JP3095167B2 - Multi-channel Fourier transform spectrometer - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a Fourier transform spectrometer, and more particularly to a multi-channel Fourier transform spectrometer using an image sensor.

(Prior Art) A visible Fourier transform spectrometer using an image sensor, particularly a solid-state image sensor, instead of a photographic plate in holographic Fourier transform spectroscopy and using a triangular optical path common path interference system as an optical system has been proposed. (Appl. Opt. 23 , 269 (198
Four)).

This method is called multi-channel Fourier transform spectroscopy because it simultaneously detects interference fringes on multiple channels. In this multi-channel Fourier transform spectroscopy,
Since there is no mechanical drive part, a compact and robust optical system can be assembled compared with the conventional one, and it can meet the demands of field use,
In addition, since multi-channel detection is used, there are many advantages such as time-resolved measurement of dynamic or transient phenomena.

FIG. 1 shows a principle diagram of the conventional example. The optical system includes a light source (S), two mirrors (M1) and (M2), a beam splitter (BS), and a lens (L). In P), the imaging signal is processed, and the data is subjected to Fourier transform to obtain a spectral distribution.

In the optical system, light emitted from the light source (S) is divided into two clockwise and counterclockwise light beams by a beam splitter (BS). Here, when the mirror (M1) is moved by a distance a in parallel from the position where the wavefronts of the two light beams coincide (broken line) (solid line position), the two light beams move in a direction orthogonal to the optical axis. Shift. The two separated light beams are converted into a lens (L)
(Focal length: f) on the image sensor (IS) placed at the rear focal plane of the lens (L), And an interference fringe (also referred to as an interferogram) is formed here. Since this interference fringe is observed on the rear focal plane of the lens (L), it is an interference fringe localized on infinity, that is, an equi-tilt interference fringe. Since this equi-tilt interference fringe is located at the position of the Fourier transform with respect to the intensity, due to its shift invariant property, even when the light source (S) has a spread, the light wave emitted from any point of the light source has the same interference fringe at the same position. Form. Therefore, with this optical system, the visibility of the interference fringes does not decrease due to the spread of the light source (the emission area of the measurement sample).

(Problems of the Prior Art) However, in the above-described optical system, the following three problems arise because a triangular optical path must be formed before the lens (L).

(1) It is difficult to adjust the optical system.

(2) Even if it is miniaturized, there is a limit to this miniaturization, and when the optical system is configured in a dark box, it is bulky and inconvenient in handling, and it is difficult to adapt to field use.

(3) The throughput is limited by the structure itself,
Restricting the resolution.

The throughput of the optical system is given by (area of the light source x solid angle of the lens viewed from the light source). The necessity of setting a triangular optical path in front of the lens causes the distance from the light source to the lens to the effective diameter of the lens. Can not be shortened so much that the throughput cannot be increased unless the F value of the lens is increased. Even if the F value of the lens is increased, the throughput is limited by vignetting caused by the beam splitter (BS) and the mirrors (M1) and (M2).

An object of the present invention is to solve the above problems.

(Means for Solving the Problems) The multi-channel Fourier transform spectrometer of the present invention comprises:
A basic feature is that it is provided with a polarizer, a polarization splitting birefringent element, an analyzer, a convergent lens, and an imaging element located behind the lens along the optical path.

(Operation) The light from the measurement sample is transmitted by the polarizer with only a linearly polarized light component in a specific direction, and the transmitted light is polarized and split into an extraordinary ray and an ordinary ray by a polarization splitting birefringent element to form two light beams.
Emitted from this birefringent element. Each of the two light beams transmits only the same polarization component as that transmitted by the polarizer by the analyzer, and the converging lens forms the two transmitted light beams as interference fringes on the image sensor.

In order to form interference fringes, a polarization splitting birefringent element is used, and a polarizer, a polarization splitting birefringent element, an analyzer, a convergent lens, and an imaging element can be arranged on a straight line.

(Embodiment) FIG. 2 shows a first embodiment.

In FIG. 2, (S) is a light source, (L1) is a light source (S).
Lens that forms an image of the object at the aperture of the aperture (AP),
(L2) is a collimator lens, (P1) is a polarizer, (W) is a Wollaston prism that divides a polarized light component transmitted through the polarizer (P1) into two light beams, and (P2) is a polarizer (P1). ), An analyzer that transmits the same polarized light component as (), (L3) is an analyzer (P
2) An imaging lens that forms an image of the phase difference distribution of the two light beams transmitted through as an intensity interference fringe. (IRIS) is a two-dimensional infrared image sensor, and an infrared image sensor (IRIS) is formed by the imaging lens (L3). , Interference fringes are formed on the light receiving surface.

(FG) is an image capture device that captures video signals from an infrared image sensor (IRIS), A / D converts them, and performs frame processing. (CPU) controls the image capture device (FG) and captures captured image data. Is a computer device for storing and performing arithmetic processing.

As shown in FIG. 3, the Wollaston prism is a prism 1 having an apex angle α on one side of a uniaxial crystal whose end face is polished parallel to the optical axis, and a prism 2 having the same apex angle whose optical axis is orthogonal to the other. This is a multiple image element joined in the same plane. The light that is perpendicularly incident on the Wollaston prism (W) travels straight with polarization components orthogonal to each other as an ordinary ray o and an extraordinary ray e. Was an ordinary ray o, but an extraordinary ray e
Then, the extraordinary ray e becomes the ordinary ray o and is separated. Two light separation angle θ = 2ψ at this time is represented by _{sinψ = (n o -n e)} tanα + 0 (α 3) ... (1-1). Here, n _o is the refractive index of the ordinary ray, n _e is the refractive index of extraordinary ray.

In FIG. 2, the light emitted from the aperture (AP) is collimated by a collimator lens (L2), and is converted into a parallel light by a polarizer (P
According to 1), only the 45-degree component passes through the paper surface. This linearly polarized light is split into two linearly polarized light beams orthogonal to each other by a Wollaston prism (W) and emitted with a separation angle θ. The two split light beams form interference fringes on the infrared image sensor (IRIS) by the imaging lens (L3). However, since the two split light beams do not interfere as they are because their polarization directions are orthogonal, the analyzer (P2) transmits only the same polarization component as the polarizer (P1) to generate interference fringes. I have.

The interference fringe intensity distribution I (x) generated at this time is such that the observation surface is in a direction perpendicular to the optical axis which bisects the angle formed by the two light waves with the light receiving surface of the infrared image sensor (IRIS). ), The wave number is σ, and the spectral distribution of the light wave (light source) is S
Expressed as (σ), And S (σ) is given by the Fourier transform of equation (1-2).

When the polarization directions of the two polarizing plates (P1) and (P2) are orthogonal to each other, the phase difference between the two light beams is shifted by π.
The intensity distribution Ir (x) of the interference fringes generated in this case is Becomes

That is, if the two polarizing plates are perpendicular to each other, the brightness of the interference fringes is inverted. This can be used to improve the SN ratio. That is, an arithmetic operation for subtracting the expression (1-3) from the expression (1-2) is performed. Difference intensity distribution Id
(X) And the signal component is doubled at the same time as the background component is removed. Of course, this arithmetic operation is performed on a computer device (CPU), but the SN ratio is significantly improved.

The optical system of the first embodiment using the Wollaston prism (W) observes interference fringes localized in the Wollaston prism (W). Since this interference fringe is an equal thickness interference fringe,
When the light source has a spread, the interference fringes caused by the light emitted from a point off the optical axis are slightly shifted laterally on the observation surface with respect to the interference fringe caused by the light emitted from a point on the optical axis. . For this reason, the visibility of the interference fringes decreases. Also,
There is also a decrease in visibility due to crystal aberration. In practice, in order to prevent this decrease in visibility, the spread (area) of the light source is limited. The aperture (AP) shown in FIG. 2 is provided for this purpose.

The allowable solid angle Ωmax is given by the following formula with respect to the maximum wave number σmax: Ωmax = 2π / (θx · σmax) (1-5) However, this value is the same as that of a conventional Michelson-type Fourier transform spectrometer. Compared to the conventional one with a common path triangular optical path, this optical system limits the spread of the light source, but the collimator lens (L2)
For example, if the collimator lens (L2) has a small F value, the throughput can be made equal to or greater than that of the common path triangular optical path.

The features of the first embodiment are summarized as follows.

The optical elements can be arranged in a straight line, and the adjustment of the optical system is extremely easy.
It is easy to handle and use, and adapts well to field use. And it is possible to obtain the same throughput as that of the common path triangular optical path.

Next, a more specific description will be given, and data of the obtained results will be shown.

The collimator lens (L2) shown in FIG. _{2 is} made of CaF _{2 and} a plano-convex lens having a focal length of 120 mm and a diameter of 50 mm. The imaging lens (L3) is also made of CaF ₂ and has a focal length of 100 mm and a diameter of 50 mm.
Is a biconvex lens. Wollaston prism (W) is 30
× 30 × 30mm ^3, the separation angle is used as one degree of MgF ₂ made.
Polarizing plates (P1) and (P2) are both Cambridge Physical Sci
ences Ltd. · IGP227,40mmφ, using infrared grid polarizer of CaF ₂ substrate made. The light source (S) uses a nichrome line, and the lens (L1) that forms an image of the nichrome line on the aperture of the aperture (AP) is made of CaF ₂ and has a biconvex lens with a focal length of 40 mm and a diameter of 30 mm. Was. The size of the aperture of the aperture (AP) was 3 mmφ.

The infrared image sensor (IRIS) has a 512 x 512 pixel IR
-A CDS type, IR-5120A manufactured by Mitsubishi Electric Corporation (the size of one pixel is 26 × 20 μm ² ) was used. The video signal of this infrared image sensor (IRIS) is converted into a digital signal of 8 bits per pixel by a pixel capture device (FG) and transferred to a computer device (CPU), where apodization, intra-frame integration and FFT calculation are performed. And outputs a spectrum distribution. In order to further improve the SN ratio, integration of 30 frames and integration of 40 lines in the same frame are performed.

 4 to 12 show the results obtained in this specific example.

FIG. 4 is an interferogram of a nichrome wire of a light source. (A) is the intensity distribution of the interference fringes represented by the equation (1-2), (b) is the intensity distribution of the interference fringes represented by the equation (1-3), and (c) is the difference between the two. 4) where the background component is removed and the signal component is 2
You can see that it has doubled. FIG. 5 shows this spectrum distribution.

It is a result of examining a polymer thin film (polystyrene, polyethylene, polypropylene).

FIG. 6 shows a spectrum when a thin film of polystyrene is placed in the aperture of the aperture (AP). (A) is a transmission spectrum and (b) is a reference light spectrum. The absorption spectrum of polystyrene obtained from these two spectra is shown in FIG. The absorption on the right near 3.4 μm is due to the stretching vibration due to the saturated CH group in the main chain, and the absorption on the left is due to the stretching vibration due to the unsaturated C = H group in the benzene ring.

7 (a) and 7 (b) show the transmission spectrum and absorption spectrum of polyethylene.

The fluctuations above and below the baseline seen in the above absorption spectra (FIGS. 6 (c), 7 (b) and 8 (b)) are considered to be due to the effect of multiple reflections on the thin film.

Next, the results when the intersection angle (θ) is changed by the magnification of the lens (L3) will be described.

FIG. 9 (a) is an interferogram of light transmitted through a 3.45 μm interference filter when the lateral magnification is 1.1 ×,
(B) is an interferogram when the lateral magnification is 0.71, and (c) is when the magnification is 0.58. The corresponding spectra are shown in FIGS. 10 (a), (b) and (c). From FIG. 10, it can be seen that the half-width of the interference filter increases as the wavenumber resolution increases. Incidentally, the wave number resolutions in FIGS. 10 (a), (b) and (c) are 63 cm ^-1 , 40 cm ^-1 and 32 cm ^-1 , respectively.

For further verification, the influence of a change in magnification on the above absorption spectrum of polystyrene (FIG. 6 (c)) was examined. The results are shown in FIG. It can be seen that the resolution of the two absorption peaks was improved by improving the wavenumber resolution, and finer resolution was possible. However, since the number of channels does not change, the resolution is the same.

FIG. 12 shows the reproducibility of the absorption spectrum. It is an example of the absorption spectrum performed with the polystyrene thin film.
From FIG. 12, it can be seen that reproducibility is particularly high especially in the C-H absorption region. Some variation is seen where the spectral intensity of the reference light is weak. However, the standard deviation is
0.18 percent at 3.3 μm, 0.91 percent at 3.5 μm,
It is only 1.65% at 2.9μm. It is considered that the cause of the variation is mainly the change of the atmosphere due to convection and the change of the nichrome wire itself because the nichrome wire was left in the air as it is.

 Next, a second embodiment will be described.

FIG. 13 is a principle diagram showing the configuration of the second embodiment.
(S2) is a light source, (P3) is a polarizer, (SP) is a Savart plate as a polarization splitting birefringent element, (P4) is an analyzer, (L4) is a lens, and (IRS) is after the lens (L4). This is an infrared image sensor in which a light receiving surface is made coincident with a side focal plane. (FG) is an image capture device, and (CPU) is a computer device, which are equivalent or equivalent to those of the first embodiment.

As shown in FIG. 14, the Savart plate (SP) is composed of two plane-parallel plates of a uniaxial crystal whose normal to the boundary surface forms a 45-degree component with the optical axis. It is a double image element joined so as to be orthogonal. The light IL that is perpendicularly incident on the Savart plate (SP) has two polarization components orthogonal to each other, and the ordinary ray O goes straight as it is, while the extraordinary ray E refracts at the interface and goes straight. In the second crystal, the ordinary ray O in the first crystal is refracted at the bonding interface to become an extraordinary ray E and travels straight, while the extraordinary ray E in the first crystal is
Is refracted at the junction interface to become an ordinary ray O and travels straight in a direction parallel to the normal incident light IL. That is, the two light beams emitted from the Savart plate (SP) travel in the same direction as the incident light IL and in a direction perpendicular to the plate surface. During this time 2 light, the thickness of one plane-parallel plate t, a refractive index n _o of the ordinary ray, the refractive index of the extraordinary ray When n _e, Only lateral separation occurs.

Now, in FIG. 13, the light emitted from the light source (S2) is transmitted only by the polarizer (P3) in the direction perpendicular to the paper surface. This linearly polarized light is split into two linearly polarized lights orthogonal to each other by a Savart plate (SP), separated by a distance d given by the equation (2-1), and emitted in a direction parallel to the optical axis. The two rays that are separated and travel in parallel are the analyzer (P
By 4), only the same polarization component as the transmission component by the polarizer (P3) is transmitted, and the lens (L4) having the focal length f is superimposed on the infrared image sensor (IRIS) to form interference fringes.

As in the case of the first embodiment, the light and darkness of the interference fringes is reversed when the polarization directions of the polarizing plates (P3) and (P4) are parallel to each other and when they are orthogonal to each other. Therefore, easily,
Background components can be removed.

In the case of the second embodiment, the source doubling system is equivalent to that of the triangular optical path common path shown in the conventional example. For this reason, even if the light source is spread, the visibility does not decrease. Further, unlike the case of the triangular optical path, since the Savart plate (SP) can be placed immediately before the lens (L4), there is almost no vignetting in the optical system, and a very large throughput can be obtained.

In the case of the second embodiment, only one lens is required in the optical system. Compared to the first embodiment using the Wollaston prism, in the first embodiment, the collimator lens (L2) and the interference fringes localized in the Wollaston prism (W) are placed on the image sensor (IRIS). A lens (L3) for forming an image is essential. In the second embodiment (FIG. 13) using a Savart plate (SP), the lens (L4)
Is the same as the collimator lens (L2), the total length of the optical system can be reduced by the length of the imaging system.

 The features of the second embodiment are summarized as follows.

Since the optical elements can be arranged in a straight line, the adjustment of the optical system is extremely easy and the number of elements is small, so when configuring the entire optical system in a dark box, it is more compact with a cylindrical shape, and of course it can be hand-held. In, handling,
It is easy to use and suitable for field use.

The highest throughput can be obtained. Increasing the size of the Savart plate can further increase the throughput in proportion thereto. A large amount of light flux is equivalent to a large amount of signal as it is, and the SN ratio is also increased in proportion thereto.

In connection with the above, here is a comparison of the throughput of the three.

First, the case of the Savart plate. As an image sensor, the length of one pixel is 25 μm and the total length is 1024 channels.
It is a 6.6 mm multi-channel image sensor (for example, a one-dimensional solid-state image sensor). The Savart plate is the same 15 × 15 × 15 mm ³ made of Tio ₂ as used in the following specific examples,
The lens has a focal length of f = 60 mm. At this time, the maximum wave number that can be measured is 11300 cm ⁻¹ . The lens and the light source are placed immediately before, before and after the Savart plate (see FIG. 15). Since the average refractive index of TiO ₂ is 2.5, the equivalent position of the light source is 15 / 2.5 = 6 mm from the exit surface. At this time, the allowable size of the light source with no vignetting on the exit surface is 12.5 mm. Solid angle is 17.4stradmm ^2.

Next, in the case of the triangular optical path common path listed in the conventional example, the system shown in FIG. 16 is used. Light beam width H is 50mm, mirror (M
1), (M2) 54mmφ, Beam splitter (BS) 70.7m
mφ. Also in this case, it is assumed that the position of the light source is at the position indicated by the triangle △ in the figure, and the position of the lens is at the position indicated by the square □. In this case, the vignetting of the luminous flux is
It is determined by the beam splitter (BS) that recombines after the light beam makes a round in the interferometer. This beam splitter (BS)
Is 316 mm from the light source, but if a lens with f = 60 mm is used as before, the beam splitter (B
Due to vignetting by S), only about 1/3 of the total luminous flux reaches the sensor. Conversely, the focal length of the lens needs to be f = 160 mm in order for light to reach everything on the sensor, and the focal length f = 223 mm to make the size of the light source 12.5 mm.
Lens is required. The solid angle at this time is 1.27stradm
a m ^2.

When the Wollaston prism of the second embodiment is used,
The radius r of the allowable aperture of the Wollaston prism is
From equation (1-5), let N be the numerical aperture. Given by

Here, the focal length of the collimator lens (L2) is f = 60
If mm, r = 3.75 mm. When r = 3.75, the visibility of the maximum wave number σmax becomes 0, so it is actually necessary that the value be sufficiently smaller than this. Therefore, if r = 1 mm, the solid angle is 0.44 stradmm ² .

From the above, it can be seen how the second embodiment using the Savart plate has a large throughput. As described above, if the size of the Savart plate is increased, a higher throughput can be realized, and the SN ratio can be significantly improved.

Next, a specific example of the second embodiment is shown to show an example of the performance.

The light source (S2) is a nichrome wire as in the first embodiment.
Both polarizer (P3) and analyzer (P4) are manufactured by Polaroid.
HR type. The Savart plate (SP) is made of TiO ₂ having a size of 15 × 15 × 15 mm ³ and an average refractive index of 2.5. Lens (L
4) is a plano-convex lens made of fused quartz with a focal length of 80 mm and a diameter of 50 mm. For the infrared image sensor (IRS), N214-06 near infrared vidicon manufactured by Hamamatsu Photonics KK is used. The video signal from the near-infrared vidicon is controlled and processed by a frame grabber / image capturing device (FG) and a computer device (CPU), as in the first embodiment, to obtain a spectral distribution. The S / N ratio is improved by integrating 30 frames and integrating 40 lines in the same frame.

FIG. 17 shows the intensity distribution of the interference fringes of the nichrome line as the light source. (A) is an interference fringe represented by the equation (1-2), (b)
Is the interference fringe represented by the equation (1-3), and (c) is the difference between them, which is given by the equation (1-4). The background component is removed and the signal component is doubled. I understand. Similarly, FIG. 18 shows the intensity distribution of interference fringes obtained from a semiconductor laser having an oscillation wavelength of 1.3 μm. FIGS. 18 (a), (b) and (c) show equations (1-2) and (1), respectively.
-3) and (1-4).

FIGS. 19 (a) and 19 (b) show spectral distributions obtained by Fourier transforming the interference fringes of FIGS. 17 (c) and 18 (c), respectively. The wave number resolution of the spectrum obtained with this system was 100 cm ^-1 . The spectral wavelength range is 1-2 μm, which is determined by the spectral characteristics of the near-infrared vidicon and the polarizing plate.

In order to show the usefulness of the second embodiment, an example in which it is used for near-infrared diffuse reflection analysis will be described. FIG. 20 is a diagram showing the principle of construction of the apparatus, and the same reference numerals as those in FIG. 13 indicate the same or corresponding elements. (SB) is a sample base table on which the sample is placed, and the lens (L5) is a lens that focuses the diffused light beam of the light source (S2) on the sample base table (SB).
The focusing lens (L5) has a focal length of 50 mm,
The diameter is 60mm.

Near-infrared diffuse reflectance analysis is non-destructive and pre-treated (pulverization only)
Because of its simplicity, it is widely used for quantitative analysis of protein, moisture, oil, etc. in agricultural products and foods. Particularly in Canada, near-infrared spectroscopy has been officially approved for the determination of wheat protein.

The absorption in the near infrared region (0.8-2.7 μm) is the overtone absorption or the binding absorption of the normal vibration of the molecule. Since these absorptions are much weaker than the normal vibration in the mid-infrared region, there is no nonlinearity in the absorption. In addition, since the overtone absorption of the bond contained in organic compounds such as CH, NH, OH, and CO exists in the near infrared region, the near infrared region is suitable for quantitative analysis of organic substances. . However, since the overtone absorption and the coupling absorption have a wide spectrum width and many absorption bands, the absorption peaks overlap and show a complicated absorption spectrum. This makes it difficult to identify structurally important spectra, and is not suitable for qualitative analysis.

Diffuse reflection analysis is often used to measure the absorption spectrum in the near infrared region. In this method, a sample is made into a powdery state, and diffuse reflection light that has been transmitted multiple times through the powder is observed. Diffuse reflected light also satisfies Beer's law, and concentration and absorbance are proportional.

However, a disadvantage of this method is that the diffuse reflection light from the sample is weak. Conventional dispersive spectrometers use an integrating sphere, but are not suitable for field use. On the other hand, the spectrum obtained by the near-infrared diffuse reflectance analysis does not have a fine structure, and therefore does not need to have high resolution. Therefore, multi-channel Fourier transform spectroscopy is optimal for performing near-infrared diffusion analysis in field use.

The results measured with the system of FIG. 20 are shown in FIGS. 21 to 26. In the diffuse reflection spectrum shown here, the vertical axis represents the ratio of the diffuse reflected light intensity of the sample to the reflected light intensity of the white plate having no special absorption in the near-infrared region. (1 / R). This value is on an absorbance scale and is proportional to the amount of the component.

FIG. 21 (a) shows interference fringes of the reference light, (b) shows interference fringes due to flour, FIG. 22 (a) shows the spectrum of the reference light,
(B) shows the spectrum of flour, and FIG. 23 shows the diffuse reflection spectrum of flour.

The peak near 1.4 μm is the first harmonic of the normal vibration of the O—H atomic group, the peak near 1.7 μm is the first harmonic of the normal vibration of the CH atomic group, and the peak near 1.2 μm is the CH atomic group. Is the second harmonic of the reference vibration of. It can be seen from the overtone absorption of the CH atomic group that the higher-order absorption is smaller.

FIG. 24 and FIG. 25 show the diffuse reflection spectra of starch and rice powder. The height of each absorption peak correlates with the amount of each component.

FIG. 26 shows the diffuse reflection spectrum of flour moistened with water. In comparison with FIG. 23, it can be seen that the peak of the overtone absorption of the primary OH group increases with respect to the peak of the overtone absorption of the primary and secondary CH groups. This indicates that the absorption by the OH atomic group is increased as the water content increases.

In the above embodiment, the image sensor is in the infrared region. However, it goes without saying that the image sensor in the visible region and / or the ultraviolet region can be applied in exactly the same manner. The image sensor is preferably a solid-state image sensor, more preferably a one-dimensional solid-state image sensor particularly considering field use.

In the second embodiment, TiO ₂ is used for the Savart plate. However, calcite is used to construct a multi-channel Fourier transform spectroscopy system for visible and ultraviolet. The image sensor (imaging device) may be an imaging tube, but is preferably configured by a solid-state imaging device.

In the above-described embodiment, the interference fringes are sampled by the two-dimensional image sensor and the image capturing device adopting the standard TV scanning method. Therefore, the resolution is determined by the number of TV scanning lines or the pixel of the image capturing device. It is considered that the number is considerably restricted. In an embodiment, the resolution is
128 (the number of pixels of the image capture device is 256, phase correction is not performed, and the method of taking the absolute value after FFT is used, so it is half of 256), but the bandwidth is 5120c
For m ⁻¹ , the wavenumber resolution is 40 cm ⁻¹ . Although not actually used in any of the above-described embodiments, a one-dimensional infrared solid-state imaging device that can be used is being developed. For example, the Mitsubishi Electric Group has a 4096 × 4 element Schottky Barrier-type platinum silicide array detectors have been announced. For example, when this image sensor is used, the resolution is improved 16 times, and the wave number resolution is 2.5 cm ⁻¹ . Also, if a method of obtaining the real part without taking the absolute value after the FFT is performed by performing phase correction as performed in a normal Fourier transform spectroscope, the resolution is further improved by a factor of 2, and the wave number resolution is increased by 1.25. cm ^-1 is achieved.
This is equivalent to the performance of a currently known FTIR (infrared Fourier transform spectrophotometer), and the spectrophotometer according to the present invention can be almost equivalent to the conventional example in terms of resolution, and is non-driven and extremely non-driven. It has a compact system configuration and also has excellent features such as high-speed sampling.

(Effects of the Invention) As described above, according to the present invention, the throughput of the light beam can be made equal to or higher than that of the conventional example, and the optical system can be assembled in a straight line, so that the adjustment of the optical system becomes easy and the optical system is easily adjusted. The effect is that the system can be made more compact with a cylindrical shape suitable for field use.

In the infrared region, various molecules absorb due to vibration between atoms in the molecule or rotation of the molecule, so quantitative analysis and qualitative analysis of substances using this wavelength range are widely performed. If an infrared solid-state imaging device is used, such an infrared spectroscopy system can be made into a sensor according to the present invention, and a wide variety of field-use substances can be measured.

[Brief description of the drawings]

FIG. 1 is an optical path diagram showing a main part of a conventional example, FIG. 2 is a diagram showing a first embodiment of the present invention, FIG. 3 is an explanatory diagram of the function of a Wollaston prism, FIG. FIG. 4B is a graph showing the intensity distribution of the interference fringes of the nichrome line represented by the equation (1-2), and FIG.
A graph showing the intensity distribution of the interference fringes of the nichrome line represented by the expression 3), FIG. 4 (c) is a graph showing the intensity distribution of the interference fringes with these backgrounds removed, and FIG. 5 is a spectrum distribution of the nichrome line. Fig. 6 (a) is a diagram showing a transmission spectrum of polystyrene, Fig. 6 (b) is a diagram showing a reference light spectrum of polystyrene, Fig. 6 (c) is a diagram showing an absorption spectrum of polystyrene, FIG. 7 (a) shows a transmission spectrum of polyethylene.
FIG. 7 (b) shows the absorption spectrum of polyethylene, FIG. 8 (a) shows the transmission spectrum of polypropylene, FIG. 8 (b) shows the absorption spectrum of polypropylene, FIG. 9 (a) ) Is a diagram showing an interference fringe caused by a 3.45 μm interference filter at a lateral magnification of 1.1 times, FIG. 9B is a diagram showing a case of 0.71 times, and FIG. 9C is a diagram showing a case of 0.58 times. Fig. 10 (a) is a diagram showing a spectrum when the lateral magnification is 1.1 times with the interference fringes by the 3.45 µm interference filter, Fig. 10 (b) is a diagram showing the spectrum when the magnification is 0.71 times, and Fig. 10 ( c) is a diagram showing a spectrum when the magnification is 0.58 times, and FIG. 11 (a) is a graph showing the magnification of the absorption spectrum of polystyrene.
FIG. 11 (b) shows a spectrum when the magnification is 1.1 times.
Is a diagram showing a spectrum at 0.71 times, and FIG.
FIG. 12 shows the spectrum of polystyrene at 0.58 times, FIG. 12 shows the reproducibility of the absorption spectrum of polystyrene, FIG. 13 shows the second embodiment of the present invention, and FIG. 14 shows the function of the Savart plate. FIG. 15 is a diagram for explaining the throughput of the Savart plate, FIG. 16 is a diagram for explaining the throughput of the conventional example, and FIG. 17 (a) is nichrome represented by the formula (1-2). FIG. 17B is a graph showing the intensity distribution of the interference fringes of the line, and FIG.
FIG. 17 (c) is a graph showing the intensity distribution of the interference fringes of the nichrome line represented by the expression 3), FIG. 17 (c) is a graph showing the intensity distribution of the interference fringes from which these backgrounds have been removed, and FIG. FIG. 18 (b) is a graph showing the intensity distribution of the interference fringes of the semiconductor laser expressed by equation (2), and FIG.
-3) is a graph showing the intensity distribution of the interference fringes of the semiconductor laser expressed by the equation, FIG. 18 (c) is a graph showing the intensity distribution of the interference fringes from which these backgrounds have been removed, and FIG. 19 (a) is nichrome. Diagram showing the spectral distribution of the line,
FIG. 19 (b) is a diagram showing a spectrum distribution of the semiconductor laser, FIG. 20 is a diagram showing a device configuration to which the second embodiment is applied, FIG. 21 (a) is a diagram showing interference fringes of the reference light, FIG. 21 (b)
Is a diagram showing interference fringes of flour, FIG. 22 (a) is a diagram showing a spectrum of reference light, FIG. 22 (b) is a diagram showing a spectrum of flour, FIG. 23 is a diagram showing a diffuse reflection spectrum of flour FIG. 24 is a diagram showing a diffuse reflection spectrum of starch, FIG. 25 is a diagram showing a diffuse reflection spectrum of rice, and FIG. 26 is a diagram showing a diffuse reflection spectrum of flour having a high moisture content. P1 ... Polarizer, W ... Wollaston prism, P2 ... Analyzer, L3 ... Lens, IRIS ... Infrared image sensor as image sensor, FG ... Image capture device, CPU ... Computer device P3 ... Polarizer, SP ... Savart plate, P4 ...
Analyzer, L4 ... Lens, IRS ... Infrared image sensor as an image sensor.

──────────────────────────────────────────────────続 き Continuation of front page (56) References JP-A-2-245622 (JP, A) JP-A-59-105508 (JP, A) JP-A-2-147784 (JP, A) JP-A 62-245 251627 (JP, A) (58) Field surveyed (Int. Cl. ⁷ , DB name) G01J 3/45

Claims (1)

(57) [Claims] Along a light path from a light source, a polarizer, a Savart plate as a polarization splitting birefringent element, an analyzer, a convergent lens, and an infrared solid located at a rear focal plane of the lens. A multi-channel Fourier transform spectrometer, comprising: an imaging element.