JPH0458892B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a spectroscopy method based on a completely new principle, that is, Faraday rotation Fourier spectroscopy that utilizes the phenomenon of Faraday rotation, which has not been used as a spectroscopy method in the past. (Prior art) Conventional spectroscopy methods include spectroscopy using A-diffraction gratings, prisms, etc., spectroscopy using B-interference spectrometers, and spectroscopy using the coherent effect of wavelength tunable lasers. . The degree of resolution of these spectroscopy methods differs depending on the measurement target, but theoretically it is possible to achieve very high resolution with each method. It is necessary to manufacture it with an accuracy of about 1/100 of a wavelength, and
It is necessary to suppress vibrations over a distance of several tens of meters to a wavelength, which is extremely difficult in practice. Furthermore, in the B spectroscopy, if the light source, mirror, or its movable parts could be constructed with infinite precision, the resolution would improve within the range of the coherent length, but this is not realistic. Furthermore, since C spectroscopy is sometimes used within a range called coherent spectroscopy and the light source is a laser, it is generally limited to spectroscopy of coherent light. Regarding brightness, in spectroscopy A, brightness is determined by the brightness of the object to be measured (in the case of luminescence) and the light source, and in the case of spectroscopy such as laser light, it is determined by the brightness of all the components that enter the spectrometer or spectrometer. It cannot be said that it is absolutely impossible to take in light, so there are cases where it is bright and cases where it is dark. However, under typical usage conditions, most of the light is lost at the entrance slit.
Since the size of the refraction grating is not so large, F
The current situation is that the value cannot be very large. In addition, B spectroscopy allows for bright light, and for this reason, infrared Freeier spectrophotometers and the like are widely used. However, with the principle used here, it is extremely difficult to manufacture a spectrophotometer for application to short wavelengths, such as the ultraviolet region, due to mechanical accuracy. (Problems to be Solved by the Invention) However, since the spectrometers used for these spectroscopy include components whose resolution is determined by the mechanical precision and mechanical vibration of the components, ultra-high resolution is not possible. In addition, in many cases the configuration uses optical components such as mirrors, prisms, and diffraction gratings, so the measurement wavelength range is limited by these optical components. There was a drawback. The purpose of this invention is to develop a new principle that enables measurement from the vacuum ultraviolet region to the far infrared region without using components that cause mechanical vibration or optical components that limit the measurement wavelength range. The objective is to provide a spectroscopy method based on (Means for Solving the Problems) Therefore, according to the present invention, the light to be detected is polarized and then introduced into a Faraday rotating medium to which a magnetic field is applied, and the polarization plane of the polarized light is changed in the Faraday rotating medium. is rotated multiple times to extract a superimposed signal of wavelength information with different rotation angles depending on the wavelength, separate a specific polarized light component from the superimposed signal, and calculate the light amount of the polarized light component corresponding to the magnetic field strength from the polarized light component. detecting the amount of light, performing inverse Fourier transform on the amount of light based on the magnetic field strength to convert it into intensity information based on the Weerde constant, and converting the Weerde information into wavelength or frequency spectrum information to spectrally analyze the detected light. Features. (Function) As described above, according to the spectroscopy method according to the present invention,
As a result of rotating the plane of polarization of light multiple times, from twice to many times, in a Faraday rotating medium, the difference in the Faraday rotation angle depending on the wavelength is determined as a function of the change in magnetic field strength, and this is once converted into a function of the Weerde constant. This is a method in which the wavelength and frequency functions (ie, spectral intensity distribution) are determined after the conversion. Therefore, this invention can be applied to conventional refractive index dispersion (prism),
It utilizes properties that are completely different from the angular dispersion of diffraction phenomena (diffraction gratings) and the interference effect of light (interferometers).
Its distinctive feature is that the principle of spectroscopy is different from conventional methods. This principle will be explained below. Principle of Faraday rotation Fourier spectroscopy The Faraday rotation medium used in the present invention may in principle be any medium as long as it is transparent to the wavelength of the target light or has a transmittance that allows the light to be detected. Faraday rotation is a fundamental phenomenon that occurs in any material depending on the size of Weerde's constant. This Werde's constant has been measured for various substances, and for example, some of the measurements for liquids are shown in Attached Table 1. Therefore, for example, pure water may be used as a medium because it is transparent from the ultraviolet to the near infrared, or He, which is not listed in this table, because it has low absorption in the short wavelength range. Gases such as rare gases may also be used. I will now explain using pure water as an example. Water is transparent in the range from at least 0.19 um ultraviolet to 1 um infrared. In particular, in the case of water, the permeation distance is determined by the concentration of impurities, and it is easy to achieve a permeation length of 100 m with pure water. Furthermore, since a photomultiplier tube can be used to detect light in the ultraviolet to visible range, measurements can be made even if the transmission distance in water is very long, for example 10 km. On the other hand, an experiment was carried out to investigate the Faraday rotation of pure water, and it was confirmed that by applying a magnetic field of about 10 ⁴ Oe, the polarization plane could be rotated by 90 degrees with about 5 cm of pure water. From these facts, it can be seen that in a very long Faraday rotation medium of water, for example 100 meters or longer in some cases, light undergoes many rotations of the plane of polarization without being almost absorbed. For example, 10 ⁴
If a magnetic field of Oe is applied to water and the water is passed through 100 meters, 248 nm light will have a polarization plane rotation angle of 400 x 360 degrees. Incidentally, the Weerde constant of this pure water is as shown in Attached Table 1, and its change from ultraviolet to near-infrared is shown in FIG. 2 in comparison with the wavelength dependence of the refractive index of quartz. In Figure 2, the vertical axis on the left shows Weerde's constant (V (mjn/cm・Oe)) blotted on a logarithmic scale, the vertical axis on the right shows the refractive index n, and the horizontal axis shows the wavelength. λ(um) is shown as a blot. In addition, due to the influence of the resonance term, values slightly deviate from λ ^-2 in the ultraviolet and near-infrared regions, but it can be seen that the Weerde constant is increasing in both cases. In the case of solids and liquids, this Werde's constant V(λ) is usually given by V(λ)=(a/λ ² )+ΔV(λ). However, a is a constant and ΔV(λ) is a resonance term, which is often a Δ constant V(λ)0.
In the case of gas, it is usually given by V(λ)=a/λ+b/λ ³ . However, a and b are constants. In this way, in the case of a solid or a liquid, the Werde constant V(λ) is inversely proportional to the square of the wavelength. Therefore, an analyzer detects a linearly polarized light component in one direction from a superimposed signal of optical information with different Faraday rotation angles depending on the wavelength component, and the obtained polarized light component signal is converted into an electrical signal by photoelectric conversion. An electric signal is subjected to inverse Fourier transform as a function of magnetic field strength to convert it into a signal as a function of Weerde's constant, and a wavelength or frequency signal can be derived from this signal as a function of Weerde's constant to obtain a spectral intensity distribution. A spectroscopy method that obtains a spectrum by Fourier transforming a signal that utilizes the wavelength dispersion of the Faraday rotation angle is called Faraday Rotation Fourier Transform spectroscopy.
Spectroscopy (also abbreviated as FRFT spectroscopy). The principle of obtaining this spectrum will be explained using FIG. 1. FIG. 1 is a diagram schematically showing an example of the configuration of a spectrophotometer used to perform such Faraday rotation Fourier spectroscopy. LS is the light source of this spectrophotometer, CL is the lens system that converts the light into a parallel beam, P1 is the polarizer that converts the incident light into linearly polarized light, and P2 separates the polarization plane of the output light into P and S components. This polarizer is called an analyzer. PS is a power supply that applies a uniform magnetic field to the Faraday rotary element by applying current to the air-core coil L, and can sweep the magnetic field strength with enough accuracy to ensure the accuracy of the desired spectrum. FR is a power supply with low absorption. It is a Faraday rotation element with a large Werde constant, and in this case, it is formed by placing water in a quartz container, for example. Furthermore, D1 and D2 are photoelectric detectors, and CP is a computer. Furthermore, although not shown in the figure, when a parallel beam is not used, an image relay lens is used to sequentially transmit the light source image in order to prevent reflection on the container wall due to the spread of light due to diffraction. system can be used. S is a sample, and the light from the light source LS passes through this sample and becomes the detected light to the Faraday rotation element.
It is configured to be incident on the FR. (1) Case of a single spectrum First, consider the case of a single spectrum. The sweep speed k of the magnetic field strength H is kept constant. This sweep speed k does not necessarily have to be constant. The electric field strength of the light is a function A of the Werde constant V depending on the wavelength.
It can be expressed as (V). If the angular frequency of the change in light intensity when the length of the Faraday rotation element FR is l is ω=kVl, then the electric field of the light of the polarization component orthogonal to the incident light can be expressed as E=Asinωt. Therefore, the output of the photoelectric detector D1 of the spectrophotometer can be obtained in the form of A ² (V) sin ² wt in response to changes in the magnetic field strength. The output of the detector D2 is a signal obtained by subtracting the output of the detector D1 from A ² (V). (2) Case of two line spectra Next, the case where there are two line spectra will be explained. The amount of light transmitted from the analyzer P2 after passing through the Faraday rotation element FR is A ² (V) sin ²
(VHL). V and L are the Weerdet constant and length of the Faraday rotation element FR,
Since the Welde constant V has a different value for each wavelength, if the magnetic field is swept at a constant speed k (H = kt), the output of the photoelectric detector D1 will be A in response to changes in the magnetic field strength. It can be expressed as ² (V) sin ² ωt (however, ω=kVl). Since the Werde constant V is a function of the respective wavelengths λ ₁ and λ ₂ , the obtained signals can be expressed as A ² (V ₁ ) sin ² ω ₁ t (ω ₁ =kV ₁ l) and A ² (V ₂ )
If it can be decomposed into terms of sin ² ω ₂ t (ω ₂ = kV ₂ l), we can obtain a spectral distribution expressed by blotting the Werde constant V on the horizontal axis and the intensity on the vertical axis, that is, intensity information based on the Werde constant. It will be done. Figure 8 shows two wavelengths λ ₁ = 248 nm and λ ₂ =
Indicates the output light amount of analyzer P ₂ when measuring 546.1 nm light, A indicates the combined light amount of two wavelengths, and B
shows the spectrum of each wavelength component obtained by inverse Fourier transform of this combined light quantity. Furthermore, Figure 3A
The horizontal axis is the time axis, the vertical axis is the light amount, and the state within the sweep time T at a constant magnetic field sweep speed is shown. The Werde constant V ₁ at 20°C at the wavelength λ ₁ = 248 nm is 1084 min/cm・Oe, and λ ₂ =
The Werde constant V ₂ at 546.1nm is 0.0156min/cm・Oe
This is a small value of about 1/7. In this way, the Weerde constants V ₁ and V ₂ differ by a factor of about 7, so the composite output (Fig. 3A) is a waveform that is a superposition of sin ² (VHl) functions whose periods differ by a factor of 7. Therefore, if the information on this combined light amount (Fig. 3A) can be separated into a function of sin ² (VHl) for the two wavelengths λ ₁ and λ ₂ , the spectrum expressed by the Werde constant for both wavelengths λ ₁ and λ ₂ ( intensity information based on the Werde constant) can be obtained. By the way, sin ² (VHl) can be written as sin ² (VHl) = 1/2 (1-cos2VHl), so the spectrum for the above two wavelengths λ ₁ and λ ₂ is A ² (V ₁ ) sin ² (V ₁ Hl) + A ² (V ₂ ) sin ² (V ₂ Hl) =
1/2 [A ² (V ₁ ) + A ² (V ₂ )] - 1/2 [A ² (V ₁ ) cos2
V ₁ Hlt + A ² (V ₂ ) cos2V ₂ Hlt], which is the sum of the Weerdet constant function of the light intensity of the two wavelengths λ ₁ and λ ₂ and ordinary trigonometric functions, so basically it is The combined light amount can be separated into two wavelength components using the usual Fourier transform spectrum method. If we can obtain the spectrum of the function of Weerde's constant by Fourier transforming the combined light intensity in this way, then since the Weerde's constant for each wavelength is a known value, we can obtain the spectral intensity distribution by finding the wavelength or frequency from the Weerde's constant. becomes. The Fourier transform and the Weerdet constant/wavelength or frequency conversion as described above are processed and performed by the computer CP. That is, these photoelectric detectors D
1, the light intensity information of the polarized light component corresponding to the magnetic field intensity from D2 is supplied to the Fourier transform means M1 in the computer CP, where it is converted into intensity information expressed as a function of Weerde's constant, and this intensity information is converted into Weerde's constant/ It is supplied to wavelength or frequency conversion means M2 from which wavelength or frequency spectral information is derived. These Fourier transform means M1 and Werde's constant/wavelength or frequency transform means M2 can be easily constructed by a combination of ordinary circuits. In this case, the Welde constant/wavelength or frequency conversion may be calculated based on a constant table without performing computer processing. (3) Case of continuous spectrum Next, the case of continuous spectrum will be explained. Fourth
The figure is a waveform diagram showing the amount of light when sweeping a curved magnetic field at a constant sweep speed from analyzer P2, with the horizontal axis representing the time axis and the vertical axis representing the amount of light. These are a transmitted signal and a reflected signal when the analyzer P is arranged so that no signal is transmitted when the magnetic field is zero. This assumes that there is no loss in the medium of the Faraday rotation element FR considering an ideal case. As is clear from this figure, for a continuous spectrum, the periods gradually cancel each other out and asymptotically reach 1/2 of the amount of incident light, and the line spectral component becomes a periodic function. As is well known, the general relationship between these Fourier integral transforms and their inverse transforms is: Fourier integral transform g(ω)=1/2π~ ^∞ _-∞ e ^-i 〓 ^t (t) dt Its inverse transform ( t)=~ ^∞ _-∞ e ⁱ 〓 ^t g(ω)dω. Here, ω=kVl, and k is the sweep speed of the magnetic field H. Therefore, in this case ω is a function of the Weerde constant V, and the magnetic field H space and the Weerde constant V
It is nothing but a Fourier transform between spaces. still,
Here, the sweep speed of the magnetic field does not need to be constant as described above. Furthermore, in each of the above cases, only the direction in which the light travels, that is, the axial component, contributes to the magnetic field applied to the Faraday rotation element FR.
Although the presence of a radial component is acceptable, under conditions where the uniformity of the magnetic field determines the resolution, magnetic fields other than the axial component are undesirable because they become elements that darken the light source. (Effects of the Invention) As is clear from the above, the Faraday rotational Fourier spectroscopy of the present invention makes it possible to improve the mechanical precision of the components of a spectrophotometer or the components whose resolution is determined by mechanical vibrations. In addition to not containing any optical components, it is possible to perform spectroscopy using a method that does not include optical components to control the measurement wavelength range, making it possible to improve resolution and enable measurement wavelength ranges from the vacuum ultraviolet region to the far infrared region. It can be done. Furthermore, according to the spectroscopy method of the present invention, spectroscopy is possible regardless of whether the light is coherent or incoherent. Further, according to the spectroscopy method of the present invention, the spectroscopy is completed within the time required for one sweep of the magnetic field, so measurements can be made in an extremely short time. As mentioned above, one of the features of this invention is its extremely high resolution. In the case of a prism spectrometer that uses the difference in refractive index of materials, the above-mentioned
The refractive index of 248nm and 546nm is 1.5085.
1.4601, making use of a difference in refractive index of only 3.3%. Therefore, even if the device is made larger, it is impossible to increase the wavelength dispersion to a large extent. In the spectroscopic method of this invention, the wavelength range is limited only by the absorption of the Faraday element or polarizer, and it is possible to measure a wide range of ultraviolet, visible, and near-infrared wavelengths at once. Furthermore, since the resolution can be improved by increasing the number of Faraday rotations, the resolution can be easily improved by using a long cell as a Faraday rotation element. An attempt was made to draw the proximity line 3131.5485A (6 ³ P ₁ −6 ³
Consider the separation of D ₁ ) and 3131.8391A (6 ³ P ₁ −6 ¹ D) as an example. The Weerde constant at this wavelength is given by 0.0556, and the rate of change of the Weerde constant with wavelength is dV=-2dλ/λ ³ =-2Vdλ/λ=-1.03x10 ^-5 . Assuming the strength of the magnetic field is 10 ⁵ Oe,
The Faraday rotation of 52.5 m long pure water is 90 degrees, and the resolution of the two closely spaced line spectra, which have a slight wavelength difference of 0.2906〓, can be completely separated by an analyzer. . Furthermore, due to the irreversibility of Faraday rotation, the rotation of the plane of polarization is added regardless of the direction in which the light travels, so it is easy to shorten the length of the Faraday rotation element by performing multiple reflections within the multi-pass cell. becomes. Although the magnetic field applied to the Faraday rotation element is required to be uniform in the radial direction, the rotation angle of the polarization plane can be expressed as θ(λ)=~ ^l _O V(λ)H(x)dx, so Since the axial magnetic field is determined by the integral of the magnetic field strength along the optical path,
This is not a strict requirement. rather stabilized
If a HeNe laser or a mercury emission line with a well-known wavelength is measured simultaneously or alternately, the Faraday rotation angle θ is given by θ = VHl, and the Werde constant V is known. The magnetic field strength has been measured, and the Fourier transform spectrophotometer must be constantly calibrated. Such methods are very useful for maintaining and improving system stability. Furthermore, since Faraday rotation itself only affects light with components parallel to the lines of magnetic force, there are many advantages, such as deviations from the parallelism of light within the Faraday rotation element do not affect the results. According to the spectroscopy method of the present invention, spectroscopy can be performed in a very bright state. In other words, in spectrometers that use dispersive elements such as diffraction gratings and prisms, losses due to entrance and exit slits are large, and the system measures light while losing most of the light. The spectrophotometer used in the method can be said to be very bright. The light that has passed through the polarizer in the incident section as parallel light reaches the detector without undergoing any loss other than the loss due to absorption or scattering by the pure water of the Faraday rotating substance. When added, the separated mutually orthogonal components become the incident intensity, which is a mutually complementary signal, and can be obtained by inversely transforming the original spectrum from both, making it possible to use all transmitted light as information. It is a very bright spectroscopic system. Let's compare it with an interferometer, which is known as a bright spectroscopic system. It is clear that the fabrication of UV-visible spectrophotometers with variable length arms is very difficult in practice. Furthermore, in the case of Michelson interferometers and Fabry-Perot interferometers, there is a contradiction between widening the free spectral range and improving resolution, and it is difficult to measure a wide wavelength range with high resolution. This cannot be done without sacrificing performance. In addition, when considering normal usage such as absorption measurement, the light source is a normal white light source, so it is practically impossible to increase the coherent length, that is, the length that allows interference, to be more than a few meters. The scaling law between the size of and resolution improvement holds only within a certain range. (Example) Example based on simulation In order to measure actual emission and absorption spectra, the basic signal waveform of Faraday rotational Fourier spectroscopy and its spectrum expressed by the Weerde constant will be examined by computer simulation. The simplest spectroscopy of two line spectra was described above to explain the principle, so below we will discuss spectroscopy of white light, four line spectra existing at arbitrary intervals, and Lorentzian spectral shape. We will present simulation results regarding the spectroscopy of a near doublet with , and discuss the basic signal shape and its processing. (a) Completely white light For simplicity, if the light to be separated is a completely white light with respect to wavelength in Werdet constant space (V space), the Fourier transform term will be constant without the need for simulation, The analyzer output is constant at 1/2 of the input light intensity.
However, white light, whose intensity is constant regardless of wavelength in wavelength space, is deformed due to the wavelength dependence of Werde's constant. Assuming that the Werde constant is completely inversely proportional to the square of the wavelength, it can be considered that the input light intensity has a spectrum proportional to the square of the wavelength in V space. Therefore, the signal detected by the photoelectric detector, the spectrum expressed in V space by Fourier transform, and the normal spectrum with wavelength on the horizontal axis are shown in Figures A, B, and C, respectively. That's right. In FIGS. 5A, B, and C, the vertical axis plots the amount of light, and the horizontal axis plots time t, Werde's constant V, and wavelength λ, respectively. (b) Set of line spectra with irregular intervals As a simple example, consider the four line spectra shown in Figure 6B with arbitrary intensities and irregular intervals. It is common practice to place some type of pre-spectrometer in front of a Faraday rotating Fourier spectrophotometer to attempt to analyze a specific spectral population with high resolution while reducing background light. Naturally, overlapping signals of sin ² ωt with different amplitudes as shown in FIG. 6A are detected by the photoelectric detector. If the clearness ratio of the analyzer is infinitely large and does not transmit any orthogonal components, it is possible to convert the time-varying information to be detected into an accurate spectrum even if the sampling range is not very large. However, even if the extinction ratio of the analyzer you are actually using is slightly lacking in accuracy, it is possible to compensate for these factors that reduce the resolution of the device by widening the sampling range, and in this respect, the scaling rule can also be applied. is valid. The vertical axes of FIGS. 6A and 6B plot the amount of light, and the horizontal axes plot the time t and Werde's constant V, respectively. (c) Close doublet spectrum with finite linewidth When trying to measure using a high-resolution spectrometer, the line spectrum actually has a Doppler width, pressure width,
It has some kind of spectral shape due to its fine structure. If we are aiming for truly high-resolution spectroscopy, it is of great significance to accurately measure the shapes of these extremely narrow spectra. Proximity of Mercury 2
Using the example of heavy lines, we will explain how Faraday rotational Fourier spectroscopy has a high resolution, and how the scaling law is more effective than other spectroscopy methods, making it possible in principle to construct an ultra-high resolution spectrometer. The explanation regarding this point is as explained above. In order to examine the state of the signal in such a case, a hypothetical close double line spectrum was examined. For simplicity, the shape of the spectrum is assumed to be Lorentzian. In this case, FIG. 7A shows the transmitted light signal,
FIG. 7B shows the spectrum obtained by Fourier transformation. As is clear from these figures, on top of the periodic signal determined by the line spectrum, the signal has minute vibrations due to the base portion of the spectrum. The facts shown by these analyzes (a) to (c) indicate that it is Fourier spectroscopy itself, and one of the differences from infrared Fourier spectroscopy, etc., which uses ordinary light interference, is that , the white light component gradually approaches not the zero level but one-half of the amount of incident light. Another feature is that it does not utilize interference phenomena, so the signal detected by the photoelectric detector with a line spectrum becomes a periodic signal that does not attenuate, which is different from interferometry, which is limited by the coherent length. It's on. (Summary) In this invention, we have proposed Faraday rotation Fourier spectroscopy based on a new principle that is completely different from conventional spectroscopy using dispersive elements or interference spectroscopy. According to studies using pure water as a Faraday rotation element, the refractive index dispersion is basically several tens to several times larger than the refractive index dispersion.
It was found that spectroscopy can be performed with essentially high resolution by using the Welde constant, which has a wavelength dispersion of more than 100 times. In addition, since it is a system with little optical loss in principle and does not utilize interference effects, a large scaling law is established, and if a large-scale equipment is constructed and ultra-high resolution spectroscopic experiments are performed, it is possible to move from the ultraviolet to the near-infrared. It is expected that the accuracy of the spectrum in the outer region can be significantly improved. On the other hand, if an infrared Fourier spectrometer is used in combination, the optical spectroscopy from ultraviolet to infrared will be completely different from conventional methods, allowing the Fourier spectrometer to cover the entire range and perform consistent measurements. It became possible. Furthermore, Faraday rotation Fourier spectroscopy can be applied to optical, electrical, and
It can be performed only by a combination of computer techniques.
This is significant in that it is possible to produce a spectrometer with significantly improved performance without any particularly difficult technical innovations in the actual production of the device. 【table】

[Brief explanation of drawings]

Fig. 1 is a diagram showing a spectrophotometer to explain the principle of Faraday rotational Fourier spectroscopy of the present invention, Fig. 2 is a curve diagram showing the Weerde constant of water, and Fig. 3 A and B are diagrams of the spectrophotometer of the present invention. FIG. 4 is a curve diagram showing the combined light amount from the analyzer and the spectrum of wavelength components obtained from the analyzer for explanation, FIG. 4 is a curve diagram showing the light amount from the analyzer for explanation of this invention, and FIG. 5 A, B and C to FIG. 7A and B are curves showing computer simulation embodiments of the present invention, respectively. LS...Light source, CL...Lens system, P1...Polarizer, P2...Analyzer (or polarizer), FA...Faraday rotation element, PS...Power supply, D1, D2...
Photoelectric detector, CP...computer, M1...Fourier transform means, M2...Wuelde constant/wavelength or frequency conversion means.

Claims (1)

[Claims] 1. After polarizing the light to be detected, the light is guided into a Faraday rotating medium to which a magnetic field is applied, and the plane of polarization of the polarized light is rotated multiple times in the Faraday rotating medium to adjust the rotation angle depending on the wavelength. extracting a superimposed signal of wavelength information with different wavelengths, separating a specific polarized light component from the superimposed signal, detecting the light amount of the polarized light component corresponding to the magnetic field strength from the polarized light component, and determining the light amount based on the magnetic field strength. Faraday rotation Fourier spectroscopy, which comprises performing inverse Fourier transform to convert into intensity information based on the Welde constant, and converting the Welde information into wavelength or frequency spectrum information to separate the detected light.