIG. 10 is a diagram illustrating spectroscopy by the optical measuring device 100 of FIG. 5. As mentioned above, in the wavelength swept light L1, the time t and the wavelength λ correspond one to one, so the time waveform I _REF (t) can be converted into the frequency domain spectrum I _REF (λ). Can be done.

The time waveform I _OBJ (t) of the object light L2 also has a one-to-one correspondence between the time t and the wavelength λ. Therefore, the processing unit 400 can convert the waveform I _OBJ (t) of the object light L2 indicated by the output of the light receiving device 320 into the spectrum I _OBJ (λ) of the object light L2.

The arithmetic processing unit 400 calculates the transmission spectrum T(λ) of the object OBJ based on the ratio I _OBJ (λ)/I _REF (λ) of the two spectra I _OBJ (λ) and I _REF (λ). be able to.

It is assumed that the relationship between the wavelength λ and the time t in the wavelength swept light L1 is expressed by a function λ=f(t). Most simply, the wavelength λ varies linearly with time t according to a linear function. When the time waveform I _OBJ (t) of the object light L2 decreases at a certain time t _x , the transmission spectrum T (λ) means that it has an absorption spectrum at the wavelength λ _x = f (t _x ).

Note that the processing in the arithmetic processing device 400 is not limited to this. After calculating the ratio T(t)=I _OBJ (t)/I _REF (t) of two time waveforms I _OBJ (t) and I _REF (t), the variable t of this time waveform T(t) The transmission spectrum T(λ) may be calculated by converting λ to λ.

Next, the advantages of the optical measurement device 100 will be explained.

FIG. 11 is a diagram illustrating the intensity distribution of the wavelength swept light L1 irradiated onto the surface of the sample 2.
As described above, at the output end 229 of the rod integrator 228, the beams BM1 to BMn of wavelengths λ1 to λn each have a top hat-shaped profile, and form a light source image at the output end 229. The critical illumination system 312 projects this light source image onto the surface of the sample 2 while maintaining the intensity distribution. Note that the critical illumination system 312 may have a magnification depending on the size of the output end 229 and the size of the sample 2, and therefore may perform enlarged projection or reduced projection. The surface of the sample 2 is sequentially irradiated with a plurality of beams having different wavelengths and having a top-hat illuminance distribution as the wavelength swept light L1.

FIG. 12 is a diagram illustrating the first effect of the optical measurement device 100. The first effect corresponds to issue 1. FIG. 12 shows the relative positional relationship between the object OBJ and the beam of the wavelength swept light L1. As described above, the wavelength swept light L1 is irradiated multiple times while moving the object OBJ. Since the beam irradiated onto the surface of the object OBJ has a top hat-shaped intensity distribution, the energy of stray light is small even if the beam is irradiated close to the end of the object OBJ. Therefore, the above problem 1 caused by stray light can be solved.

FIG. 13 is a diagram illustrating the second effect of the optical measurement device 100. The second effect corresponds to issue 2. In this embodiment, the illumination image of the output end 229 of the rod integrator 228 is directly projected onto the surface of the object OBJ using the critical illumination system 312. Therefore, beams of a plurality of wavelengths are irradiated onto substantially the same location with a top hat-shaped intensity distribution. Thereby, beams with different wavelengths will feel the same thickness, and the above-mentioned problem 2 can be solved.

Next, the design of the rod integrator 228 will be explained. As described above, the rod integrator 228 is designed so that beams of multiple wavelengths have a top-hat beam profile at the output end 229. The top-hat beam profile will now be explained.

FIG. 14 is a diagram illustrating a top-hat type beam profile. In this embodiment, the top-hat beam profile can be defined from two viewpoints: edge steepness and top flatness.

A "top hat beam profile" does not mean a complete top hat shape, but may include a profile similar to a top hat shape as shown in FIG. 14.
A top-hat beam profile can be understood from two characteristics: the steepness of the skirt (edge) and the flatness of the peak. Regarding issue 1, the steepness of the tail is important, and regarding issue 2, the flatness of the peak is important.

For example, if the object is flat, only issue 1 can be focused on, and in that case, the intensity distribution at the output end of the rod integrator only needs to have at least a steep tail. is also included in the beam profile similar to the top hat type.

Alternatively, in the case of a device with sufficient measures against stray light, only issue 2 can be focused on, and the intensity distribution at the output end of the rod integrator only needs to have at least flatness of the peak. A beam profile similar to a top hat type also includes a beam profile similar to that of a top hat type.

The steepness is calculated as follows using the beam width corresponding to 40% intensity (40% width) and the beam width corresponding to 10% intensity (10% width) when the peak of the intensity distribution is normalized to 100%. It shall be defined as follows.
Steepness = (10% width - 40% width) / 10% width

In order to solve problem 1, it is preferable that the steepness of the edge is 0.2 or less. In other words, in this embodiment, a top-hat beam profile has an edge steepness of 0.2 or less.

Flatness shall be defined by the CV value (Coefficient of Variation) in a range where the intensity is 40% or more (within a 40% width).

In order to solve problem 2, it is preferable that the flatness of the peak is 25% or less. In other words, in this embodiment, a top hat beam profile has a flatness of 25% or less.

Next, a specific example of beam profile evaluation will be explained. FIG. 15 is a diagram showing measurement results of the beam profile of irradiation light generated by the optical measurement device 100 according to the embodiment. The rod integrator 228 has a regular hexagonal cross section that circumscribes a 0.69 mm circle, and has a length of 100 mm. The critical illumination system 312 is composed of two lenses with a focal length of 750 mm, and projects the illumination image of the output end 229 of the rod integrator 228 onto the imaging surface of the image sensor at the same magnification.

As described above, since light propagates in multiple modes within the rod integrator 228, speckles occur in the illumination image at the output end 229. Therefore, the profile of the imaging surface (that is, the sample surface) on which the image is formed also has a mottled pattern. Therefore, when evaluating whether a beam has a top-hat profile, it is necessary to remove the influence of this speckle by smoothing processing.

Although the method of smoothing processing is not particularly limited, an example thereof will be described below.

FIG. 16 is a diagram illustrating beam profile smoothing processing. In process S100, the speckle size (full width) Δ of the original data before processing is obtained. In this example, one speckle spreads over approximately 5 pixels, as shown in the right diagram of process S100.

In process S102, the original data, the data obtained by shifting the original data by Δ pixels in the + direction, and the data obtained by shifting the original data by Δ pixels in the − direction are averaged. At this time, a weighted average may be taken.

In process S104, a moving average is taken using a window with a width of 2Δ. In this example, since Δ=5 pixels, moving average processing of 10 pixels will be performed. Then, in step S106, the peak is normalized to 1 (100%).

The above is an example of beam profile smoothing processing. Those skilled in the art will understand that the smoothing technique is not limited to this. For example, processes S102 and S104 may be interchanged. Furthermore, for final standardization, the process S102 may be simple addition.

As another smoothing method, image data may be converted to frequency domain information and filtered. Alternatively, the beam profile may be evaluated using an image sensor having a pixel pitch larger than the speckle width.

Next, the results of studies on rod integrators 228 with different cross-sectional shapes and lengths will be explained.

FIG. 17 is a diagram showing the profiles of multiple beams with different wavelengths. The measurements were performed under the same conditions as described in FIG. 15. The channel numbers correspond to the fiber position numbers in FIG. 7, and indicate channels on the outer periphery where uniformity is difficult to achieve. The figure on the left of each channel shows the profile before the smoothing process, and the figure on the right shows the profile after the smoothing process.

The results of evaluating the steepness of edges and flatness of peaks for the profile after smoothing processing are as follows.

・CH01
X direction steepness 0.08
Y direction steepness 0.07
X direction flatness 15.5%
Y direction flatness 16.0%

・CH02
X direction steepness 0.108
Y direction steepness 0.067
X direction flatness 15.3%
Y direction flatness 16.9%

・CH03
X direction steepness 0.090
Y direction steepness 0.105
X direction flatness 15.8%
Y direction flatness 16.0%

・CH04
X direction steepness 0.090
Y direction steepness 0.080
X direction flatness 20.6%
Y direction flatness 20.6%

Next, the length dependence of profile conversion by the rod integrator 228 will be explained.

FIG. 18 is a diagram showing a beam profile obtained by a rod integrator 228 having a hexagonal cross section. The beam profile at the output end 229 was simulated by changing the length of the rod integrator 228 to 50 mm, 100 mm, and 150 mm. Since the simulation cannot take multimode into account, the intensity distribution does not have a mottled shape with multiple peaks as in the actual measurement, but the trend matches well with the actual measurement.

In a rod integrator with a hexagonal cross section, when the length is 50 mm, asymmetry is seen at the peak portion, but it can be said to be sufficiently flat for practical use.

Next, the results of studies on rod integrators 228 with various cross-sectional shapes will be explained. We considered circles, triangles, squares, and pentagons.

FIG. 19 is a cross-sectional view of the rod integrator 228. In the figure, the beam profile at the output end was calculated by simulation for the beams coupled to the positions of channels CH03 and CH05 at the input end. The lengths of the rod integrator 228 were 50 mm, 100 mm, and 150 mm.

FIG. 20 is a diagram showing a beam profile obtained by a rod integrator 228 with a circular cross section. A circular rod integrator 228 requires a longer length for uniformity than a hexagonal or other rod integrator. Furthermore, even if the length is up to 150 mm, the center portion is weak in strength and has a donut shape, which is inferior in flatness compared to other shapes of the same length.

FIG. 21 is a diagram showing a beam profile obtained by a rod integrator 228 having a triangular cross section. In the case of a triangle, flattening is possible with a short length. However, since the positional misalignment between the coupler 226 and the rod integrator 228 (misalignment between the center positions of the coupler 226 and the rod integrator 228) and direction misalignment (rotational misalignment) are severe, strict restrictions must be placed on misalignment. . Furthermore, in order to enclose a circular bundle fiber, a triangle with one side having a length of bundle diameter x √3 is required, and a large optical system is required.

Figure 22 shows the beam profile obtained by a rod integrator 228 with a rectangular cross section. Like a triangle, a rectangle can also be flattened in a short length. However, like a triangle, a rectangle is also greatly affected by positional and directional misalignment.

FIG. 23 is a diagram showing a beam profile obtained by a rod integrator 228 with a pentagonal cross section. In the case of a pentagon, the length required for uniformity is longer than that of a triangle or square, but shorter than that of a circle. It can be said that positional and directional deviations have less influence on the beam profile than triangles and squares.

To summarize the above, the closer the cross section is circular, the longer the length of the rod integrator 228 required to obtain a flat profile is, and the closer the cross section is triangular, the shorter the rod integrator 228 required.

Circles are least susceptible to misalignment and directional misalignment, and the closer they are to triangles, the more severe the effects of misalignment and directional misalignment. This is because triangles are not point-symmetric, and in the case of quadrilaterals, there is a large difference between the length of the diagonals and the length of the line segment connecting the midpoints.

Therefore, if a triangular or rectangular rod integrator 228 is adopted, the sample 2 of each beam may be transported due to misalignment of the sample 2, specifically positional misalignment (displacement perpendicular to the conveyance direction) or angular misalignment (displacement in the conveyance direction). The passing distance within the two regions will be different, which may affect measurement reproducibility. In particular, when the shape and irregularities of the sample 2 are irradiated with a relatively large irradiation diameter, the influence becomes significant.

From this point of view, the closer the cross section of the rod integrator 228 is circular, the less the influence of the above-mentioned conveyance deviation will be reduced. However, the circular shape requires a long length for uniformity, and a length of about 100 mm is insufficient for uniformity. .

On the other hand, in the case of pentagons and hexagons, even 100 mm can be made uniform to some extent. In particular, a hexagonal shape is line symmetrical, point symmetrical, and close to a circle, so it is suitable as the cross section of the rod integrator 228.

In particular, it is practically not easy to hold a thin rod with a diameter of 1 mm or less or a long rod with a diameter of several hundred mm or more. Such rods cause losses due to bending due to their own weight. In order to reduce bending, it is necessary to increase the number of supporting parts, but since it is held with an adhesive or the like having a higher refractive index than glass, there is a problem that light leaks from the holding parts and increases loss. From this point of view, in practical terms, the rod integrator 228 should be at most 500 mm or less, preferably 200 mm or less. With the hexagonal rod integrator 228, a beam profile that can withstand practical use can be obtained with a length of 50 mm to 100 mm, so the loss caused by adding the rod integrator 228 can be suppressed.

Although the embodiments of the present disclosure have been described using specific terms, this description is merely an example to aid understanding, and does not limit the scope of the present disclosure or claims. The scope of the present invention is defined by the claims, and therefore embodiments, examples, and modifications not described here are also included within the scope of the present invention.

L1 Wavelength swept light L1a Broadband pulsed light L2 Object light 2 Sample L3 Reference light 100 Light measuring device 200 Light source device 210 Pulse light source 220 Pulse stretcher 222 Splitter FB fiber 224 Delay line 226 Coupler 228 Rod integrator 229 Output end 310 Irradiation optical system 312 Critical illumination system 314 Beam splitter 316 Mirror 320 Light receiving device 322 First light receiver 324 Second light receiver 326, 328 A/D converter 400 Arithmetic processing unit

Claims (5)

a light source device that generates wavelength swept light;
an optical system that irradiates the target object with the wavelength swept light;
Equipped with
The light source device includes:
a pulsed light source that generates pulsed light including a continuous spectrum;
a splitter that spatially splits the pulsed light into a plurality of n beams (n≧2) according to wavelength;
a plurality of n fibers giving different delays to the n beams;
a coupler that is a bundle fiber or a multi-core fiber that combines the n beams output from the n fibers;
a rod integrator whose input end is coupled to the output end of the coupler;
including;
The optical measurement device is characterized in that the optical system includes a critical illumination system that projects a light source image at the output end of the rod integrator onto the object. The optical measurement device according to claim 1, wherein the rod integrator has a polygonal cross section. The optical measurement device according to claim 1, wherein the rod integrator has a hexagonal cross section. The optical measurement device according to any one of claims 1 to 3, characterized in that the length of the rod integrator is longer than 100 mm. 4. The optical measuring device according to claim 1, wherein the length of the rod integrator is shorter than 500 mm.

Images