WO2008041361A1 - Measurement device and measurement method using nuclear magnetic resonance method - Google Patents

Abstract

A measurement device (300) locally measures current at a particular position of a sample (115) by using the NMR method. The measurement device includes: a magnet (113) which applies an electrostatic field to the sample (115); a small-size RF coil (114) smaller than the sample (115), which coil applies an oscillation magnetic field for excitation to the sample (115) and acquires an NMR signal generated at a particular position of the sample (115); and a current calculation unit (303) which calculates a difference between the frequency of the NMR signal acquired by the small-size RF coil (114) and the frequency of the excitation oscillation magnetic field and calculates the current at the particular position of the sample (115) from the difference.

Description

 Specification

 Measuring apparatus and measuring method using nuclear magnetic resonance method

 Technical field

 [0001] The present invention relates to a measuring apparatus and a measuring method using a nuclear magnetic resonance method, and more particularly to a technique for measuring a current at a specific portion of a sample using the nuclear magnetic resonance method.

 Background art

 [0002] Non-patent Documents 1 and 2 include conventional methods for measuring the surface current distribution of a sample.

 [0003] In Non-Patent Document 1, when measuring the current distribution in the surface direction of a fuel cell, the electrodes are divided into “divided electrodes”, and the individual currents flowing through each divided electrode are insulated. A method of measuring is described.

 [0004] Non-Patent Document 2 describes a method of measuring the strength of a magnetic field using a Hall element. Here, the Hall element is an element having a characteristic that the electric resistance of the element changes according to the magnetic field strength applied to the element. In Non-Patent Document 2, this Hall element is moved closer to the fuel cell and spatially scanned to measure a spatial map of the magnetic field strength and analyze it as an inverse problem to obtain the current distribution. A method has been proposed.

 Non-patent document 1: Kazuo Onda et al., “Mechanical properties measurement of solid polymer fuel cell and analysis / measurement of current distribution”, Proceedings of the 1st 3rd Fuel Cell Symposium, 2006, p. 2 3 4-2 3 7

 Non-Patent Document 2: Masaaki Izumi, Yuji Goto, “Research and Development on Measurement Technology and Modeling of Polymer Electrolyte Fuel Cell”, NEDO Fuel Cell ■ Summary of Hydrogen Technology Development Interim Report Meeting, February 1999 2 Announced on 7th, p. 3 9-4 0

 Disclosure of the invention

 [0005] However, the methods described in Non-Patent Documents 1 and 2 described above have room for improvement in the following points.

[0006] First, in the split electrode method described in Non-Patent Document 1, a split electrode is incorporated. It is necessary to manufacture a fuel cell, and since it is actually measured with a device for measurement, there is a possibility that the measurement result may be different from the actual machine that does not use the divided electrode, and the reliability of the experimental data is improved There was room for. In addition, each time a new cell is designed and manufactured, the split electrodes must be designed and manufactured again, which is not practical in terms of increasing development costs.

 [0007] Further, in the method using the Hall element described in Non-Patent Document 2, the magnetic field generated by the current flowing in the electrode is measured, but the magnetic field strength is almost equal to the strength of the geomagnetism. It is a weak value. In order to accurately measure such a weak magnetic field strength, the Hall element is required to have high resolution and high reproducibility.

 [0008] Also, for example, when a Hall element is used for measurement of a fuel cell, the Hall element is sensitive to temperature changes, and is installed in and around a fuel cell that generates heat, and a magnetic field is generated by the Hall element. In order to measure the non-linearity of the element, the relationship between the current or resistance value flowing through the Hall element measured at each temperature and the applied magnetic field strength can be used as a calibration curve in advance and applied to the fuel cell. In this case, the temperature of the hole element itself must be measured with very high accuracy, and a very time-consuming method of calculating the magnetic field from the calibration curve must be taken. Furthermore, there is a problem that it is difficult to measure the true hall element temperature.

 [0009] For these reasons, the method using the Hall sensor has been studied, but it is still in the research stage and is far from practical use.

 [0010] As described above, with the conventional technique, it is difficult to locally measure the current distribution in the surface of the sample.

 [001 1] According to the present invention,

 A device that locally measures the current at a specific part of a sample using nuclear magnetic resonance,

 A static magnetic field application unit that applies a static magnetic field to the sample;

A small RF coil smaller than the sample, which applies an oscillating magnetic field for excitation to the sample and obtains a nuclear magnetic resonance signal generated at a specific location of the sample; Calculating a difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coil and the frequency of the excitation oscillating magnetic field, and calculating a current of the specific portion of the sample from the difference;

 A measuring device is provided.

[0012] Further, according to the present invention,

 An apparatus for acquiring a current distribution in a plane of a solid polymer electrolyte membrane of a fuel cell using a nuclear magnetic resonance method,

 A static magnetic field applying unit that applies a static magnetic field to the solid polymer electrolyte membrane; and an oscillating magnetic field for excitation applied to the solid polymer electrolyte membrane, and generated at a specific location of the solid polymer electrolyte membrane. A plurality of small RF coils that are smaller than the solid polymer electrolyte membrane,

 For the plurality of small RF coils, a difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coil and the frequency of the oscillating magnetic field for excitation is calculated, and from the difference, the solid polymer electrolyte membrane A current distribution acquisition unit for acquiring an in-plane current distribution;

 A measuring device is provided.

[0013] Further, according to the present invention,

 A method of locally measuring the current at a specific part of a sample using a nuclear magnetic resonance method.

 First, an excitation oscillating magnetic field is applied to a specific location of the sample placed in a static magnetic field using a small RF coil smaller than the sample, and a nuclear magnetic resonance signal generated at the specific location is acquired. Steps,

 A second step of calculating a difference between the frequency of the nuclear magnetic resonance signal acquired in the first step and the frequency of the excitation magnetic field, and obtaining a current at the specific location of the sample from the difference;

 A measurement method is provided.

[0014] In the present invention, an excitation oscillating magnetic field is applied locally using a small RF coil smaller than the sample, and emitted from the location where the excitation oscillating magnetic field is applied. The obtained nuclear magnetic resonance signal is obtained, and the current at a specific portion of the sample is obtained from the obtained nuclear magnetic resonance signal. By applying an oscillating magnetic field for excitation by limiting the region to be measured with a small RF coil, the local current in a predetermined region of the sample can be measured in a short time.

 [0015] In addition, the measurement accuracy can be improved by using a nuclear magnetic resonance signal with high frequency resolution when obtaining the current. In addition, by calculating the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation, and obtaining the current from the difference, the measurement using the Hall element described above with reference to Non-Patent Document 2 is performed. Compared with methods that use absolute values, the effects of changes in the environment around the element, such as the temperature environment, and the need for a calibration curve can be reduced, so measurement accuracy can be further improved.

 [001 6] Here, the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation can be specifically obtained as (i i) below with respect to (i) below.

 (i) The RF oscillator as an oscillating magnetic field for excitation has (remembered) “frequency that is the reference when no current is flowing”

 (i i) “Amount by which the frequency of the nuclear magnetic resonance signal is increased or decreased by the magnetic field formed by the flow of current”

 The above (i i) is measured, for example, as a phase change amount.

[001 7] At this time, when the current is not flowing, the frequency of the excitation oscillating magnetic field is equal to the frequency of the nuclear magnetic resonance signal under the static magnetic field created only by the magnet, and the current flows. Current flows and flows using the difference between the magnetic field generated and the frequency of the nuclear magnetic resonance signal measured under both the static magnetic field originally applied by the magnet. Measure the frequency difference from the phase difference when not.

 [0018] In the present specification, the "static magnetic field" is a completely stable magnetic field as long as the magnetic field is stable in time so that the nuclear magnetic resonance signal and current can be stably acquired. It may not be a magnetic field, and there may be some variation within that range.

[001 9] In the present invention, the obtained current is expressed as a spatial distribution. Therefore, it can be expressed as the current density divided by the area where the current flows.

 [0020] The measurement apparatus of the present invention may further include a detection unit that detects a real part and an imaginary part of the nuclear magnetic resonance signal, and the current calculation unit is detected by the detection unit. The difference between the frequency of the nuclear magnetic resonance signal and the frequency of the excitation oscillating magnetic field may be calculated using the imaginary part and the imaginary part.

 Further, in the measurement method of the present invention, in the second step, the real part and the imaginary part of the nuclear magnetic resonance signal are detected, and the frequency of the nuclear magnetic resonance signal and the frequency using the real part and the imaginary part are detected. The difference from the frequency of the oscillating magnetic field for excitation can also be calculated.

 As a result, the frequency difference can be obtained more simply and reliably.

 [0021] Note that, from the viewpoint of the difference in frequency when the amount of phase change at a certain time interval is converted per unit time, using the detected real part and imaginary part, the excitation vibration magnetic field is used as a reference. The amount of change in phase at a certain time interval of the nuclear magnetic resonance signal or the difference between the two frequencies may be calculated.

 [0022] In the present invention, for each sample type, a storage unit that stores information indicating the correlation between the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the excitation vibration magnetic field and the current, for example, calibration curve data The current calculation unit may acquire the information corresponding to the sample to be measured from the storage unit, and calculate the current based on the information.

 [0023] The sample to be measured in the present invention can be, for example, a film.

 At this time, it is possible to grasp the local current in the film.

 [0024] Further, the measurement apparatus of the present invention includes a plurality of the small RF coils, and the plurality of small RF coils apply the excitation oscillating magnetic field to a plurality of locations of the sample, and the nuclear magnetic resonance. A signal may be acquired, and the current calculation unit may be configured to calculate currents at the plurality of locations of the sample.

[0025] By doing this, it is possible to measure multiple points of current simultaneously with a simple configuration. For example, if the sample is a film, information on the current distribution of the film can be obtained. Duplicate The number of small RF coils can be arranged arbitrarily, and can be arrayed according to the shape of the object to be measured.

 [0026] Here, the small RF coil applies, for example, the pulsed oscillating magnetic field, and also outputs a FID (Free Inductive Decay) signal corresponding to the oscillating magnetic field for excitation. The current calculation unit can acquire the real part and the imaginary part of the FID signal. At this time, it is only necessary that the magnetization vector is excited with an excitation pulse that can obtain a significant FID signal compared to noise, and the angle at which the excitation pulse excites the magnetization vector (the direction of the static magnetic field is changed). The angle of tilting as a reference) is arbitrary.

 By making this angle arbitrary, “It is possible to shorten the recovery time of the magnetization vector related to the relaxation time constant and to irradiate the excitation pulse with a shorter repetition time and to measure the current distribution in a short time. Is possible.

 [0027] Further, the small RF coil can apply an excitation oscillating magnetic field in the following sequence, for example, and can also acquire an echo signal corresponding to the excitation oscillating magnetic field.

 (a) 90 ° pulse, and

 (b) A 180 ° pulse applied after the elapse of the pulse of (a).

[0028] The excitation oscillating magnetic field is a pulse sequence including the above (a) and (b), and the current calculation unit uses the echo echo method to obtain the real part and imaginary part of the echo signal. The phase of the signal can be converged. Also, as will be described later, measurement errors due to magnetic field inhomogeneities can be effectively reduced. For this reason, the measurement accuracy of the real part and the imaginary part of the nuclear magnetic resonance signal can be further improved. For example, in the second step, the real part and the imaginary part of the echo signal are detected, and the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the excitation oscillating magnetic field is calculated using the real part and the imaginary part. You may calculate it.

[0029] In this specification, "FID signal" and "echo signal" are nuclear magnetic resonance signals that correspond to the excitation oscillating magnetic field and can detect the real part and the imaginary part. Any signal that functions as a signal may be used.

 [0030] In addition, the "pulse sequence" in the present invention is a sequence that defines a timing diagram for setting a time at which an excitation oscillating magnetic field is applied and its interval. Here, the timing diagram also includes a procedure table for performing necessary operations in time series.

 [0031] In addition to the above pulse sequence, another sequence in which a step of applying a 180 ° pulse is added at a time just before the 90 ° pulse (a) may be executed. By comparing the intensity of the N MR signal acquired with the 90 ° pulse (a) with the intensity of the N MR signal acquired by selecting the time at the 1 80 ° pulse (b) as appropriate, the irradiation from the RF coil It can be determined whether the intensity of the oscillating magnetic field for excitation corresponds exactly to 90 ° and 1800 °. The intensity of the two pulses is 1 to 2, or the irradiation energy is 1 to 4, or the pulse application time is 1 to 2, and the magnetization vector is excited to 90 ° and 1 80 ° respectively. This is an important factor in improving the accuracy and reproducibility of measured values. As a result, even if the relationship between the two pulses becomes inappropriate due to device abnormality or immature adjustment, the abnormality can be detected before the measurement is performed, and the measured value can be made more probable. .

 [0032] In the present invention, an RF signal generation unit that generates an RF signal for generating the excitation oscillating magnetic field in the small RF coil, and an echo signal obtained by the small RF coil are detected, and An echo signal detection unit that sends an echo signal to the current calculation unit; and a branch unit that connects the small RF coil, the RF signal generation unit, and the echo signal detection unit; and the small RF coil and the RF A switch circuit that switches between a state in which the signal generation unit is connected and a state in which the small RF coil and the echo signal detection unit are connected can be provided.

[0033] By doing this, the loss of the excitation high-frequency pulse signal applied to the sample from the small RF coil is reduced, and as a result, the pulse angle of 90 ° pulse and 1 80 ° pulse can be accurately controlled. It becomes. In addition, the pulse sequence of the excitation oscillating magnetic field may include the following (a), (b), and (c).

 (a) 90 ° pulse,

 (b) 1 80 ° pulse applied after the time of pulse in (a), and

(c) n 1 80 ° pulses that start after time 2 of pulse (b) has elapsed and are printed at time 2 intervals.

 N is a natural number.

[0035] By using the pulse sequence consisting of (a) to (c) above, the current at a specific part of the sample is measured using an echo signal corresponding to the pulse (b) or (c), and Using a plurality of echo signals corresponding to the pulses of b) and (c), the amount of the protic solvent in the sample at the specific location can be measured.

 [0036] At this time, the measurement apparatus of the present invention includes a solvent amount calculation unit that calculates the amount of the protonic solvent in the sample based on the nuclear magnetic resonance signal acquired by the small RF coil, A switching unit that switches between a first measurement mode for measuring current and a second measurement mode for measuring the amount of the protonic solvent in the sample, and when in the first measurement mode, the current calculation The unit calculates the current at the specific location of the sample based on the difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coil and the frequency of the oscillating magnetic field for excitation, and the second measurement mode The solvent amount calculation unit may calculate the amount of the protonic solvent at the specific location in the sample based on the nuclear magnetic resonance signal acquired by the small RF coil. so That.

 [0037] In this way, by switching between the first and second measurement modes, it becomes possible to measure the local amount of the protic solvent in the sample in a short time in addition to the current measurement.

[0038] More specifically, in the second measurement mode, the small RF coil acquires an echo signal corresponding to the excitation oscillating magnetic field, and the solvent amount calculation unit calculates from the intensity of the echo signal. , Calculate the T ₂ relaxation time constant and calculate Τ ₂ relaxation time constant From the number, the amount of the protic solvent at a specific location in the sample can also be calculated.

 [0039] Thus, in the measurement method of the present invention, two physical quantities can be measured for a specific portion of the sample by a single pulse sequence in one measurement apparatus. For example, in the present invention, in the first step, the small RF coil applies the excitation oscillating magnetic field in a pulse sequence including (a), (b) and (c), and An FID signal corresponding to the pulse of (a) or an echo signal corresponding to the pulse of (b) or (c) is acquired, and in the second step, the FID signal corresponding to the pulse of (a) or the The difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation is calculated using the real part and the imaginary part of the echo signal corresponding to the pulse of (b) or (c). ) And above

(c) The T ₂ relaxation time constant is calculated from the intensity of a plurality of echo signals corresponding to the pulse of (c), and the proton solvent at the specific location in the sample is calculated from the calculated 前 記₂ relaxation time constant. The amount of can also be calculated. More specifically, the above

The current can be measured with the first pulse-corresponding signal in (b), and the moisture content can be measured using the subsequent n-pulse-corresponding signal group in (c). The measurement of the two physical quantities may be performed at the same time, or may be performed at different timings such as alternating. For example, in the second step, the acquisition of an echo signal for calculating the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation, and the echo signal for calculating the T ₂ relaxation time constant Acquisition may be performed alternately. Further, in the above first Nisutetsu flop, the nuclear magnetic and acquisition of echo signals for calculating the difference between the frequency of the resonance signal and the frequency of the excitation oscillating magnetic field, T ₂ relaxation time E co one for calculating the constants Signal acquisition may be performed simultaneously.

[0040] In the above configuration, a small RF coil is used to (i) apply an excitation vibration magnetic field locally, and (ii) obtain an echo signal emitted from a location where the excitation vibration magnetic field is applied. The T ₂ relaxation time constant (lateral relaxation) Sum the time constant) and measure the water content based on this. Since the pulse-echo method is applied with a small RF coil to limit the area to be measured, the local water content can be measured in a short time.

[0041] Further, in the present invention, the 90 ° pulse is in the first phase, and the n number of the 180 ° pulses are in the second phase shifted by 90 ° from the first phase. Can do. In an actual measurement system, magnetic field inhomogeneities of the static magnetic field and excitation oscillating magnetic field occur, which may cause measurement errors in the T ₂ relaxation time constant. The pulse sequence with the above configuration is 90 ° with the first phase as a 180 ° pulse. Since the one in the shifted second phase is used, by applying a 180 ° pulse, the nuclear magnetization is reversed in the rotating coordinate system, and this causes a measurement error factor due to the above magnetic field inhomogeneity. It will be resolved. Second phase 1 8 0 ° pulses, since it is periodically applied, so that every time the measurement error factors are eliminated, it is possible to reliably obtain accurate T ₂ relaxation time constant.

[0042] In the present invention, a plurality of the small RF coils are provided, and in the second measurement mode, an excitation oscillating magnetic field is applied to a plurality of locations of the sample and an echo signal corresponding to the excitation oscillating magnetic field is acquired. And it is good also as a structure which calculates the ₂ relaxation time constant with respect to several places of the said sample, and calculates | requires the moisture content in the said several places of the said sample based on the said ₂ relaxation time constant. Furthermore, after obtaining the moisture content at the plurality of locations, the moisture content distribution of the sample may be presented based on the moisture content at the plurality of locations of the sample.

[0043] Further, in the measurement apparatus of the present invention, a gradient magnetic field application unit that applies a gradient magnetic field to a sample, and a proton in the sample based on a nuclear magnetic resonance signal acquired by the small RF coil. A mobility calculation unit for calculating the mobility of the organic solvent, and a first measurement mode for measuring the current of the sample and a third measurement mode for measuring the mobility of the protonic solvent in the sample. A switching unit; and in the third measurement mode, the small RF coil applies the excitation vibration magnetic field to the sample, and the nuclear magnetic resonance corresponding to the excitation vibration magnetic field and the gradient magnetic field. The signal is acquired, and the mobility calculation unit supports different gradient magnetic fields. The mobility of the specific portion of the sample may be calculated based on the information of the nuclear magnetic resonance signal obtained as described above.

 [0044] In the measurement method of the present invention, a third step of applying an excitation vibration magnetic field and a gradient magnetic field to a specific portion of the sample and acquiring a nuclear magnetic resonance signal generated at the specific portion; Based on the information of the nuclear magnetic resonance signal obtained in the first step and the information of the nuclear magnetic resonance signal obtained in the third step, the mobility of the specific portion of the sample is calculated. And in the first step and the third step, a local magnetic field is applied to a specific portion of the sample using the small RF coil, and a nucleus is formed from the specific portion. A magnetic resonance signal is acquired, and in the first step, application of a gradient magnetic field to the sample is executed according to a predetermined pulse sequence; in the third step, the first step It is also possible to perform the application of the gradient magnetic field of step with different sizes according to a predetermined pulse sequence

In this way, by switching between the first and third measurement modes, it becomes possible to measure not only the current but also the mobility of the protonic solvent at a specific location in the sample.

 [0046] In the above configuration, the small RF coil is used to (i) apply the excitation oscillating magnetic field and gradient magnetic field locally, and (ii) apply the excitation oscillating magnetic field and gradient magnetic field. The nuclear magnetic resonance signal emitted from the sample is acquired, and the mobility at a specific location of the sample is measured from the NMR signal obtained corresponding to different gradient magnetic fields. Since the spin echo method and gradient magnetic field NMR method are applied by limiting the measurement target area with a small RF coil, the local mobility of the protonic solvent in a predetermined region of the sample can be reduced in a short time. It can be measured.

[0047] The "mobility" measured by the present invention refers to a physical property value representing the movement of a protonic solvent in a sample. Such physical property values include parameters such as self-diffusion coefficient and mobility (movement speed). Main departure According to Ming, one of these parameters can be obtained.

 [0048] In the present invention, "different gradient magnetic fields" includes a case where one gradient magnetic field is zero, that is, measurement performed without applying a gradient magnetic field. For example, the first step may be performed without applying a gradient magnetic field.

 [0049] In the present invention, the gradient magnetic field application unit may take various modes. For example, it can be a gradient magnetic field application coil arranged away from a small RF coil, or a planar coil provided in the same plane as the small RF coil. Alternatively, a pair of gradient magnetic field application coils arranged with a small RF coil in between may be used. Alternatively, these configurations may be arbitrarily combined.

 [0050] For example, in the present invention, the pair of gradient magnetic field applying coils may have a substantially half-moon planar shape, and half-moon strings may be arranged to face each other toward the small RF coil. In this way, high-precision local measurement can be performed while saving space. In this specification, a substantially half-moon shape means that a pair of planar coils have a chord-like straight line region, and a gradient magnetic field that inclines in a direction perpendicular to the straight line region by arranging them in a sample. The configuration can be applied, and if the application of such a gradient magnetic field is possible, the plane shape of the moon shape of the coil may be larger or smaller than half a month.

 [0051] In the present invention, the sample may include a matrix made of a solid or a gel. At this time, the solvent amount calculation unit and the mobility calculation unit can be configured to calculate the amount and mobility of the protonic solvent contained in the matrix, respectively. Examples of such a sample include a film containing moisture, for example, a solid electrolyte film used for a fuel cell or the like.

[0052] Further, in the present invention, by adopting a configuration capable of measuring the amount of protonic solvent or mobility at a specific location in the sample in addition to the electric current, the phenomenon occurring at the specific location of the sample can be varied. It becomes possible to grasp it. In particular, when applied to the measurement of solid electrolyte membranes in fuel cells, it is possible to accurately grasp the current in the power generation state and the presence of the protonic solvent. [0053] Further, in the present specification, the protonic solvent refers to a solvent that dissociates itself to generate proton. Protonic solvents include, for example, water; alcohols such as methanol and ethanol;

 Carboxylic acids such as acetic acid;

 Phenol; and

 Liquid ammonia is mentioned. Among these, water and alcohol are solvents that can more easily measure the mobility or the amount of solvent in the present invention.

[0054] It should be noted that any combination of these components, or a conversion of the expression of the present invention between methods, devices, etc. is also effective as an aspect of the present invention.

[0055] As described above, according to the present invention, it is possible to locally measure a current at a specific portion of a sample.

 Brief Description of Drawings

[0056] FIG. 1 is a flowchart showing a procedure for measuring current in the embodiment.

 FIG. 2 is a diagram for explaining a compensation function of the CPMG method.

 FIG. 3 is a flowchart showing a procedure for measuring a water content in the embodiment.

 [Fig. 4] A diagram for explaining the principle of acquiring the N MR signal by the spin echo method.

FIG. 5 is a diagram showing an example of a pulse sequence for self-diffusion coefficient measurement.

 FIG. 6 is a flowchart showing the procedure for measuring the self-diffusion coefficient in the embodiment.

FIG. 7 is a diagram showing a schematic configuration of a measuring apparatus according to an embodiment.

 FIG. 8 is a diagram showing an example of an LC circuit that performs application of an oscillating magnetic field for excitation and detection of an NMR signal of the measurement apparatus in the embodiment.

 FIG. 9 is a diagram showing a configuration of a switch unit of the measuring apparatus in the embodiment.

 FIG. 10 is a diagram for explaining a phase difference shift of an N M R signal.

 FIG. 11 is a perspective view showing an arrangement example of a plurality of small RF coils of the measuring apparatus in the embodiment.

FIG. 12 is a diagram showing a configuration of an output unit of the measurement apparatus in the embodiment. FIG. 13 is a diagram showing a schematic configuration of a measuring apparatus according to an embodiment.

 FIG. 14 is a diagram showing a configuration of a control unit of the measurement apparatus in the embodiment.

 FIG. 15 is a flowchart showing a measurement procedure in the embodiment.

 FIG. 16 is a diagram showing a schematic configuration of a measuring apparatus according to an embodiment.

 FIG. 17 is a diagram showing a configuration of a G coil of the measuring apparatus in the embodiment.

 FIG. 18 is a flowchart showing a measurement procedure in the embodiment.

 FIG. 19 is a diagram showing a schematic configuration of a measuring apparatus in an embodiment.

 FIG. 20 is a diagram for explaining a method of measuring current in the example.

 FIG. 21 is a view showing a sample used in an example.

 FIG. 22 is a diagram showing an echo signal in the example.

 FIG. 23 is a diagram showing a temporal change in phase difference in the example.

 FIG. 24 is a diagram showing an echo signal in the example.

 FIG. 25 is a diagram showing a temporal change in phase difference in the example.

 FIG. 26 is a diagram showing an echo signal in the example.

 FIG. 27 is a diagram showing a temporal change in phase difference in the example.

 FIG. 28 is a diagram showing the relationship between current and frequency shift in the example.

 FIG. 29 is a diagram showing an echo signal in the example.

 FIG. 30 is a diagram showing a temporal change in phase difference in the example.

 FIG. 31 is a diagram showing the relationship between current and frequency shift in the examples.

 FIG. 32 is a diagram for explaining a method of measuring current in the example.

 FIG. 33 is a diagram showing a part of a configuration of a current measuring apparatus in an example.

 FIG. 34 is a diagram for explaining a method of measuring current in the example.

 FIG. 35 is a diagram showing a relationship between current and frequency shift in the example.

 FIG. 36 is a diagram showing a relationship between current and frequency shift in the example.

 FIG. 37 is a diagram showing a relationship between current and frequency shift in the example.

 FIG. 38 is a diagram showing a relationship between current and frequency shift in the example.

 FIG. 39 is a diagram showing a relationship between current and frequency shift in the example.

FIG. 40 is a diagram showing a relationship between current and frequency shift in the example. FIG. 41 is a diagram showing a configuration of a measurement apparatus in the embodiment.

 FIG. 42 is a diagram showing a time change of the F I D signal in the example.

 FIG. 43 is a diagram showing a time change of the phase difference in the example.

 FIG. 44 is a diagram for explaining the time change of the phase difference in the example.

 FIG. 45 is a diagram showing the time change of the F I D signal in the example.

 FIG. 46 is a diagram showing a time change of the phase difference in the example.

 FIG. 47 is a diagram showing a temporal change in phase difference in the example.

 FIG. 48 is a diagram showing a time change of the F I D signal in the example.

 FIG. 49 is a diagram showing a temporal change in phase difference in the example.

 FIG. 50 is a diagram showing a temporal change in phase difference in the example.

 FIG. 51 is a diagram showing a time change of the F I D signal in the example.

 FIG. 52 is a diagram showing a temporal change in phase difference in the example.

 FIG. 53 is a diagram showing the time change of the phase difference in the example.

 FIG. 54 is a diagram showing the relationship between current and frequency shift in the example.

 FIG. 55 is a diagram showing directions of a static magnetic field H 0 and a magnetic field H i used for the analysis in the example.

 FIG. 56 is a perspective view showing the positions of the copper plate, RF detection coil, and water sample in the magnetic field analysis in the example.

 FIG. 57 is a diagram showing a coordinate system used for magnetic field analysis in an example.

 FIG. 58 is a diagram showing an analysis result in the example.

 FIG. 59 is a diagram showing the relationship between the frequency shift amount Δω and the current I in the example. FIG. 60 is a diagram showing the configuration of the small surface coil used in the example.

 61 is a diagram showing the relationship between the current I and the frequency shift amount Δω in the example. FIG. 62 is a cross-sectional view showing a schematic configuration of Μ Ε in the example.

63 is a diagram showing current flowing through the collector electrode and the PEM in the example. FIG. FIG. 64 is a diagram showing an analysis result of a frequency shift amount Δω in the example. FIG. 65 is a diagram showing a configuration of a small surface coil used in Examples.

 FIG. 66 is a diagram showing a coil placed on a force-pong mesh in an example.

FIG. 67 is a diagram showing a measurement result of a frequency shift amount Δω in the example.

 FIG. 68 is a cross-sectional view showing a schematic configuration of Μ Ε に used in the analysis in this example.

FIG. 69 is a diagram showing an analysis result of a frequency shift amount in the example.

 FIG. 70 is a cross-sectional view showing a schematic configuration of M EA in an example.

 FIG. 71 is a diagram showing a frequency shift obtained by measurement and analysis in the example.

 FIG. 72 is a cross-sectional view showing the arrangement of M EA and small coils in the example.

 FIG. 73 is a diagram showing the measurement timing of P G S E and C P M G in an example.

FIG. 74 is a diagram showing a time change of a current flowing through M E A in the P G S E measurement of the example.

 FIG. 75 is a diagram showing the time change of the voltage applied to M E A in the P G S E measurement of the example.

 FIG. 76 is a diagram showing the measurement result of the frequency shift amount and the analysis result of the frequency shift amount in the PGSE measurement of the example.

 FIG. 77 is a diagram showing the time change of the frequency shift amount in the PGSE measurement of the example.

CO

 FIG. 78 is a diagram showing temporal changes in echo signal intensity measured by the anode coil in the CPMG measurement of the example.

 FIG. 79 is a diagram showing temporal changes in echo signal intensity measured by the cathode side coil in the CPMG measurement of the example.

 FIG. 80 is a diagram showing the relationship between the water content in P EM and the signal intensity in the example.

[Fig.81] Phenomena occurring in PEM during water electrolysis operation of MEA in Example It is a figure explaining.

 BEST MODE FOR CARRYING OUT THE INVENTION

 Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, common components are given the same reference numerals, and explanations are omitted as appropriate.

[0058] (Measurement principle)

 First, the measurement principle of a current measurement method in an embodiment described later will be described with an example. Hereinafter, the current measurement mode is also referred to as a first measurement mode.

 [0059] (A) Current measurement

 Figure 1 is a flowchart showing an overview of the current measurement procedure. In Fig. 1, the following steps are performed sequentially, and the current at a specific part of the sample is measured locally using the nuclear magnetic resonance (N M R) method. In the NMR method, the motion of nuclear magnetization can be detected as an NMR signal by the spin resonance phenomenon of a nucleus placed in a magnetic field. If the NMR signal is measured using a small surface coil (small RF coil), local NMR measurement around the coil becomes possible.

 Step 3 0 1: Place the sample in the space where the magnet is placed and apply a static magnetic field to the sample.

 Step 3 0 3: Apply a vibrating magnetic field for excitation to a specific part of a sample placed in a static magnetic field using a small RF coil that is smaller than the sample, and generate a nuclear magnetic resonance (NMR) signal generated at the specific part. get,

 Step 3 0 5: Calculate the difference between the frequency of the nuclear magnetic resonance signal acquired in Step 3 0 3 and the frequency of the oscillating magnetic field for excitation.

 Step 3 07: Obtain the current at a specific part of the sample from the difference obtained in Step 3 0 5, and

 Step 3 0 9: Then, output the result.

 Hereinafter, each of steps 3 0 3 to 3 0 7 will be described in detail.

[0060] (i) Step 3 0 3 (Application of high frequency pulse for excitation and acquisition of NMR signal) )

 In this step, a high frequency pulse applied to the measurement target nucleus in the sample is applied as an oscillating magnetic field for excitation. In addition, the NMR signal emitted from the measurement target nucleus in the sample is acquired by the nuclear magnetic resonance phenomenon caused by the oscillating magnetic field for excitation.

[0061] The NMR signal is specifically an echo signal corresponding to the excitation high-frequency pulse. It is preferable that the phase of the echo signal is converged so that the frequency difference in Step 3 0 5 can be obtained reliably. Further, it is preferable to apply the high-frequency pulse in a pulse sequence in which the phases of the echo signals are matched.

 A specific example of such a pulse sequence will be described later with reference to FIG.

 [0062] Further, the NMR signal is detected by separating the real part and the imaginary part by a phase sensitive detection method. As a result, the calculation of the frequency difference in step 300 is easily performed.

[0063] (i i) Step 3 0 5 (Calculation of frequency change)

 In this step, the difference (frequency shift) between the frequency of the NMR signal acquired in step 303 and the frequency of the oscillating magnetic field for excitation is obtained.

Specifically, the phase difference Δ0 is obtained by calculating the real part and the imaginary part of the echo signal acquired by the phase sensitive detection method. Then, the frequency shift Δω is converted as a phase difference Δø per unit time.

[0065] (i i i) Step 3 0 7 (Calculation of current)

 In this step, the current is calculated from the frequency difference Δω obtained in Step 3 0 5. Below, the calculation principle of the current will be explained.

[0066] If a current flows through the measurement target, a magnetic field in direct proportion to the current j! "!" Is generated according to Bio-Savart's law. The magnetic field strength is inversely proportional to the distance r ⁿ (where n is a power) between the position where the current flows and the measurement position.

On the other hand, in the nuclear magnetic resonance phenomenon, the resonance frequency ω of nuclear magnetization is directly proportional to the magnetic field strength H. If the magnetic resonance signal is acquired by the small detection coil, the small detection coil In other words, the magnetic field strength H in the region measured by the magnetic field is indirectly measured as the magnetic resonance frequency ω.

[0068] In a magnetic field vector H _Q that is stable both spatially and temporally created by a magnet, if a current is passed to create a magnetic field vector Η, the magnetic field strength Η at a certain position is It is expressed by the combined vector (Ho + Hj) of both. Since the magnetic field vector Ho is constant, when the resonance frequency ω of nuclear magnetization increases or decreases by Δω, the magnetic field strength at a certain position! ”!” Is related to the current j and the distance.

[0069] Therefore, for example, by acquiring in advance an experimental method or the like the relationship between the current j flowing in a specific part of the sample and the frequency difference Δω, the frequency difference Δω obtained in step 305 is obtained. The current j flowing through the sample can be obtained.

[0070] Furthermore, by arranging multiple small coils on the sample and measuring the increase / decrease Δω of the resonance frequency of the nuclear magnetization at multiple positions in the sample, the current j and the position r where it flowed are analyzed by inverse problem analysis. Can be obtained.

[0071] At this time, the NMR detection method has a frequency resolution of the order of p pm, and this makes it possible to capture changes in magnetic field strength with high resolution and high sensitivity. For example, when the frequency of the oscillating magnetic field for excitation is 43 MHz, a resolution of about 10 Hz can be obtained sufficiently.

[0072] A specific example of the excitation high-frequency pulse applied in step (i) above (i) is shown below.

 In actual measurements, magnetic field inhomogeneities due to sample and instrument characteristics may occur, and frequency differences may not be obtained accurately. Therefore, in the following embodiment, the spin echo method is used, and the high frequency pulse for excitation is, for example, a pulse sequence including a plurality of pulses including the following (a) and (b).

(a) 90 ° pulse and

 (b) 1 80 ° pulse applied after the time of pulse in (a)

[0073] By applying the excitation oscillating magnetic field according to the pulse sequences (a) and (b) above, the phase of the echo signal converges, and the measurement error due to such magnetic field inhomogeneity is effectively reduced. . In addition, the phase of the corresponding echo signal Since current can be suppressed, the current can be obtained more accurately. The reason for this will be described below with reference to FIG.

[0074] In Fig. 4, the resonance excited magnetization vector M- _y relaxes with time. The time change of the magnetic resonance signal actually observed at this time is relaxed by the spin-lattice relaxation time constant T and another time constant Τ ₂ * that cannot be expressed by the spin-spin relaxation time constant T ₂ alone. To go. This is shown immediately after the 90 ° excitation pulse as the time variation of the signal intensity at the bottom of Fig. 4. In general, the actually measured magnetic resonance signal intensity indicated by this wavy line decays rapidly, and its time constant Τ ₂ * is shorter than Τ ₂ . ( _{2) The} reason why the decay signal actually observed decays faster than the decay curve due to relaxation is because of the non-uniformity of the external static magnetic field created by the static magnetic field magnet, the magnetic field in the sample due to the magnetic properties and shape of the sample This is because a uniform magnetic field is not secured across the entire sample due to inhomogeneities. This time change of the actually measured magnetic resonance signal is called free induction decay, “Free Induction decay j” in English, and “FID” in English.

[0075] "Spin echo" is a method for correcting the phase shift due to magnetic field inhomogeneity as a sample or device characteristic. This is because the phase of the magnetization vector M is disturbed on the xy plane by applying a 180 ° excitation pulse with twice the excitation pulse intensity after the 90 ° excitation pulse. middle reverses the disturbance of the phase, after 2 Te time is a method of obtaining an echo signal to get on and converges the phase T ₂ decay curves on.

 [0076] When the direction along the static magnetic field is the Ζ direction for the sake of convenience, the 180 ° excitation pulse applied in (b) above may be either the X direction or the Y direction, either the 180 ° excitation pulse. Can be used.

It should be noted that after the elapse of time 2 in the above (b), a further 180 ° pulse may be applied and current measurement may be performed using an echo signal corresponding thereto. However, when performing current measurement using multiple echo signals, it is effective to irradiate the 180 degree excitation pulse multiple times in the Y-axis direction so that the strongest possible echo signal can be observed. The reason is that the magnetization vectors in Fig. 2 (a) to Fig. 2 (d) described later Is shown in the movement of Le.

 [0077] By adopting these methods, the phase of the magnetization vector can be converged and an echo signal as strong as possible can be obtained. With such an echo signal, the real and imaginary parts of the NMR signal can be detected with higher accuracy, and the amount of phase change from the reference frequency can be determined reliably.

 [0078] The current measurement principle has been described above.

 Next, the measurement principle of the distribution of protonic solvent in the sample and the movement of the protonic solvent (mobility) in the sample using the N MR method will be described in the case where the protonic solvent is water. An example will be described. These can be measured using a current measuring device, as described later in the fourth and fifth embodiments. In the following description, detailed description of the steps common to the current measurement described above will be omitted as appropriate.

 [0079] First, a method for measuring the amount of moisture will be described. The moisture content measurement mode is hereinafter also referred to as a second measurement mode.

 (B) Measurement of water content

In the following embodiments, more C PMG (Garr-Purcel l- Meiboom-Gi II) method described below, calculates the T ₂ (transverse) relaxation time constant, then the conversion table of the "T ₂ and water content" Use this to calculate the local moisture content of the sample and grasp the moisture distribution.

FIG. 3 is a flowchart showing an outline of moisture content measurement.

 In the water content measurement shown in FIG. 3 as well, as in the current measurement described above, first, the sample is placed in the space where the magnet is arranged, and a static magnetic field is applied to the sample (S 10 02). In this state, an excitation oscillating magnetic field (high frequency pulse) is applied to the sample via a small RF coil, and the corresponding N MR signal (echo signal) is acquired (S 104).

Next, a T ₂ relaxation time constant is calculated from this echo signal (S 10 06). And from the obtained T ₂ relaxation time constant, to measure the local water content in the sample (S 1 0 8). Specifically, to get the de-one data showing the correlation between the water content and the T ₂ relaxation time constant of the sample, from this data and the T ₂ relaxation time constant, contact to a particular location in the sample Find the local water content. After that, the result is output (S 1 1 0). By performing the above procedure (Step 104 to Step 110) via each small RF coil, the distribution of moisture content can be grasped.

[0082] Hereinafter, step 104 to step 108 will be specifically described.

 (i) Step 104 (Applying excitation RF pulse and acquiring N MR signal

)

The excitation high-frequency pulse in step 104 is preferably a pulse sequence composed of a plurality of pulses, and an echo signal group corresponding to the pulse sequence is acquired. In this way, the T ₂ relaxation time constant can be obtained accurately.

 [0083] The pulse sequence preferably includes the following (a), (b) and (c).

 (a) 90 ° pulse and

 (b) 1 80 ° pulse applied after the time of pulse in (a), and

(c) n 1 80 ° pulses (n is a natural number) that starts after the lapse of time 2 of the pulse in (b) and is applied at intervals of time 2

 (A) and (b) above are the same as (A) current measurement.

[0084] By applying the excitation oscillating magnetic field according to the pulse sequences of (a) to (c) above, the phase of the echo signal converges, and measurement errors due to such magnetic field inhomogeneities are effectively reduced. . In addition, since the variation in phase of the corresponding echo signal can be suppressed, the amount of water can be determined more accurately. The reason for this will be described with reference to FIGS. 2 (a) to 2 (d).

[0085] Hydrogen nuclei placed in a static magnetic field have a net magnetization vector in the direction along the static magnetic field (for convenience, the Z direction), and RF of a specific frequency (this is called the resonance frequency). By irradiating a wave from the outside in the X-axis direction perpendicular to the Z-axis, the magnetization vector tilts in the positive direction of the Y-axis, and a nuclear magnetic resonance signal (referred to as the NMR signal) can be observed. At this time, the excitation pulse in the X-axis direction irradiated to obtain the maximum intensity N MR signal is called a 90 ° pulse. And the magnetization vector After tilting in the positive direction of the Y axis with a 90 ° pulse, irradiate an external 1 80 ° excitation pulse in the `` axis direction '' after a while, and reverse the magnetization vector `` with the axis symmetric '' Let As a result, after two hours, the magnetization vector converges on the “positive direction” of the Υ axis, and an N MR signal with a large amplitude is observed.

 [0086] Since the magnetization vector is thus reversed "with the Y axis as the axis of symmetry", the following compensation function appears. Figures 2 (a) to 2 (d) illustrate the compensation function of the spin echo method. The coordinates shown in the figure are a rotating coordinate system.

 [0087] Consider P and Q as nuclear magnetization in a small region where the inhomogeneity of the static magnetic field is negligible in the sample. Let the magnetic field at P be stronger than the magnetic field at Q. At this time, as shown in Fig. 2 (a), when a 90 ° pulse is applied in the x'-axis direction, the nuclear magnetization of P and Q precesses from the same location (y'-axis) in the rotating coordinate system. As time passes, the phase of P advances from the phase of Q (Fig. 2 (b)).

 [0088] Therefore, when a 1 80 ° pulse is applied in the y'-axis direction after the 90 ° pulse has elapsed, the nuclear magnetization of P and Q rotates by 1 80 ° around the y'-axis, The arrangement is symmetric with respect to the y 'axis before application (Fig. 2 (c)).

 [0089] In this arrangement, the nuclear magnetization P, which had a more advanced phase, has a phase that is later than Q, so that both nuclear magnetizations reach the y 'axis at the same time. (Fig. 2 (d)).

 [0090] Since this relationship holds for nuclear magnetization in all regions of the sample, all nuclear magnetization gathers on the y 'axis at this time, resulting in a large N MR signal.

[0091] As described above, by first applying a 90 ° pulse in the χ'-axis direction and then applying an 80 ° pulse in the y'-axis direction, as shown in Fig. 2 (c), P, The nuclear magnetization of Q is reversed in the x 'y' plane. This reversal of the nuclear magnetization produces a good compensation function. For example, due to (a) non-uniformity of magnetic field, (b) non-uniformity of excitation pulse intensity irradiated by RF coil, the positions of P and Q are shifted to the position above or below the x 'y' plane. Even if the nuclear magnetization is reversed in the X 'y' plane. By rotating, the phase shift is compensated.

[0092] From the above, after two hours, the magnetization vector converges on the “positive direction” of the Y axis, and an echo signal having a large amplitude is observed. Furthermore, in the above (c), after 3 hours, the magnetization vector is irradiated with an external 180 ° excitation pulse in the “Y-axis direction” and converged again on the “positive direction” of the vertical axis. Then, an echo signal with a large amplitude is observed after 4 hours. In addition, continue to irradiate 1 80 ° pulse at the same interval of two. During this period, the peak intensities of the even-numbered echo signals ( ₂ ), (4), (6), and (3) are extracted, and fitted with an exponential function in Step 106, so that CP2 (horizontal) is relaxed by the CPMG method. It is possible to calculate a constant.

[0093] (ii) Step 1 06 (Calculation of T ₂ relaxation time constant)

( _{2) The} relaxation time constant can be accurately measured by using the spin echo method described above with reference to FIG. The intensity of the echo signal SSE when using spin echo is expressed by the following equation (A) when TR >> TE.

 [0094] [Equation 1] ヽ

 Formula (A)

j

j

 [0095] In the above formula (A), p is the density distribution of the target nuclide as a function of the position (X, y, z), TR is the 90 ° excitation pulse repetition time (from about 10 Oms to about 10 s), TE is the echo time (2ms, about 1 ms to 10 Oms), and A is a constant that represents the RF coil detection sensitivity and device characteristics such as amplifier.

[0096] In step 1 06, as described above, Step 1 04 plurality of echo signal groups riding on the obtained T ₂ decay curves on the (on 2, 4 Te, Te 6, ■ ■ ■) an exponential By fitting with Τ _{2, the} relaxation time constant Τ ₂ can be obtained from the above equation (Α).

 [0097] (i i i) Step 1 08 (Calculation of water content)

Returning to Figure 3, in step 108, the water content is calculated from the T ₂ relaxation time constant. The amount of water in the sample and the T ₂ relaxation time constant have a positive correlation, and the τ ₂ relaxation time constant increases as the amount of water increases. Since this correlation varies depending on the type and form of the sample, it is desirable to prepare a calibration curve for a sample of the same type as the sample to be measured whose moisture concentration is known in advance. In other words, it is desirable to measure the relationship between the moisture content and the τ ₂ relaxation time constant for a plurality of standard samples with known moisture content, and obtain a calibration curve representing this relationship in advance. By referring to the calibration curve created in this way, the amount of water in the sample can be calculated from the measured value of the τ ₂ relaxation time constant.

 Next, calculation of mobility will be described. The mobility measurement mode is also called the third measurement mode below.

 (C) Calculation of mobility

 In the following embodiment, by applying a gradient magnetic field and measuring the self-diffusion coefficient of water molecules by the PGS E (Pulsed-Gradient Spin-Echo) method, the mobility of local water molecules in the sample is calculated.・ To understand the mobility distribution of water molecules.

 [0099] After a specific nuclear spin in a liquid molecule is excited by magnetic resonance, a pair of gradient magnetic field pulses (pulsed gradient magnetic field) are applied with an interval of several Om s. N nuclei move due to Brownian motion or diffusion, and the phase of the nuclear spin does not converge, so the intensity of the N MR signal decreases. The self-diffusion coefficient of a specific molecular species can be measured by associating the gradient magnetic field pulse changed stepwise with the decrease in the intensity of the NMR signal. This is the principle of measurement of the self-diffusion coefficient by the PGS E method.

 FIG. 5 is a diagram showing an example of a PGS E sequence used for measuring the self-diffusion coefficient. In the sequence shown in FIG. 5, a pair of gradient magnetic field pulses Gz with the same application time and intensity are applied to the spin echo sequence described above with reference to FIG. N For example, a spin echo signal is acquired as the MR signal.

 [0101] The peak intensity S of the obtained N MR signal is the applied pulse gradient magnetic field intensity G z

Depending on [gauss / m], application time d, and pulse interval △, It is related to the z-direction self-diffusion coefficient D z [m ² / s] by the equation.

I n (S / S ₀ ) =-r ² D z A ² d G z ² (II)

[0102] In the above formula (II), S ₀ represents the normal N MR signal strength when G z = 0. D, △, and G z indicate the pulse width of the gradient magnetic field pulse, the time interval between the pair of gradient magnetic field pulses, and the magnetic field gradient (z direction) of the gradient magnetic field pulse, respectively. Also,: Γ represents the gyromagnetic ratio, which is a value intrinsic to the nucleus. For example, in the case of hydrogen nucleus 1 H, the gyromagnetic ratio 42. 577 X 1 0 ² [1 / ga

O “U S S ■ S”

 Note that FIG. 5 illustrates a sequence in the case of d = 1.5 ms Δ = 34.5 ms. For example, by applying a magnetic field to the sample in such a / Lus sequence, the self-diffusion coefficient D z can be stably calculated using the peak intensity S of the N MR signal.

 [0104] FIG. 6 is a flowchart for measuring the mobility of a specific portion of a sample using the PGS E method as described above, and includes the following steps.

 First, place the sample in a static magnetic field created by a magnet or the like, and apply the magnetostatic field to the sample. In this state, an oscillating magnetic field for excitation is applied to the sample according to a predetermined pulse sequence via the small RF coil, and the corresponding N MR signal is obtained via the small RF coil (S202).

 Next, an oscillating magnetic field for excitation and a gradient magnetic field are applied to the same region in the sample, and a corresponding N MR signal is acquired via a small RF coil (S 204).

 [0106] Fig. 6 shows a case where no gradient magnetic field is applied in step 202, but in step 202, a gradient magnetic field having a magnitude different from that in step 204 is applied in accordance with a predetermined pulse sequence. May be executed. At this time, for example, it is preferable to set the magnitude of the gradient magnetic field in step 202 to a value close to zero.

[0107] Next, the self-diffusion coefficient D is calculated from a plurality of N MR signals obtained by changing the gradient of the pulse gradient magnetic field stepwise (S206). In step 206 Later, based on the self-diffusion coefficient D calculated in step 206, other parameters indicating the mobility of water in the sample may be calculated. After that, the result is output (S208).

 By performing such operations (Step 202 to Step 208) via each small RF coil, the distribution of the self-diffusion coefficient can be grasped.

[0108] Details of each step will be described below.

 (i) Step 202 and Step 204 (Application of excitation oscillating magnetic field, application of gradient magnetic field and acquisition of N M R signal)

 In Step 202 and Step 204, an excitation oscillating magnetic field and a gradient magnetic field are applied to the sample according to a predetermined pulse sequence. The excitation oscillating magnetic field is a pulse sequence composed of a plurality of pulses, and the gradient magnetic field is a pair of pulse sequences corresponding to the excitation oscillating magnetic field.

[0109] Regarding the gradient magnetic field, as described above, in step 202, the gradient magnetic field is set to a value close to zero or close to zero, and in step 204, a predetermined gradient magnetic field is applied.

[0110] Further, the pulse sequence is preferably composed of the following (a) to (d).

 (a) 90 ° pulse of oscillating magnetic field for excitation,

 (b) Gradient magnetic field pulse that starts after the elapse of the pulse time of (a) and is applied for a certain time d,

 (c) 1 80 ° pulse of an oscillating magnetic field for excitation applied after the elapse of the time of the pulse of (a), and

 (d) Gradient magnetic field pulse that starts after elapse of the pulse time of (c) and is applied for a certain time d.

 However, when the gradient magnetic field is set to zero in step 202, the above sequence (b) is not performed.

[0111] More specifically, as shown in FIG. 5 described above, (b) the time to finish applying the gradient magnetic field pulse and (d) the time to start applying the gradient magnetic field pulse are as follows. , (C) 1 80 ° pulse (even though it has a width of 120 microseconds. Considering 60 microseconds at the center as the axis of symmetry), the time is equal ((34.5 m s- (1) 5 ms) / = Λ 6.5 ms), and (b) gradient magnetic field pulse application time d and (d) gradient magnetic field pulse application time d are both Make equal (d = 1.5 ms).

[0112] Then, the N MR signal corresponding to the pulse sequence is measured. The peak intensity S of the NMR signal is measured by the spin echo method. Specifically, as shown in Fig. 5, the peak intensity S of the echo signal that appears in time 2 is measured. The peak intensity S may be an average value of NMR signal intensities measured not only in the 2 hour NMR signal intensity but also in the surrounding time. This method can reduce variations in measured values caused by noise included in the NMR signal.

 [0113] In this way, by applying the gradient magnetic field stepwise and detecting the degree of decrease in the N MR signal corresponding to the increase in the magnetic field gradient, the self-diffusion coefficient D of the proton in the sample is obtained. Calculated.

 [0114] In step 204, the step of executing the excitation oscillating magnetic field and the gradient magnetic field according to a predetermined pulse sequence, and the step of acquiring the NMR signal corresponding to this pulse sequence are performed once or a plurality of times. Run once.

 [0115] (i i) Step 206 (Measurement of self-diffusion coefficient D)

 In step 206, the self-diffusion coefficient D of water at a specific location of the sample is obtained from the peak intensity of the NMR signal obtained in steps 202 and 204. The self-diffusion coefficient D of the proton is expressed by the above formula (I I) using the peak intensity S of the NMR signal obtained by the PGS E method.

[0116] from the N MR signal peak intensity S in the case of applying the N peak intensity of the MR signal S _Q and the gradient field G when applied with no gradient magnetic field G, using the above formula (II), the sample The self-diffusion coefficient D of the inner proton can be obtained. For example, the same location in the sample is measured by changing the magnitude of the gradient magnetic field G, and I n ( By plotting the relationship between S / S ₀ ) and 1 ² DA ² d G ² , the self-diffusion coefficient D can be obtained from the slope of the plot.

[0117] Note that (B) measurement of moisture content and (C) measurement of mobility described above may be performed by switching measurement modes. It is also possible to calculate the distribution of water molecule movement based on the water content calculated in each measurement mode and the water mobility.

 [0118] Hereinafter, a method for measuring a local current using the above-described measurement principle and an apparatus for realizing this method will be described more specifically.

 [0119] (First embodiment)

 FIG. 7 is a diagram showing a schematic configuration of the measuring apparatus 300 according to the present embodiment. Each component of the measuring apparatus 300 is realized by an arbitrary combination of hardware and software, mainly a CPU, a memory, and a program that implements the components shown in FIG. It will be understood by those skilled in the art that there are various variations in the implementation method and apparatus.

 [0120] Measuring device 300 is a device that locally measures the current at a specific location of sample 1 15 using the NMR method.

 Sample mounting table 1 1 6 for mounting sample 1 1 5

 A static magnetic field application unit (magnet 1 1 3) that applies a static magnetic field to sample 1 1 5, an oscillating magnetic field for excitation to sample 1 1 5, and N generated at a specific location on sample 1 1 5 Small RF coil 1 1 4 smaller than sample 1 1 5 to acquire MR signal, and

 The current calculation unit 303 calculates the difference Δω between the frequency of the N MR signal acquired by the small RF coil 1 1 4 and the frequency of the oscillating magnetic field for excitation, and calculates the current at a specific location of the sample 1 1 5 from the difference. Is provided.

[0121] In addition, the measurement apparatus 300 includes an RF oscillator 10 02, a modulator 104, an RF amplifier 10 06, a preamplifier 1 1 2, a detector 301, an A / D converter 1 1 8, and a switch unit 1. 61, Pulse control unit 1 08, Timekeeping unit 1 28, Sequence table 1 27, Calculation unit 1 30, Data reception unit 1 31, Storage unit 305, Output unit 1 35 Etc.

 Further, the measuring device 300 may have a structure described later with reference to FIG.

 [0122] Sample 1 1 5 is a sample to be measured. The sample 1 15 can be in various forms such as a membrane, a solid such as a lump, a liquid, agar, a gel such as a jelly. In the case of a membranous substance, the measurement result of local water content can be obtained stably. In particular, when a membrane having a property of retaining moisture in the membrane, such as a solid electrolyte membrane, is used as a sample, the measurement result can be obtained more stably.

 [0123] The sample mounting table 1 1 6 is a table on which the sample 1 1 5 is mounted, and a sample having a predetermined shape and material can be used.

 Magnet 1 1 3 applies a static magnetic field to sample 1 1 5 (S 3 0 1 in FIG. 1). With this static magnetic field applied, an oscillating magnetic field for excitation is applied to the sample, and the current is measured.

 [0124] The small RF coil 1 1 4 applies an excitation oscillating magnetic field to a specific part of the sample 1 1 5 and acquires an NMR signal corresponding to the excitation oscillating magnetic field (S 3 0 3 in Fig. 1). ) Specifically, the NMR signal is a high-frequency pulse for the excitation oscillating magnetic field to generate nuclear magnetic resonance.

 [0125] There may be one or a plurality of small RF coils 1 14 disposed inside the sample, on the sample surface or in the vicinity of the sample. A configuration in which a plurality of small RF coils 114 are arranged will be described later in the second embodiment.

[0126] The small RF coil 1 1 4 is preferably 1/2 or less, more preferably 10 or less, of the size of the entire sample. By using such a size, it becomes possible to accurately measure the local mobility of the proton solvent in the sample in a short time. The sample size can be, for example, the projected area when the sample is placed, and the exclusive area of the small RF coil 1 14 is preferably 1/2 or less of the projected area, more preferably By setting it to 1/10 or less, accurate measurement can be performed in a short time. The size of the small RF coil 1 14 is preferably, for example, a diameter of 1 O mm or less. As the small RF coil 1 1 4, for example, the one shown in FIG. 33 (a) described later in the embodiment can be used. By using the planar coil as shown in the figure, it is possible to limit the measurement region and perform local measurement. The measurement area of such a spiral coil has a width of about the coil diameter and a depth of the coil radius, for example. In addition, this coil is different from an ordinary solenoid coil, and is flat, so that NMR signals can be acquired simply by pasting on a flat sample.

 [0128] The small RF coil 1 1 4 is not limited to a flat spiral coil, and various types of RF coils can be used. For example, a planar 8-shaped coil (sometimes called a butterfly coil, Doub I e _D type coil, etc.) is also available. The figure 8 coil includes two spiral coils, and the NMR from the sample, even if the coil's spiral axis is parallel to the direction of the main magnetic field of the magnet, or when both are at an angle. The signal can be detected. The spiral coil has sensitivity in the axial direction of the wound coil, whereas the figure 8 coil has sensitivity in the same plane as the wound coil.

 [0129] Returning to FIG. 7, the oscillating magnetic field (exciting oscillating magnetic field) applied by the small RF coil 1 1 4 is the RF oscillator 1 02, the modulator 1 04, the RF amplifier 1 06, the pulse controller 1 08, It is generated by the linkage of the switch unit 1 61 and the small RF coil 1 1 4. In the present embodiment, the RF pulse generator that generates an RF pulse for generating an oscillating magnetic field for excitation in the small RF coil 1 14 includes an RF oscillator 10 2, a modulator 1 04, and an RF amplifier 1 06. Composed. The excitation oscillating magnetic field oscillated from the RF oscillator 102 is modulated by the modulator 104 based on the control by the pulse control unit 108 and becomes a pulse shape. The generated RF pulse is amplified by the RF amplifier 106 and then sent to the small RF coil 1 14.

[0130] The reference frequency is matched to the resonance frequency of the NMR signal when no current is flowing. This resonance frequency is stored in the RF oscillator 102. [0131] In addition, the pulse control unit 10 8 controls the above cooperation so that the excitation oscillating magnetic field applied to the sample 1 1 5 by the small RF coil 1 1 4 is executed according to the pulse sequence described above. To do.

 [0132] The pulse control unit 1 0 8 is connected to the sequence table 1 2 7 and the time measuring unit 1 2 8, and the sequence data acquired from the sequence table 1 2 7 and the measurement time in the time measuring unit 1 2 8 are used. Based on this, a high frequency pulse is generated. The sequenceable 1 2 7 stores the sequence data of the high frequency pulse when measuring the current. Specifically, the sequence table 1 27 stores a timing diagram in which the generation time and interval of the high-frequency pulse are set, and the intensity of the high-frequency pulse to be applied based on the timing diagram.

 The small R F coil 1 1 4 applies this R F pulse to a specific location of the sample 1 1 5 placed on the sample mounting table 1 1 6. The small RF coil 1 1 4 acquires the NM R signal of the applied RF pulse. The NMR signal is, for example, an echo signal corresponding to an oscillating magnetic field for excitation. The frequency of the echo signal changes from the reference frequency described above due to the magnetic field formed by the flow of current. For this reason, the current flowing through the sample 115 is obtained from the difference in the frequency of the measured echo signal by acquiring the relationship between the frequency change (difference) and the current value in advance. The frequency difference can be obtained by converting the amount of phase change at a certain time interval per unit time.

 [0134] The exciting oscillating magnetic field applied to the sample 1 1 5 by the small RF coil 1 1 4 is, for example,

 (a) 90 ° pulse, and

 (b) A pulse sequence consisting of 180 ° pulses applied after the elapse of the pulse time of (a).

[0135] When the small RF coil 1 14 is used, it may be difficult to adjust the excitation pulse intensity of the above (a) and (b). For example, in the region to be measured, that is, the region surrounded by the small RF coil 1 1 4 It is difficult to excite so that the whole is at a uniform excitation angle, that is, the intensity ratio of the excitation magnetic field in (a) and (b) is constant. There is a case. If the excitation angle ratio in (a) and (b) varies, an appropriate spin echo signal cannot be obtained, and accurate measurement of the current becomes difficult.

 [0136] Therefore, in such a case, the step of applying the 1 80 ° pulse at a time just before the 90 ° pulse (a) in addition to the above pulse sequence in the case of the pulse control unit 1 08 force To execute another sequence with And by comparing the behavior of the decay curves of the 1 80 ° pulse (b) corresponding to these two sequences, the excitation pulse intensities of the 90 ° pulse (a) and 1 80 ° pulse (b) are accurate. Or not. As a result, even if the excitation pulse intensity deviates due to an abnormality in the device, an abnormality can be detected before the measurement is performed, and the measured value can be made more accurate. Alternatively, (a) a 90 ° pulse is in the first phase, and (b) 1 80 ° pulse is in a second phase that is 90 ° away from the first phase.

 [0137] Next, detection of the NMR signal will be described.

 The N MR signal detection unit detects the N MR signal acquired by the small RF coil 1 14 and sends this NMR signal to the calculation unit 130. The NMR signal detector includes a preamplifier 1 1 2, a detector 301 and an A / D converter 1 1 8. The detected NMR signal is amplified by the preamplifier 1 1 2 and then sent to the detector 301.

[0138] The detector 301 is configured to detect the real part and the imaginary part of the NMR signal by the phase sensitive detection method. In order to accurately separate the N MR waveform acquired by the detector 301 into the real part and the imaginary part, the phase difference between the sine wave and the cos wave of the fundamental wave that is the source of demodulation is precisely 90 degrees. It is preferable to adjust to. By adjusting the two fundamental waves so that the phase difference is exactly 90 degrees, the phase difference can be calculated more accurately using the real part and imaginary part tan- ¹ described later. Note that the reference wave that is the source of demodulation is Made by the 90 ° hybrid described.

 [0139] The detector 30 1 sends the detected real part and imaginary part to the A / D converter 1 1 8.

 The A / D converter 1 1 8 converts the N MR signal from A / D and sends it to the data reception unit 1 3 1. The calculation unit 130 including the current calculation unit 303 acquires the data sent to the data reception unit 1 3 1.

 [0140] The application of the excitation oscillating magnetic field and the detection of the NMR signal have been described above, but these can be realized by an LC circuit including a small coil.

 FIG. 8 is a diagram showing an example of such an LC circuit. In Fig. 8, the coil part (inductance part) of the resonant circuit is a small RF coil with a diameter of 1.4 mm. In the nuclear magnetic resonance (NMR) method, the number density and spin relaxation time constant can be measured by detecting the movement of nuclear magnetization as an N MR signal by the spin resonance phenomenon of the nucleus placed in a magnetic field. The spin resonance frequency in the magnetic field of 1 Tes Ia is about 43 MHz, and an LC resonance circuit as shown in Fig. 8 is used to selectively detect the frequency band with high sensitivity.

 [0141] Returning to FIG. 7, the switch section 1 6 1 is provided at the branch section connecting the small RF coil 1 1 4, the RF amplifier 1 06 and the preamplifier 1 1 2, and the small RF coil 1 1 4 and RF The first state where the signal generator (RF amplifier 106) is connected, and

It has the function of switching the second state in which the small RF coil 1 1 4 and the NMR signal detector (detector 30 1) are connected. In other words, the switch section 1 6 1 serves as a “transmission / reception switching switch”. The role of this is to transmit the excitation pulse amplified by RF p ower-amp to the small RF coil 1 1 4, disconnecting the receiving preamplifier 1 1 2 and protecting it from a large voltage, and N after excitation When receiving the MR signal, the noise generated by the large amplification transistor leaking from the RF amplifier 106 is blocked so as not to be transmitted to the receiving preamplifier 1 1 2. When measuring using a small RF coil 1 1 4, the switch section 1 6 1 is required for the following reasons in order to handle weak signals. On the other hand, in large measurement systems that do not use small RF coils 1 1 4 ”Can be dealt with sufficiently. The cross diode is a diode that is turned on when a voltage equal to or higher than a predetermined value is applied, and is turned off when the voltage is lower than the predetermined value.

 [0142] The reason why the “transmission / reception switching switch”, that is, the switch section 1 6 1 is particularly necessary when the small RF coil 1 1 4 is used is as follows.

 (i) The sample volume that can be detected by the small coil of this measurement system is smaller than that of the large coil. This detectable sample volume is approximately (coil inner area x coil radius depth). In order to measure weak NMR signals that decrease in proportion to volume with low noise and high sensitivity, it is necessary to block noise leaking from the large amplification transistor of the RF amplifier 10 6 in the transmission system. Necessary. In the receiving system, it is necessary to use a highly sensitive preamplifier 1 1 2. When using the high sensitivity preamplifier 1 1 2, preamplifier 1 1 2 must be disconnected so that preamplifier 1 1 2 can be protected from the high voltage excitation pulses sent to the small coil during transmission.

 (ii) When exciting the nuclear magnetization in the sample volume, with the appropriate excitation pulse power, specifically, the relationship between the intensity of the 90 degree pulse and the 180 degree pulse is 1 to 2, or the irradiation energy is It is necessary to excite nuclear magnetization so that the relationship between 1 to 4 or pulse application time is 1 to 2. If the excitation pulse power cannot be adjusted properly, the target pulse sequence of the spin echo method will not be obtained, and as a result, an appropriate spin echo signal cannot be acquired. Reliability decreases. This phenomenon is prominent when switching between transmission and reception of a small coil using a conventional cross diode. In the case of a large coil, the excitation pulse intensity is very large and the loss at the cross diode can be considered to be negligibly small, but in the case of a small coil, the excitation pulse intensity is smaller than that of the large coil. The loss at the cross diode cannot be ignored. For this reason, in order to obtain an appropriate excitation pulse intensity, a “transmission / reception switching switch” with minimal loss is required.

[0143] By providing a switch section 1 6 1 at the above branch section, a small RF coil 1 1 The loss of the oscillating magnetic field signal applied from 4 to the sample 1 15 is reduced, and as a result, the pulse angles of the 90 ° pulse and 1 80 ° pulse can be accurately controlled. Accurate control of the pulse angle is an important technical problem in order to reliably obtain the compensation effect in the spin echo method. In this embodiment, this problem is solved by the arrangement of the switch unit 161.

 [0144] In addition, the RF detection coil for local measurement is miniaturized, and low noise during NMR reception is an important factor for ensuring the accuracy of measurement. When receiving NMR signals, the noise that enters the preamplifier 1 1 2 mainly includes an RF wave transmission system. RF leakage from the RF amplifier 1 06 that amplifies the excitation pulse and high power There is noise generated by the amplifier. When receiving the N MR signal, it is necessary to cut off the excitation wave leaking from the transmitter side with the switch unit 161 and receive the N MR signal with low noise. In the present embodiment, such a problem is also solved by the arrangement of the switch portion 161.

 [0145] The switch unit 1 61 can employ various configurations. FIG. 9 is a circuit diagram showing an example of the configuration of the switch unit 1601.

 [0146] The apparatus configuration around the sample has been described above. Next, the N MR signal processing block will be explained.

 Returning to FIG. 7, the real part and the imaginary part of the N MR signal (echo signal) detected by the detector 301 are acquired by the data reception unit 1 31 and sent to the calculation unit 1 30. The calculation unit 130 has a current calculation unit 303. The current calculation unit 303 acquires the real part and the imaginary part of the echo signal detected by the detector 301, calculates the phase difference between the echo signal and the excitation oscillating magnetic field using these, and from this phase difference, the echo is calculated. The difference between the frequency of the signal and the frequency of the oscillating magnetic field for excitation (frequency shift amount) Δω is calculated (S 305 in FIG. 1).

[0147] Specifically, tan- ¹ (Re / I mg) is calculated from the detected real part and imaginary part. This value corresponds to the phase difference ΔΦ [rad] of the NMR signal. ΔΦ is the phase difference between the measured NMR signal and the reference wave (phase Φ ₀ ) that travels at a frequency that does not change with time, as shown in Fig. 10. Here, the reference frequency is The resonance frequency of the NMR signal in a state where no flow is flowing is set in advance.

 The current calculation unit 303 obtains Δω from the amount of change per unit time of the obtained phase difference ΔΦ. Then, by referring to the relationship between Δω and the current, the value of the current of the sample 115 at the measurement location is calculated (S 307 in FIG. 1). In addition, the current calculation unit 303 may calculate the current density by dividing the obtained current value by the area where the current flows.

 Here, the measuring apparatus 300 includes a storage unit 305 that holds information indicating the correlation between the current and the frequency difference for each type of the sample 1 15. The storage unit 305 stores, for example, data on correlation between the frequency difference Δω and current obtained experimentally. More specifically, this is calibration curve data of the frequency difference Δω and the current. The current calculation unit 303 in the calculation unit 130 acquires calibration curve data corresponding to the sample to be measured from the storage unit 305, and calculates a current corresponding to the frequency difference Δω based on the calibration curve data.

 [0150] The current calculated by the current calculation unit 303 is presented to the user by the output unit 135 (S309 in FIG. 1). The presentation format can take various forms, and there are no particular restrictions on display, printer output, file output, etc.

 [0151] Fig. 41 shows the RF oscillator 10 02, modulator 10 04, RF amplifier 106, pulse control unit 10 08, switch unit 161, small RF coil 1 1 4, in the measuring apparatus 300 shown in Fig. 7. FIG. 3 is a diagram showing an example of a more detailed configuration regarding the cooperation of a preamplifier 1 1 2, a detector 301 and an A / D converter 1 1 8. This configuration can also be applied to the measuring apparatus shown in FIGS. 13 and 16 described later.

 In FIG. 41, the modulator 104 includes a mixer 1 77, a mixer 1 79, and a synthesizer 1 81. The detector 301 includes a mixer 1 83, a mixer — 1 85, and a distributor 1 87. The A / D converter 1 1 8 includes a first _A / D converter 1 189 and a second A / D converter 191.

In FIG. 41, a 90 ° hybrid 1 7 1 and a distributor 1 73 are further arranged in this order between the RF oscillator 1 02 and the modulator 1 04, and the 9 0 ° hybrid 1 7 1 Distributor 1 75 is placed between the detector and detector 301 Has been.

 [0154] In this configuration, the waveform output from RF oscillator 102 is changed to two waveforms of the same frequency but different in phase by 90 ° by 90 ° hybrid 1 71. Based on these two reference waveforms, the NMR signal is detected and becomes R e a I and I m i g n a r y components.

 [0155] Here, the two waveforms output from the 90 ° hybrid 1 7 1 are specifically a sine wave and a cos wave, and the two waveforms are accurately orthogonal to each other. It is an important point in seeking.

 [0156] In Fig. 41, the signal name of A / D converter 1 1 8 is named Rea I and I ma ginary, but this is a convenient expression, I ma ginary And it may be the opposite of Real. In the opposite case, the phase obtained by a r c t a n is only shifted by ± 90 °, and there is no problem in obtaining the “phase change” that increases or decreases with time.

 [0157] In addition, the "N MR signal strength" required when determining the water content and mobility described later in the fourth and fifth embodiments is the component of the acquired Rea I and I ma ginary. Based on

(R ea I "2+ I ma ginary" 2) ^Λ — 1/2

 Can be converted to that intensity. In other words, this calculation corresponds to finding the radius of the circle in FIG.

[0158] Next, the function and effect of this embodiment will be described.

 When a small surface coil is used as in this embodiment, since the measurement area is small, the static magnetic field uniformity in the measurement area is high, and the echo signal can be observed over a very long time. As a result, the frequency shift amount can be measured with high frequency resolution.

[0159] Further, by converging the phase of the NMR signal using the spin echo method, the phase of the echo signal can be converged and acquired by the small RF coil 114. As a result, the real part and the imaginary part of the echo signal can be detected and the phase difference ΔΦ using them can be calculated more accurately. Note that this embodiment and The spin-echo method need not be used for the measurement of the frequency shift amount in other embodiments of the present specification, and the frequency shift amount can also be calculated from a simple FID (Free Induction Decay). The spin echo method has a smaller measurement area than the FID, so that the uniformity of the static magnetic field can be further improved compared to the FID.

 [0160] Also, when trying to measure the whole sample with a large solenoid coil instead of the small RF coil 1 1 4 when measuring the current, the uniformity of the static magnetic field deteriorates, the FID is short, and the echo is The shape is short and sharp. For this reason, it is difficult to perform with a large solenoid coil due to the uniformity of the magnetic field. On the other hand, in the present embodiment, since the small RF coil 114 is used, sufficient magnetic field uniformity can be obtained.

 [0161] In the chemical shift method, it is also possible to measure the frequency shift amount of the NMR signal using a large solenoid coil and applying a gradient magnetic field to excite only a specific local area. In contrast, in small surface coils, the measurement area is limited by the shape of the coil, so there is no need to use a gradient magnetic field for local excitation. For this reason, the apparatus configuration can be simplified.

 In addition, when using the Hall element described above with reference to Non-Patent Document 2, it is necessary to measure two physical quantities of the resistance value and temperature of the sensor, whereas in the method of this embodiment, Since it is only necessary to obtain the frequency obtained from the magnetic resonance signal, this is a simple method because only one physical quantity needs to be measured.

 [0163] In NMR measurement using a small surface coil, a static magnetic field may be created using a rod-shaped magnet, and the sensor unit is small and can be used as a current measurement probe that can be easily installed in the apparatus. In addition, as will be described later in the fourth and fifth embodiments, “water content” and “mobility of water molecules” in a sample such as a polymer membrane can be measured locally at the same place almost at the same time. it can.

 [0164] This embodiment and the following embodiments can be applied to, for example, local current measurement of a solid electrolyte membrane of a fuel cell.

Note that the current j measured by the method of the above embodiment is generated by the fuel cell. The principle of forming a magnetic field is the same regardless of whether the current flows in a state where the electric field is flowing or when the direct current is applied to perform water electrolysis. Therefore, by measuring the increase / decrease Δω of the resonance frequency, the spatial current during power generation and water electrolysis operation of the fuel cell can be grasped.

 [01 65] In a fuel cell using a solid polymer electrolyte membrane, the power generation state changes depending on the gas supply state, catalyst deterioration, and ionic conductivity of the polymer electrolyte membrane. When the hydrogen utilization rate is increased, the hydrogen concentration is high near the gas supply port and the generated current at that location is large. On the other hand, the hydrogen concentration is low near the gas outlet and the generated current is also small. This is because “material transport loss” increases.

 [01 66] Moreover, if the catalyst of the fuel cell deteriorates, the “activation loss” increases and the generated current decreases. The Pt catalyst deteriorates due to transient fluctuations such as starting and stopping of the fuel cell, resulting in spatial non-uniformity. Also, depending on the water content of the polymer electrolyte membrane, the ionic conductivity increases and decreases, the “ohm loss” changes, and the power generation current increases and decreases. Since this loss depends on the spatial distribution of water content, the current also has a spatial distribution even within a single polymer electrolyte membrane.

 [01 67] In fuel cell power generation, the above-mentioned “mass transport loss”, “activation loss”, and “ohmic loss” overlap to determine the final output current and voltage, which is the battery performance. The current output from the output terminal of the fuel cell can be easily measured with an ammeter. The current value is a flat "MEA (Membrane Electro Assembly)" and has a spatial distribution. It is the value obtained by integrating the current over the entire plane. The information inside the battery that is necessary to improve battery performance is the power generation state that varies from place to place, and the current that varies from place to place. Furthermore, if water content, water molecule mobility, and gas concentration can be measured for each location, the phenomenon occurring inside the battery can be understood in more detail.

[01 68] If the spatial distribution of current in the MEA plane can be measured, the state of power generation at each location can be determined, and what is the state of “material transport loss”, “activation loss”, and “ohm loss” spatially? Specifically, when the gas supply concentration, hydrogen utilization rate, gas supply pressure, humidification amount, and moisture content of the membrane are changed, the power generation state at which location changes. It is possible to examine in detail whether the resulting battery performance has increased or decreased. This can provide technical guidelines for improving battery performance.

. Measuring the spatial distribution of current in the MEA plane is necessary for developing high-performance fuel cells.

According to the present embodiment, the spatial current of the fuel cell can be grasped by measuring the increase / decrease Δω of the resonance frequency.

[0170] Note that in the measuring apparatus 300, in addition to the current measurement, the CPMG method can be used to obtain the ₂ relaxation time constant, whereby the amount of water in the sample can be calculated.

 In addition, the measuring device 300 can be further provided with a gradient magnetic field coil, and by appropriately applying a gradient magnetic field to the sample, the mobility of the target molecule can be measured by using the PGS method with the gradient magnetic field applied. it can.

 These measurements will be described in the fourth and fifth embodiments described later.

[01 71] (Second Embodiment)

 The measuring apparatus 300 described in the first embodiment may include a plurality of small RF coils 1 1 4. In the present embodiment, the plurality of small RF coils 1 1 and 4 apply an excitation oscillating magnetic field to a plurality of locations of the sample 1 1 5 and acquire a nuclear magnetic resonance signal. The current calculation unit 303 is configured to calculate currents at a plurality of locations of the sample 1 15.

 FIG. 11 is a perspective view showing an arrangement example of a plurality of small RF coils 1 14.

[0172] By providing a plurality of small RF coils 1 1 4 in the apparatus, the current distribution in the sample 1 1 5 can be measured. In this case, if the two-dimensional arrangement is performed along the surface of the sample 1 15, the two-dimensional current distribution on the sample surface can be obtained. In addition, if it is arranged three-dimensionally in sample 1 15, the three-dimensional current distribution in the sample can be obtained.

[0173] For example, a current distribution calculation unit (not shown) that calculates the current distribution in the sample 1 15 based on the current calculation result in the calculation unit 1 3 0 force current calculation unit 3 0 3 is provided. It may be. Thereby, it is possible to apply the excitation oscillating magnetic field to a plurality of locations of the sample and acquire the corresponding NMR signals. The current distribution calculation unit (not shown) calculates the current distribution in the sample based on the current at multiple points in the sample. The output unit 1 3 5 outputs this distribution.

 In this embodiment, the output unit 1 3 5 may have the configuration shown in FIG.

 In FIG. 12, the output unit 1 3 5 is a measurement data acquisition unit that acquires the current for each measurement region of the plurality of small RF coils 1 1 4 (FIG. 1 1) calculated by the current calculation unit 3 0 3. 1 3 5 A and a display section 1 3 5 B that displays the acquired current in a partitioned area of the same screen.

 In display unit 1 3 5 B, as shown in FIG. 11, the screen is divided into a plurality of regions according to the arrangement position of small RF coil 1 1 4. Each area is displayed in a predetermined color according to the current in the measurement area of each small RF coil 1 14.

 [0176] In this way, by displaying the color corresponding to the current in the measurement area of each small RF coil 1 1 4 on each area of the display unit 1 3 5 B, measurement with each small RF coil 1 1 4 It is possible to intuitively grasp the relationship between the position and the current. Thereby, it is possible to provide a measuring device that is convenient for the user.

 [0177] Further, the plurality of areas of the display unit 1 3 5 B are divided into upper and lower parts, current is shown on one side (for example, the upper half), and in the sample 1 1 5 in the corresponding place on the lower half The amount of moisture may be indicated. The measurement of the water content in the sample 1 15 will be described later in the fourth embodiment.

 [0178] Further, the measuring apparatus 300 described in the first embodiment may be an apparatus that acquires a current distribution in a plane of a solid polymer electrolyte membrane of a fuel cell by using a nuclear magnetic resonance method. Good. At this time, Sample 1 15 is a solid polymer electrolyte membrane of a fuel cell. Also, the current distribution acquisition unit (current distribution calculation unit) Force For multiple small RF coils 1 1 4, the difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coil 1 1 4 and the frequency of the oscillating magnetic field for excitation ( The frequency shift amount Δω) is calculated, and the current distribution in the plane of the solid polymer electrolyte membrane is obtained from Δω.

[0179] When the measuring device 300 is a measuring device for a fuel cell, By obtaining Δω for multiple regions within the surface of the membrane or MEA (Membrane Electrode Assembly), the operating state of the fuel cell can be diagnosed. For example, when Δ ω is measured in each region and compared with the theoretical analysis value, if the measured value of Δ ω shows a different behavior from the analysis value only at a specific measurement location, There may be a defect in ME IV at the location. In addition, when Δ ω is measured at a predetermined time interval for a plurality of measurement points, and the difference between the measured value of Δω and the analysis value becomes large at all the measurement points, Μ Μ Α There is a possibility that the output is reduced overall.

 Note that the method for obtaining Δω in a plurality of regions in the plane of M EA will be described more specifically in the examples described later.

[0180] (Third embodiment)

 In the measuring apparatus 300 (FIG. 7), by providing a plurality of small R F coils 1 14, the measurement accuracy of the frequency shift amount Δω can be further increased by the following procedure.

 [0181] Here, when the current measurement method described above in the first embodiment is used, for example, as described later in the practical example, first, “nucleus of moisture molecules without current (in which no current flows)” The frequency of magnetic resonance (this is called the reference frequency) is determined, and then the "shift amount of the nuclear magnetic resonance frequency when a current is passed" is determined.

 [01 82] In this method, when the reference frequency does not change with time (constant) (for example, when a superconducting magnet or electromagnet is used as the magnet), the reference frequency is the first to be measured. It is only necessary to acquire it, and after that, it is only necessary to perform measurement when current is passed. In this case, there is an advantage that current fluctuation over time and transient response are easy to measure.

 [0183] On the other hand, when a permanent magnet is used as the magnet, the magnetic field strength may fluctuate with time due to fluctuations in the temperature of the magnet. When the magnetic field strength fluctuates, the reference frequency cannot be said to be constant over time.

[0184] In the case of such a fluctuating magnetic field (reference frequency), “the NMR in the case of no current in a time interval that is short enough to assume that the reference frequency does not change” The method of measuring the signal and determining the frequency shift amount from the difference between the two is considered. For example, in the examples described later, this method is selected, and the time interval is set to 10 seconds.

 [0185] However, for example, when the above embodiment is used for measurement of a fuel cell, the fuel cell does not immediately enter a steady state even if it starts power generation and starts flowing current, and it takes several seconds to several minutes to reach a steady state. May be required. Also, it may be difficult to apply to current measurement in “slowly occurring phenomena” where the load varies, the gas supply fluctuates, or moisture condenses in the gas diffusion layer. The reason is that the above-mentioned “NMR measurement with and without current at a short time interval where the reference frequency can be regarded as constant” cannot be performed.

 Therefore, in this embodiment, the following (i) and (i i) are combined to estimate the reference frequency when there is no current, and the frequency shift amount is calculated accordingly.

 (i) Use multiple RF coils. Place one coil at the position you want to measure, and the other RF coil at a position far from the fuel cell so that the effect of the magnetic field generated by the current is negligible.

 (ii) “The magnetic field distribution in the permanent magnet (nuclear magnetic resonance frequency) only rises and falls uniformly in the entire space. If the fluctuation of the nuclear magnetic resonance frequency is measured at a certain point, The nuclear magnetic resonance frequency of a place can be estimated by raising and lowering the fluctuation amount.

[0187] Further, in the present embodiment, a plurality of small RF coils 1 14 are arranged as follows. Here, the case of measurement of power generation of a fuel cell is shown as an example, but the following method can be applied to the types of Sample 1 15 without any particular limitation.

 (Coil 1) First small R F coil 1 1 4: A little away from the fuel cell

Place the RF coil in a place where the influence of the magnetic field due to the current can be ignored, place the sample 1 1 5 for measuring the reference frequency in a place where the coil can receive the NMR signal, and set the nuclear magnetic resonance frequency in the magnet. measure. This serves as the “RF coil and sample for reference frequency monitoring”. (Coil 2) The remaining small RF coils 1 1 4: Place them on the sample 1 1 5 where you want to measure.

 Measurement is performed using a system that can receive NMR signals using multiple coils.

 (i) First, the NMR signal in the “no current flow” state is received by all small RF coils 1 1 4 (coil 1 and coil 2), and the magnetic field distribution (nuclear magnetic resonance frequency distribution ω) is obtained. deep.

(i-1) _Let the resonance frequency in coil 1 be ω _r (t = 0). t is time

(i -2) The nuclear magnetic resonance function at the position x where the coil 2 is placed is expressed as ω (t = 0, x

).

 (i i) Start power generation in the fuel cell, and receive NMR signals in the “current-carrying state” at all small RF coils 1 1 4 (coil 1 and coil 2). The time is t1.

 (i i i) From the NMR signal acquired in (i), obtain the reference frequency variation Δω.

 The fluctuation amount Δω (t1) at time t 1 is

 ^ monitor (t1) One ^ monitor (= 0)

It is.

 (iv) Using this variation Aco (t1), estimate the “reference frequency when there is no current” at the coil position of coil 2 using the following formula.

ω _η - _current (t1, x) = co (t1, x) + Δ ω (t1)

 (ν) “The frequency of the N MR signal actually measured by coil 2 when there is current.

_Let J _ourren t (t1, X).

 The frequency shift ΔωΟ, X) with current is

Α ω (t1, x) = ω _ourrent (t1, X) _OJ _no — _current (t1, X)

It can be calculated by The current distribution should be analyzed based on Δω (t1, x) obtained as described above. [0189] According to the present embodiment, in addition to the first embodiment, the following operational effects can be obtained. That is, by measuring the NMR signals for coil 1 and coil 2 at the same time, the reference frequency can be more accurately determined. Can be estimated. Thereby, the measurement accuracy of the frequency shift amount Δ ω (t1, x) can be improved. For this reason, for example, when used for measurement of a fuel cell, it is not necessary to measure the fuel cell by switching between “no current in a short time”. This makes it possible to measure in a more realistic power generation situation.

[0190] Coil 1 and coil 2 need not be measured simultaneously. For example, measurement may be performed by alternately switching between coil 1 and coil 2. The reference frequency w _{n for} “no current” depending on the position of the coil. It is sufficient if _{_current} (t1, x) can be estimated with a certain degree of accuracy.

 [0191] As described above, in the first to third embodiments, the measurement of the local current of the sample has been described.

 In the following embodiments, a method and apparatus for measuring the amount or mobility of a protonic solvent in a sample in addition to the local current of the sample will be described. The following embodiment is applicable to any of the first to third embodiments.

 [0192] (Fourth embodiment)

 In the present embodiment, current and water content are measured using a local magnetic resonance signal acquired by a small detection coil.

 FIG. 13 is a diagram showing a schematic configuration of the measurement apparatus of the present embodiment. The basic configuration of the apparatus shown in FIG. 13 is the same as that of the measuring apparatus 300 shown in FIG. 7 except that a solvent information calculation unit 3 09 is provided in the calculation unit 1 3 0. . Further, the difference is that a control unit 3 0 7 is provided instead of the pulse control unit 1 0 8 in FIG.

[0194] The solvent information calculation unit 3009 calculates information related to the solvent contained in the sample 1 15 and includes a water content calculation unit 1 32 in this embodiment. Based on the NMR signal acquired by the small RF coil 1 1 4 Calculate the amount of protonic solvent (water).

 FIG. 14 is a diagram showing the configuration of the control unit 307 of the apparatus shown in FIG.

 In FIG. 14, in addition to the pulse control unit 108 described above, the control unit 307 includes a first measurement mode for measuring the current of the sample 1 15 and a second measurement for measuring the water content in the sample 1 15. Includes a switching unit (mode switching control unit 169) for switching between modes. The operation signal receiving unit 129 connected to the mode switching control unit 169 receives the operator's request for the measurement mode. Then, the operation signal receiving unit 1 29 sends this request to the mode switching control unit 169.

 [0196] In the first measurement mode, the current measurement of the sample 1 15 is performed according to the procedure described in the above embodiment. That is, the current calculation unit 303 forces the calculation of the current at a specific location of the sample 1 15 based on the difference between the frequency of the N MR signal acquired by the small RF coil 1 1 4 and the frequency of the excitation oscillating magnetic field.

[0197] In the second measurement mode, a small RF coil 1 1 4 Forced NMR signal (echo signal) corresponding to the vibration magnetic field for excitation is acquired, and the solvent amount calculation unit 1 32 (moisture amount calculation unit 1 32) However, it calculates the amount of the protonic solvent (water) in the sample 1 1 5 based on the echo signal acquired by the small RF coil 1 1 4. Specifically, the water content calculation unit 1 32 calculates the T ₂ relaxation time constant from the intensity of the echo signal, and from the calculated Τ ₂ relaxation time constant, the protonic solvent at a specific location in the sample 1 15 Calculate the amount of

 In this embodiment, for example, a common pulse sequence is used for the first and second measurement modes.

 That is, the small RF coil 1 14 applies an oscillating magnetic field for excitation in a pulse sequence including the following (a) to (c).

 (a) 90 ° pulse,

 (b) 1 80 ° pulse applied after the time of pulse in (a), and

(c) n 1 80 ° pulses (n is a natural number) that starts after the lapse of time 2 of the pulse in (b) and is applied at intervals of time 2

[0199] And in the first measurement mode, the small RF coil 1 1 4 force above (b ) Or (c) The echo signal corresponding to the pulse is acquired. At this time,

Since the echo signal corresponding to the pulse (b) has the highest intensity, it is preferable to use this echo signal. Also, the current calculation unit 303 obtains the real part and the imaginary part of the force echo signal and calculates the current.

[0200] —On the other hand, in the second measurement mode, small RF coil 1 1 4 force Acquires a plurality of echo signals corresponding to the pulses (b) and (c) above. In addition, the water content calculation unit 1 32 calculates the T ₂ relaxation time constant from the intensity of the plurality of echo signals.

 [0201] FIG. 15 is a flowchart showing an example of a procedure for measuring current and water content. This measurement method includes the following steps.

 Step 301 (Step 10 02): Applying a static magnetic field to sample 1 15; Step 303 (Step 10 04): Pulse sequence including (a) to (c) above via small RF coil 1 1 4 Apply the excitation oscillating magnetic field and obtain the corresponding echo signal.

 Step 305: Using the real part and imaginary part of the echo signal corresponding to the pulse (b) or (c) obtained in step 303, the frequency of the echo signal and the frequency of the oscillating magnetic field for excitation The difference between

 Step 307: From the difference obtained in step 305, obtain the current at a specific part of the sample.

Step 1 06: Calculate the T ₂ relaxation time constant from the intensities of multiple echo signals corresponding to the pulses (b) and (c) obtained in Step 303. Step 1 08: Τ calculated in Step 1 06 ₂ Measure the local moisture content in the sample from the relaxation time constant.

 Step 309 (Step 1 1 0): After that, the result is output.

[0202] In Step 303 (Step 104), an echo signal is obtained for calculating the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation, and Τ ₂ for calculating the relaxation time constant. The acquisition of the echo signal may be performed simultaneously. For example, to calculate the difference calculation and T ₂ relaxation time constant of the frequency, both above (b) When using a pulse, in step 3 0 3 (step 1 0 4), an echo signal corresponding to the above pulses (b) and (c) is acquired. In step 300, the frequency difference is calculated by using the real part and the imaginary part of the echo signal corresponding to the pulse (b). In step 3 07, the T ₂ relaxation time constant is calculated from the intensities of multiple echo signals corresponding to the pulses (b) and (c) above. Note that the pulse in (c) above can be used for both frequency difference calculation and ( ₂₎ relaxation time constant calculation.

 [0203] According to the present embodiment, it is possible to measure not only the local current of the sample 1 15 such as a membrane, but also the water content with one apparatus by a series of measurements using a common pulse sequence. . For this reason, the state of the sample 1 15 during power generation or water electrolysis operation can be grasped in more detail.

[0204] In this embodiment, the small RF coil 1 14 is a pulse that alternately repeats a pulse for current measurement (first measurement mode) and a pulse for moisture measurement (second measurement mode) multiple times. An oscillating magnetic field for excitation can also be applied in a sequence. In other words, it performs an acquisition of echo signals for calculation unloading of the difference between the frequency and the frequency of the excitation oscillating magnetic field of the nuclear magnetic resonance signal, the acquisition of echo signals for calculating the T ₂ relaxation time constant, alternating You can also In this way, the local current and water content of sample 1 15 can be measured more stably.

 [0205] (Fifth embodiment)

 In the present embodiment, current, water content, and water mobility are measured using local magnetic resonance signals acquired by a small detection coil.

FIG. 16 is a diagram showing a schematic configuration of the measurement apparatus of the present embodiment. The basic configuration of the apparatus shown in FIG. 16 is the same as that of the measuring apparatus shown in FIG. 13. However, the solvent information calculation unit 3 0 9 of the calculation unit 1 30 is further replaced with the mobility calculation unit 1 3 3. And the movement amount calculation unit 1 3 4 is provided in the calculation unit 1 3 0. In addition to the apparatus shown in FIGS. 7 and 13, the apparatus shown in FIG. 16 has a gradient magnetic field application unit (a pair of G coils 1 5 1) that applies a gradient magnetic field to sample 1 15. And a pair of G coils 1 5 1 Further prepare.

 [0207] —The pair of G coils 15 1 is a gradient magnetic field application coil disposed away from the small RF coil 1 1 4. As shown in FIG. 17, the pair of G coils 15 1 is arranged so that a gradient magnetic field can be applied to the sample 1 15. Two G coils 15 1 are arranged with respect to one small RF coil 1 1 4, and are arranged opposite to each other with the small RF coil 1 1 4 interposed therebetween.

 [0208] Various shapes can be adopted for the G coil 151, but a flat coil is used in the present embodiment. In this embodiment, the G coil 15 1 has a half-moon shape as shown in FIG. FIG. 17 illustrates a case where a plurality of small RF coils 1 1 4 are provided in one sample 1 1 5, and a pair of G coils 1 5 1 is arranged for each small RF coil 1 1 4. Yes. The G coil 15 1 is arranged in parallel to the surface of the sample 1 15.

 [0209] Further, the G coil 15 1 is arranged above the small RF coil 1 14. As a result, a gradient magnetic field having a magnetic field gradient in the y-axis direction can be formed on the central axis of the small RF coil 1 14.

 [0210] Small R F coil 1 1 4 Between one G coil 1 5 1 and Small R F coil 1

 A shield shield (not shown) is provided between 14 and the other G coil 15 1. This shielding shield prevents noise from the G coil 15 1 from affecting the small RF coil 1 14. The shielding shield has a thickness that prevents the passage of noise and allows a magnetic field to pass.

 [021 1] When measuring the current, water content, and self-diffusion coefficient, only the small RF coil 1 1 4 is brought into contact with the sample 1 1 5.

 [0212] Returning to Figure 16, the mobility calculation unit 1 3 3 is based on the NMR signals acquired by the small RF coils 1 1 4 obtained corresponding to different gradient magnetic fields. Calculate the mobility of the protonic solvent (water).

[0213] The movement amount calculation unit 1 3 4 is based on the water amount calculated by the moisture amount calculation unit 1 3 2 and the self-diffusion coefficient calculated by the mobility calculation unit 1 3 3. Calculate the travel distance. For example, the movement amount calculation unit 1 3 4 calculates the movement amount of water molecules. A parameter storage unit in which parameters for storing are stored, and a movement amount calculation unit that reads a calculation formula stored in the parameter storage unit and calculates a movement amount of water molecules.

 [0214] The parameter storage unit stores a calculation formula for calculating the amount of movement of water molecules from the self-diffusion coefficient and the amount of water for each sample 1 15 type. Based on this calculation formula, the movement amount calculation unit can calculate the movement amount.

 [0215] In the present embodiment, the mode switching control unit 169 in the control unit 307 is a first measurement mode in which the current of the sample 1 15 is measured, and the water content in the sample 1 15 is measured. Switch between the second measurement mode and the third measurement mode that measures the mobility of water in sample 1 1 5.

 [0216] In the third measurement mode, an excitation oscillating magnetic field is applied to the small RF coil 1 1 4 force sample 1 15 and a nuclear magnetic resonance signal corresponding to the excitation oscillating magnetic field and gradient magnetic field is acquired. In addition, the mobility calculation unit 133 calculates the mobility of a specific portion of the sample 1 15 based on the information of the nuclear magnetic resonance signals obtained corresponding to different gradient magnetic fields.

 [0217] In the third measurement mode, the small RF coil 11 14 applies the excitation oscillating magnetic field in a pulse sequence including the following (a) to (d).

 (a) 90 ° pulse of oscillating magnetic field for excitation,

 (b) Gradient magnetic field pulse that starts after the elapse of the pulse time of (a) and is applied for a certain time d,

 (c) 1 80 ° pulse of an oscillating magnetic field for excitation applied after the elapse of the time of the pulse of (a), and

 (d) Gradient magnetic field pulse that starts after elapse of the pulse time of (c) and is applied for a certain time d.

[0218] Note that the gradient magnetic field applied in (b) may be zero. In addition, (a) 90 ° pulse is in the first phase and (c) 1 80 ° pulse is in the second phase that is 90 ° shifted from the first phase. Obtain correlation between peak intensity and self-diffusion coefficient D of water in sample 1 1 5 You can also.

 [0219] FIG. 18 is a flowchart showing an example of a procedure for measuring current and mobility. FIG. 18 is an example of using the gradient magnetic field applied in (b) of the above pulse sequence as an outlet, and includes the following steps.

 Step 3 0 1 (Step 1 0 2): Applying a static magnetic field to the sample 1 1 5, Step 3 0 3 (Step 2 0 2): Pulse with zero gradient field and the above (a) to (d) In sequence, apply the excitation oscillating magnetic field and acquire the corresponding echo signal (first step)

 Step 30: Difference between the frequency of the echo signal and the frequency of the oscillating magnetic field for excitation using the real and imaginary parts of the echo signal corresponding to the pulse (d) obtained in step 303 To calculate,

 Step 3 07: From the difference obtained in Step 3 0 5, find the current at a specific part of the sample (second step)

 Step 2 04: The gradient magnetic field is set to a non-zero predetermined magnitude, and the excitation oscillating magnetic field is applied in the pulse sequence including the above (a) to (d), and the corresponding echo signal is acquired (third step) ),

 Step 2 06: From the peak intensities of the NMR signals obtained in Step 2 0 2 and Step 2 0 4, the self-diffusion coefficient D of water at a specific location in Sample 1 1 5 Seeking (fourth step)

 Step 3 0 9 (Step 2 0 8): After that, the result is output.

[0220] In this procedure, the gradient magnetic field in step 2 0 2 is set to zero.

The current and self-diffusion coefficient D can be obtained by a series of measurements.

[0221] In the apparatus shown in Fig. 16 as well, the current and moisture content of sample 115 can be obtained by a single measurement by the procedure described above with reference to Fig. 15 for example.

[0222] In the present embodiment, in addition to the local current of the sample 1 15 such as a membrane, the water content and the mobility of water molecules can also be measured. For this reason, the state in the membrane at the time of power generation or water electrolysis operation can be grasped in more detail. [0223] Also, by combining the above three measurement methods, for example, the "current distribution" in the fuel cell, the "water content" in the polymer electrolyte membrane, and the "mobility of water molecules" can be Or it can measure alternately. Measurement that integrates these is useful as an integrated monitoring method for fuel cells, and is effective as a device that provides new measurement amounts to development sites aimed at improving fuel cell performance and expands the application range of NMR sensors. It is.

 [0224] According to the above embodiment, for example, the following operational effects can be obtained.

 The amount of current flowing in a sample such as a fuel cell can be converted from the frequency shift amount of the NMR signal received by the small detection coil. At that time, even with one small detection coil, if the current flowing through the sample is uniform, the amount of current can be easily converted from the amount of frequency shift.

 In addition, multiple small detection coils can be incorporated into a single unit cell, and NMR signals can be acquired from the polymer film, and the current distribution can be converted from the frequency shift amount ∆ω of the NMR signal that occurs when current flows. It becomes. This makes it possible to effectively measure the current distribution with multiple coils.

 In addition, the “local water content”, “local mobility of water molecules” and “current distribution” of the polymer membrane for fuel cells are measured locally using the same device and the same sensor. By alternately measuring the three quantities every few seconds, both values can be obtained at almost the same time (same device, same position, almost same time).

 In addition, it enables short-time measurement that measures the amount of water, mobility of water molecules, and current within a few seconds.

 In addition, non-invasive measurement using electromagnetic waves is performed simply by attaching to the surface of the polymer film.

 In addition, the amount of water, the mobility of water molecules, and the current distribution can be measured while generating power from the fuel cell.

In addition, the state of the polymer membrane can be grasped from various information such as “water content”, “mobility of water molecules”, and “current distribution”, and the power generation state or water electrolysis state in the fuel cell Monitoring can be performed to control power generation efficiency to the highest level.

 In addition, the “Magnet ■ Magnetic field gradient coil integrated rod-shaped local measurement sensor” allows the sensor to be easily installed in the fuel cell, and only the RF detection coil section needs to be in the measurement area. Measurements can be made without obstruction (Fig. 19).

 Also, by integrating the magnet and coil, the limitations on the applicability of the NMR method depending on the equipment configuration can be solved, and the scope of application can be expanded to food management and process management other than measurement of polymer films.

As described above, the embodiments of the present invention have been described with reference to the drawings. However, these are exemplifications of the present invention, and various configurations other than the above can be adopted.

For example, in the above embodiment, the case where the frequency shift amount is measured using the spin echo method when measuring the current has been described as an example. However, as described later in the embodiment, the frequency shift amount is It can also be calculated from a simple FID (Free Induct on Decay). When using F ID, for example, a small R F coil 1 1 4

 (a) 90 ° pulse

 May be applied to obtain a corresponding F I D signal, and the current calculation unit 303 may obtain a real part and an imaginary part of the F I D signal to obtain a current from a phase change amount. In this way, measurement can be performed with only 90 ° pulses, so the repetition time of the excitation pulse can be shortened, and the amount of phase change can be determined in a short time. In addition, since the F ID signal itself can be observed whether it is smaller or larger than 90 °, adjustment can be simplified compared to the spin echo method.

 In addition, when measuring the amount of water or the mobility of water together with the current, the current can be calculated from the FID signal corresponding to (a) by using the pulse sequence including (a) above. .

 Example

[0227] (Example 1) In this example, water was poured into a narrow gap, a copper plate was placed in close contact with the sample, and the change in frequency of the NMR signal when current was passed through the copper plate was measured.

 FIG. 20 is a diagram for explaining the outline of this embodiment. As shown in Fig. 20, in this example, an electric current j was passed through a copper plate to form a magnetic field! "!" And an experiment was conducted to measure the frequency change of the N MR signal from the water placed next to it. It was. By this experiment, the relationship between the current j and the frequency shift amount Δω is obtained. The copper plate simulates a fuel cell, and the water simulates a polymer membrane in the fuel cell.

[0228] The sample is a sealed container with two cover glass with dimensions of 18 mm x 18 mm and a thickness of 0.1 2 mm, with a gap of 0.5 mm, and 2.5 mm mo I in it. / encapsulating C u S0 ₄ aqueous solution L. Figure 21 shows the sample used in this example.

It is a figure which shows (CuS0 ₄ a q. 2.5 mm o I / L). A copper plate with dimensions of 2 OmmX 2 Omm and a thickness of 0.05 mm was placed directly under the sample. A steady current j can flow through this copper plate using a DC stabilized power supply. In a copper plate, the current j can be regarded as flowing uniformly in the plane.

 [0229] For the N MR signal, the interval between the 90-degree excitation pulse and the 180-degree excitation pulse was set to 5 ms, and the echo signal was measured with an echo time of 1 Oms. In this sequence, a gradient magnetic field was applied for 1 ms before and after the 180 ° excitation pulse so that the NMR signals immediately after the 90 ° and 180 ° excitation pulses did not interfere with the echo signal.

 [0230] Figure 22 shows the echo signal when the current j is zero. The NMR signal was detected by the phase-sensitive detection method, and two signals were acquired, the real part and the imaginary part. In Fig. 22, the real part and the imaginary part are indicated by "R e a I" and "I mag", respectively. The 90-degree excitation pulse was applied at t im e = 5 ms. “Pow er” in the figure is the signal strength calculated from the real and imaginary parts. From the form of “P ow e r”, it can be seen that the echo peaks at t i m e = 15 ms.

[0231] The phase difference ΔΦ [rad] of the NMR signal was calculated using ta n- ¹ (R e / I mg) based on the real part R e and the imaginary part I mg. The phase reference is the reference wave from the oscillator of the NMR device, and this frequency is set in advance to the resonance frequency of the NMR signal. It was. The phase difference between the reference wave (phase Φ ₀ ) that does not change in time and the measured NMR signal is ΔΦ. The relationship between the real part, the imaginary part, and the phase difference ΔΦ is shown in FIG.

In addition, FIG. 23 shows the phase difference ΔΦ calculated from the tan- ¹ (Re / I mg) in FIG. However, in this figure, 1 of the time when the echo signal is observed

Only between 3 3 and 17 ms is shown.

[0233] From Fig. 23, when the phase difference ΔΦ is almost constant in time (straight line) and the current] is zero, the NMR signal rotates with a constant phase difference from the reference wave. I understand.

 Next, the echo signal measured when the current j is 0.8 OA is shown in FIG. 24, and the phase difference ΔΦ calculated based on this is shown in FIG. The pulse sequence is the same as when the current j is closed.

 In the echo signal of FIG. 24, unlike FIG. 22, the real part and imaginary part of the NMR signal oscillate, and the frequency is shifted from the reference wave. In the echo signal area, the real part comes first and then the imaginary part vibrates.

FIG. 25 shows the phase difference ΔΦ calculated by obtaining tan- ¹ (Re / I mg) in FIG. In this figure, as time elapses, the phase difference ΔΦ increases (upward straight line), and you can see how the phase of the N MR signal advances from the reference wave over time. Originally, the phase difference ΔΦ seems to be a single straight line that increases with time, but the phase is expressed in the range of 2 π up to _π force, and + π. When it exceeds, it appears as a discontinuous line shifted by 2π. This is the reason for discontinuous transition from + π to 1π at time = 14.6ms.

[0237] Furthermore, Fig. 26 shows the echo signal measured at 10.8 OA when current j is applied in the reverse direction, and Fig. 27 shows the phase difference ΔΦ calculated based on this.

[0238] In the echo signal of Fig. 26, the N M R signal of the real part and the imaginary part is inverted from Fig. 24.

 (Imaginary part goes down first) You can see that the frequency is shifted from the reference wave in the opposite direction.

[0239] Figure 27 is calculated by obtaining ta n- ¹ (Re / I mg) in Figure 26. The phase difference ΔΦ is shown. In this figure, in contrast to Fig. 25, we can see that the phase difference ΔΦ decreases with the passage of time (straight-down straight line), and the phase of the NMR signal delays with the passage of time from the reference wave. it can.

 [0240] Fig. 28 shows the results of experiments conducted in increments of 0.20A with the current j flowing through the copper plate varied from 0.80A to + 0.80A. On the vertical axis of this figure, the phase difference ΔΦ of the NMR signal that changes in 1 ms is defined as “frequency shift amount of N MR signal Δω [rad / ms]”. This “frequency shift amount Δω [rad / ms]” corresponds to “slope of phase difference ΔΦ” in FIGS. 23, 25, and 27, and the graph of phase difference ΔΦ is linearly approximated by the method of least squares. It was calculated from the gradient.

 [0241] From this graph, it can be seen that the current j flowing in the copper plate and the frequency shift amount Δω [rad / ms] are in direct proportion. By using this result, the current j flowing through the copper plate can be calculated backward by measuring the shift amount Δω [rad / ms].

 [0242] N The factor that increases or decreases the MR signal frequency over time is not only the current j flowing through the copper plate, but also the increase or decrease in magnetic field strength in the case of permanent magnets. If the temperature of the permanent magnet increases or decreases, the magnetic field strength also increases or decreases in inverse proportion to it. For this reason, a time-stable magnetic field is required for frequency shift measurement.

 [0243] However, the permanent magnet has a large heat capacity, and the frequency change due to a sudden temperature change

 If it takes about 1 minute, it can be ignored. For this reason, to accurately measure the frequency shift amount of the NMR signal, the NMR signal when the current is zero is acquired, and the phase difference ΔΦ (j = 0) from the reference wave at that time is obtained in advance. After that (after 10 seconds in this experiment), the current j is applied, the phase difference ΔΦ (j) when the current flows is obtained, and the phase difference substantially caused by the current is expressed as ΔΦ (j ) -ΔΦ (j = 0) In the measurement of a fuel cell, it is only necessary to measure the phase difference when the current is changed by changing the load. By this method, the phase difference can be measured with higher accuracy.

[0244] In addition, this method can also be used by subtracting the frequency of the reference wave even if it is set to a state slightly deviated from the true resonance frequency of the N MR signal. There is a feature that can offset the “deviation”.

 [Example 2]

 In this example, the relationship between the current j and the frequency shift amount ΔΦ (j) during water electrolysis operation using ME A (Membrane Electrode Assembly) was verified.

 1 \ 1 巳 8 is a £ 1 \ 1 (Polymer Electrolyte Membrane) with electrodes joined. The ME A used here was manufactured by attaching Pt and Ir on the anode side and Pt on the cathode side electrolessly on a polymer electrolyte membrane manufactured by Asahi Glass. The dimensions of M E A are 17 mm x 15 mm square, 500; U m thick.

[0246] ME A was standardized, pulled up from the ion-exchanged water just before the experiment, and wiped off the water moderately. From the T ₂ (CPMG) relaxation time constant of ME A just before water electrolysis operation, the water content of MEA was about 10 [H ₂ 0 / S 0 ₃ -H +].

 [0247] ME A was sandwiched between a poly force single-point cell with a small surface coil and a Pt electrode, and power was applied. The voltage applied between the two electrodes was 2 to 3.5 V, and the current j was 0.10 to 0.3 OA. With this energization, ME A decomposes the water contained in the PEM and releases hydrogen and oxygen. In this experiment, water was not supplied during water electrolysis (no humidification condition). The cell temperature was 24 ° C.

 [0248] Fig. 29 shows the echo signal measured when 0.30 A of current j is passed through the MEA, and Fig. 30 shows the phase difference ΔΦ calculated based on this echo signal. From FIG. 29 and FIG. 30, it can be seen that the echo signal acquired from ME A progresses although the phase difference is slight with time.

 [0249] The current flowing in 1 \! £ 8] was set to 0.08 to +0.3 OA, and the experiment was performed in increments of 0.10 A. The current j and the frequency shift Δω [rad / ms Figure 31 shows the relationship. The method of organizing the results is the same as in FIG.

 [0250] From FIG. 31, it can be seen that the current j flowing through ME A and the frequency shift amount Δω [r a d / m s] are almost in direct proportion. Using this result, the current j flowing through M EA can be calculated backwards by measuring the shift amount ∆ω [rad / ms].

[0251] Note that comparing Fig. 28 and Fig. 31, the same current j is applied to the copper plate. It can be seen that the frequency shift amount is as small as about a quarter in the case of MEA and MEA. The reason for this is that when the copper plate is energized, the current flows uniformly, but in ME A, the Pt catalyst electrode plated on the surface and the (P t + I r) catalyst electrode Since the conductivity is non-uniform, it is assumed that the current does not flow uniformly. For this reason, the small surface coil measures the part where the current of the MEA is small, and it is assumed that the amount of frequency shift has decreased.

 [0252] As a method of further improving the measurement accuracy of this measurement method, there is a method of creating a calibration curve by measuring current and frequency shift amount using MEA with uniform electrical conductivity. In addition, multiple small surface coils are installed at multiple locations on the ME A, and the in-plane distribution of the frequency shift amount is measured. Based on this, the current distribution is solved as an inverse problem to obtain a more accurate current distribution. May be.

 [Example 3]

 In this example, when two small coils are placed in adjacent areas and measured, the frequency shift amount corresponding to each area is measured, and the current value corresponding to each area can be measured. I confirmed that there was.

[0254] Specifically, two small RF coils were used to measure the NMR of a water sample placed on two copper plates. At this time, the current values I ₂ flowing through the two copper plates were changed individually, and the relationship between the frequency shift Δω of the N MR signal acquired simultaneously by the two coils and the current value II ₂ was experimentally determined. Asked.

[0255] Figure 32 shows an overview of the entire system. The water to be measured is sandwiched between two glass covers (15 mm x 15 mm) and sealed to a thickness of 0.5 mm. Below that, two copper plates are placed in close contact, and their dimensions are 19 mm x 9 mm. Each copper plate is connected to a constant current power source, and can individually control the current amount, 1 ₂ (j _{2 in the} figure).

[0256] Two small surface coils with a diameter of 1.3 mm and three turns were placed on the sample, and the center distance between the two coils was 6 mm. Both coils are in close contact with the sample. These were inserted into a 1 Tes Ia permanent magnet, and two coils were irradiated with excitation pulses simultaneously, and NMR signals were acquired simultaneously. [0257] The coils and devices used in this example are shown in FIGS. 33 (a) to 33 (c). FIG. 33 (a) is a diagram showing a small surface coil used in this example. FIG. 33 (b) is a diagram showing a pair of polycarbonate holders used in this example. FIG. 33 (c) is a diagram showing an RF coil holder used in this example.

FIG. 34 is a diagram showing the static magnetic field H _Q applied to the sample by the permanent magnet and the direction of the magnetic field created by passing a current through the copper plate. If current I ₂ flows in the positive direction (in the direction of the arrow in the figure) through the copper plate, static magnetic field H. Is applied to the sample, and the frequency shift Δω increases.

[0259] At this time, by passing the current I, a weak magnetic field is also formed around the small coil 2 at a position farther from the small coil 1, and the frequency of the small coil 2 is slightly shifted. Similarly, if a current 1 ₂ is applied, a weak magnetic field is generated around the small coil 1 and the frequency shifts. In this way, the magnetic field formed by the current can shift the frequency not only around the copper plate through which the current flows, but also at a small surface coil at a remote location. Therefore, when the coil is placed with the dimensions shown in Fig. 32, the relationship between the current, I ₂ and the frequency shift amount is grasped, and the frequency shift amount is measured, so that the current, 1 ₂ It is necessary to verify whether can be calculated backward.

Therefore, N MR signal was obtained by changing current and I ₂ independently, and the dependence of frequency shift amount Δω on currents I and I ₂ was measured.

[0260] The NMR signal measurement pulse sequence and the frequency shift amount calculation method used in this example were in accordance with the method of Example 1.

[0261] (Current I ₂ = 0. Current at OA and frequency shift in two coils)

Figure 35 shows the frequency shift of the two coils measured with the current I ₂ = 0. OA and changing the current in increments of 0.2 A. From this result, it can be seen that the frequency shift amount of coil 1 is directly proportional to the current I, and increases with a constant positive slope. On the other hand, the frequency shift amount of the coil 2 is inversely proportional to the current I, and a constant negative slope. It can be seen that the gradient magnitude is smaller than that of coil 1. This result shows that a frequency shift occurs even with respect to the current I, which is far from the coil 2, but the frequency shift amount of the coil 2 is "insensitive".

[0262] (Current I _{2 at} current I ₂ = 0.4 A and frequency shift amount in two coils) Next, when current 1 _{2 is set} to 0.4 A, We measured the frequency shift. Figure 36 shows the frequency shift of the two coils measured by changing the current I in steps of 0.2 A. From this result, the frequency shift amount of the coil 1 is almost the same as that in Fig. 35, and is directly proportional to the current I, and the gradient is the same as in Fig. 35, but in the graph, it is shifted downward by 0.7. It is a straight line. In other words, the absolute value of the frequency shift was reduced by about 0.7 rad / ms. On the other hand, the frequency shift amount of coil 2 is about 3.92 rad / ms larger than that in Fig. 35, but it is inversely proportional to current I, its slope is negative, and its slope is You can see that they are almost the same. The frequency shift amount of coil 2 is almost equal to the frequency shift amount (4.02 rad / ms) of coil 1 at current 1, = 0.4 A in Fig. 35.

[0263] From this result, it can be said that the frequency shift of coil 1 was reduced by 0.7 rad / ms, and the frequency shift amount of coil 2 was increased by about 3.92 rad / ms by applying the current I _2. . There is no other difference.

[0264] (Current I _{2 at} current 1 ₂ = _0.4 A and frequency shift amount in two coils) Next, when current 1 ₂ is a negative value and _0.4 A, coil 1 and The frequency shift at 2 was measured. As in Fig. 36, Fig. 37 shows the frequency shift of the two coils measured by changing the current in increments of 0.2 A.

[0265] This result is the only difference between the vertical shifts seen in Figure 36. When the current I ₂ of _0.4 A was passed, the frequency shift of coil 1 was increased by 0.7 rad / ms, and the frequency shift amount of coil 2 was decreased by about 3.8 rad / ms. I understand. There is no other difference.

[0266] (Current = 0. Current I _{2 at} the time of OA and frequency shift amount in two coils) Now, with current I, the zero was measured by changing the current I ₂ at 0. 4 A increments. Figure 38 shows the frequency shift amount between the two coils.

 The shape of the copper plate is symmetric between 1 and 2, and the positions of coils 1 and 2 are the same distance from the central axis of the poly force cell. Therefore, the measurement results (Fig. 3 8) are shown in Fig. 3 Matches the result of swapping 1 and 2.

[0267] (Current I ₂ when current = 0.4 A and frequency shift amount between two coils) Next, measure current = 0.4 A and change current I ₂ in increments of 0.4 A. did. Figure 39 shows the frequency shift between the two coils. This measurement result (Fig. 39) is also consistent with the result of replacing coils 1 and 2 in Fig. 36.

[0268] (Current ^ _{2 at} the time of current ^ = -0.4 A and frequency shift amount in two coils) Next, the current I is set to negative 0.4 A, and the current l _{2 is set} to 0. Changed by 4 A increments and measured. Figure 40 shows the frequency shift between the two coils. This measurement result (Fig. 40) also agrees with the result of replacing coils 1 and 2 in Fig. 37.

[0269] (Conversion method of current amount from frequency shift amount)

 From the above experimental results, when multiple coils are placed on one sample, the frequency shift amount corresponding to each region is measured, and the current distribution can be calculated backward from the frequency shift amount in each coil. I know that there is.

 It should be noted that there are current distributions in the measurement target, and there are the following two methods for converting the current amount from the frequency shift amount of the NMR signal using multiple small surface coils.

 (i) Calculation method based on the assumption that the frequency shift amount is proportional to the current amount near the coil

 (i i) A method of solving the current distribution as an "inverse problem" by assuming the current distribution and using the relationship between the current and the magnetic field (Bio "Savar's Law) so that the frequency shifts at multiple locations are all consistent.

[0270] The above (i) is a method using the fact that the frequency shift amount strongly depends on the current amount near the coil from the above measurement result.

Also, when solving the inverse problem in (ii) above, the total current value, for example, For example, I and + I ₂ are required, but this can be easily measured. According to this method, a more accurate calculation of the current distribution can be expected.

 [Example 4]

 In Example 3 described above, the current value was measured when two small coils were placed using the spin echo method. Specifically, the current was obtained from the amount of phase change of the echo signal (equivalent to the frequency shift Δω) at a certain time interval.

[0272] The NMR signal is not limited to the echo signal, and the current can be measured by the same method even from the F ID signal.

[0273] Thus, in this example, it was confirmed that the current value when two small coils were arranged could be measured by the F I D signal.

[0274] Also in this example, as in Example 3, the apparatus described above with reference to Figs. 32 and 34 was used.

In this embodiment, the current amount I ι is supplied only to the left copper plate in FIGS. 32 and 34 and no current is supplied to the other right copper plate (current amount I ₂ = 0). Measurements were taken. Then, the current 1 ₂ to zero, to get the N MR signal (FID) by changing only the current value I, the frequency shift amount Δω is measured whether with any dependency by the current.

[0276] In the N MR signal measurement pulse sequence used in this example, a rectangular wave-shaped excitation pulse having a width of 40 3 was irradiated once every 10 seconds, and the intensity of the excitation pulse was determined by the magnetization vector. Is adjusted to excite only 90 degrees.

 The analysis results of the F ID signal obtained with the small surface coil 1 on the left side of the figure are shown below.

[0277] (Current = 0. FID waveform and phase change when OA (frequency shift)) Current ,, = 0. Fig. 42 shows the FID waveform obtained when OA. The FID at this time shows a waveform that decays with the T ₂ * relaxation time constant. It can be seen that FID can be observed significantly up to time = 2 Oms on the horizontal axis.

[0278] Based on Fig. 42, Fig. 43 shows the phase of the FID obtained by calculating arctan (Real / Imaginary). time = 1 FID up to 5ms It can be seen that the phase can be calculated without much dispersion.

 In addition, FIG. 44 shows the time course of the F ID phase (average value of 5 points) from 5.4 ms to 8.3 ms on the horizontal axis based on FIG. From Fig. 44, it can be seen that at current = 0. O A, the phase is almost constant over time.

 [0280] (F I D waveform and phase change when current = 0.4 O A (frequency shift)

)

Figure 45 shows the FID waveform obtained when the current = 0.4 OA. In the FID observed in this case, the non-uniformity of the static magnetic field becomes stronger due to the current flow, and the signal is almost invisible when the horizontal axis is around time = 12 ms. Thus, compared to Fig. 42, the FID in Fig. 45 has a waveform that decays with a shorter T ₂ * relaxation time constant.

 FIG. 46 shows the phase of F I D obtained by calculating a r c t a n (R e a l / I m a g i n a r y) based on FIG. Looking at Fig. 46, the phase can be calculated significantly up to around t i m e = 10 ms, but the phase after that has a large variance, and it cannot be said that a significant phase can be calculated. It can be seen that the non-uniformity of the static magnetic field is increased by passing an electric current, and the time during which the FID signal can be observed strongly is shortened.

[0282] The difference in calculating the amount of phase change observed with FID and echo signals is as follows. That is, in the FID, the T ₂ * relaxation time constant increases or decreases, and the observation time of the N MR signal whose phase can be calculated is increased or decreased by the amount of current. On the other hand, in the case of an echo signal, it is almost constant. It is. From the viewpoint of analyzing the obtained N MR signal, it is preferable that “the time during which the N MR signal can be analyzed so that the phase can be analyzed is constant” like the echo signal.

 [0283] Based on Fig. 46, Fig. 47 shows the amount of change in the phase of the F I D waveform when current I, = 0.4 O A. Compared to the phase change of the current I, = 0.OA shown in FIG. 44, it can be seen that the gradient of the phase change amount is larger in FIG.

[0284] Based on Fig. 43 and Fig. 46, FID phase time according to the method of Example 1 From the elapsed time, we calculated the “amount of phase change over a period of time (frequency shift Δω)”. “Phase change at a certain time interval” calculated from the FID waveform for current 1 = 0.4 OA, that is, Δω was 4. 17 rad / ms.

[0285] (Current = 0.8 8 F I D waveform and phase change when OA (frequency shift)

)

Figure 48 shows the FID waveform obtained when the current was increased to 0.8 OA. In the FID observed in this case, the non-uniformity of the static magnetic field becomes even stronger due to the flow of a larger current, and almost no signal is visible when the horizontal axis is around time = 9 ms. Thus, the FID in Fig. 48 has a waveform that decays with a shorter T ₂ * relaxation time constant compared to Figs. 42 and 45.

 Similarly to the previous method, FIG. 49 shows the phase of F ID obtained by calculating arctan (Real / Imaginary) based on FIG. In Fig. 49, it can be seen that the phase can only be calculated significantly up to t i me = 8.5 ms. The phase after that has a large variance, and a significant phase cannot be calculated.

 Based on FIG. 49, FIG. 50 shows the amount of change in the phase of the F I D waveform when the current is I, = 0.8 OA. Compared to the phase change in FIGS. 44 and 47, it can be seen that the slope of the phase change is larger in FIG.

 [0288] Based on the time lapse of the phase of F ID in Fig. 50, the current I, = 0.80A

 The “phase change at a certain time interval” calculated from the F I D waveform, that is, Δω was 8.01 r a d / ms.

 [0289] (Current ^ = —0. 4 F I D waveform and amount of phase change at OA (frequency shift))

 Next, Fig. 51 shows the F I D waveform obtained when the direction of current flow is reversed and when current = _0.4 OA. The observed FID shows that the current flows in the opposite direction and the direction of the magnetic field is reversed, so that the progression of the Real and I magnary waveforms is reversed compared to Fig. 45. I understand.

[0290] Based on Fig. 51, Fig. 52 shows the phase of the FID obtained by calculating arctan (Real / Imaginary). In this figure, the phase It can be seen that the retreat (progress with a negative slope).

 [0291] Based on Fig. 52, Fig. 53 shows the amount of change in the phase of the F I D waveform when the current is I, = –0.4 O A. Compared to the fact that the slope of the phase change amount of current = 0.40 A in Fig. 47 is positive, the slope of the phase change amount is negative in Fig. 53. .

 [0292] Based on the time course of the FID phase in Fig. 5 3, the “phase change amount at a certain time interval” calculated from the FID waveform at the current I, = —0.40 A, that is, Δ ω was 4.19 rads.

[0293] (Relationship between current I and “phase change amount (frequency shift amount) at a certain time interval”) When current I ₂ = 0.2A, current I is changed in increments of 0.2 A Figure 54 shows the amount of frequency shift in the small surface coil 1 measured in this way. From this result, it can be seen that the frequency shift amount of coil 1 is directly proportional to the current I.

 [0294] From the results of this example and example 3, the “phase change amount at a certain time interval (frequency shift amount)” due to the current is almost the same change amount for the FID and echo signal with respect to the current value. It turns out that it becomes.

 Specifically, if the frequency shift amount in coil 1 in Fig. 35 is compared with that in Fig. 54, the straight lines of both are almost the same slope, and the value of the intersection with the vertical axis is also the same. I understand. As a result, the frequency shift amount with respect to the current is almost the same regardless of whether it is an FID or an echo signal, and the relational expression (calibration equation) between the current and the frequency shift amount is created by either method. If this is done, it can be used to calculate the current value from the frequency shift amount regardless of the method.

 [0295] (Example 5)

 In this example, the magnetic field formed when a current was passed through a single copper plate and the frequency shift amount of the NMR signal generated thereby were analyzed. Then, it was confirmed that the frequency shift obtained by the analysis and the measured frequency shift were in good agreement.

 [0296] (Example 5 _ 1)

In the following way, the magnetic field formed when a current is passed through a single copper plate, and Thus, the frequency shift amount of the resulting NMR signal was analyzed. The sample was' H water.

 [0297] First, the analysis principle will be explained.

 When the current I flows through the conductor, a magnetic field Hi is formed around the conductor according to Bio-Savart's law. Its magnetic field strength is proportional to the current I. This magnetic field is obtained by analysis.

[0298] In NMR measurement, a static magnetic field H _Q is applied to a sample by a magnet. FIG. 55 is a diagram showing directions of a static magnetic field H 0 and a magnetic field H i described later.

 Also, suppose that the conductor is placed in the static magnetic field and flows from left to right in Fig. 55. This current creates a magnetic field around the conductor. As a result, the magnetic field applied to the sample is the sum of the static magnetic field Ho by the magnet and Hi formed by the current.

[0299] Furthermore, the frequency ω [H z] of the N MR signal is proportional to the magnetic field strength H [ga u s s], as shown in the following equation.

 ω = H

 = Ύ (Ho + Hi)

In the above formula, Γ is the nuclear gyromagnetic ratio [H z / gauss], and in the case of ¹ H hydrogen nucleus, it is 4260 H z / gauss.

[0300] From the above equation, in this embodiment, the magnetic field Hi that increases or decreases when the current I flows through the conductor is obtained as the frequency shift amount Δω [H z].

FIG. 56 is a perspective view showing the positions of the copper plate, RF detection coil, and water sample in the magnetic field analysis of this example.

In FIG. 56, the magnetic field Hi created by the current I flowing through the conductor can be calculated based on Bio-Savart's law. Specifically, when the conductor is placed in a vacuum (magnetic permeability is ^{4 π X 1 0- 7 N /} A 2), conductors position _{(x p, y p, z} p

The magnetic field Hi (x _p , y _p , z _p ) created in) is expressed by the following equation. FIG. 57 is a diagram showing a coordinate system in the following equation.

[0302] [Equation 2]

[0303] The symbols in the above formulas represent the following, respectively.

 Hi: Magnetic field strength at position r [A / m] (vector)

r: Position of point P in space (x _p , y _p , z _p ) [m] (vector)

r ': position of point Q on the coil (x _q , y _p , z _p ) [m] (vector)

 I: Current [A] (scalar)

 t: Unit vector indicating the direction of current flow (The copper plate is uniformly upward in the figure.) [-] (vector)

 [0304] Also, when performing the integration of the above equation, it was calculated numerically using the following approximate calculation method.

 In other words, assuming that the copper plate is rectangular and the current flows uniformly in the copper plate, the copper plate is divided into small elements, and it is assumed that the current divided equally flows through the elements. Specifically, there is a small element at point Q, a current flows for each element, and these multiple elements at point Q create a magnetic field at point P. The magnetic field at point P was calculated by integrating all elements at point Q using the above equation. In this analysis, the longitudinal direction of the copper plate (L direction in Fig. 57) was equally divided into 64, and the width direction (W direction in Fig. 57) was equally divided into 32.

[0305] Based on the above assumptions, we analyzed the z-direction distribution of the x-direction magnetic field H _x formed on the copper plate center (x _p = Omm, y _p = Omm) when the current I [A] flows .

FIG. 58 shows the analysis results. Figure 58 shows the results when the current I is changed to _ 1 [A], -0.5 [A], 0 [A], 0.5 [A], and 1 [A]. In FIG. 58, the unit of the frequency shift amount Δω on the vertical axis is [rad / ms]. Furthermore, the experimental results of Example 5_2 described later In order to contrast with the results, H z was multiplied by 27Γ to convert to rad, and s was divided by 1 000 to convert to ms.

The “measurement position” indicated by the arrow in FIG. 58 is the position of the water sample measured by the RF detection coil, and z _p = 0.45 mm. From Fig. 58, the distribution of the frequency shift Δω at this position is almost flat. The distribution near the copper plate is very steep, but it can be seen that the distribution is almost flat just by moving slightly away from the copper plate. This flat area becomes the measurement area. The frequency shift Δω in this region increases or decreases with the magnitude of the current. By measuring this frequency shift amount Δω, the current can be calculated backward.

 FIG. 59 is a diagram showing the relationship between the frequency shift amount Δω and the current I at the position of the water sample measured by the RF detection coil. From Fig. 59, it can be seen that the relationship between the two is directly proportional. It can also be seen that the current I can be easily converted from the frequency shift Δω.

 [0308] (Example 5 _ 2)

 In this example, the amount of frequency shift was measured using a small surface coil and compared with the result of Example 5-1.

 [0309] A copper plate, an RF detection coil, and a water sample (pure water) were placed in the positional relationship shown in Fig. 55, and the relationship between the current I flowing through the copper plate and the frequency shift amount Δω was measured. .

 As a small surface coil, a copper wire with a polyurethane film wire diameter of 5 Om was used, and it was manufactured by winding it three times in a spiral shape with an outer diameter of 1.3 mm. FIG. 60 is a diagram showing a small surface coil produced in this example.

 As the N MR measurement system, a base system manufactured by MRL Technology was used with high sensitivity. The magnet used was a modified Halbach magnetic circuit with a magnetic field strength of 1. O Tes Ia and an air gap of 45 mm manufactured by Neomax.

[0310] FIG. 61 is a diagram showing the relationship between the current I [A] flowing through the copper plate and the measured frequency shift amount Δω [rad / ms]. Fig. 61 shows the magnetic field of Fig. 59. The relationship (solid line) obtained from the analysis results is also shown.

 From Fig. 61, it can be seen that the measured values in this example are in good agreement with the analysis values in Example 5-1. Therefore, the frequency shift Δω of the NMR signal corresponding to the current value of the copper plate (_ 1 A to 1 A) can be measured experimentally, and it can be seen that the relationship between the two is directly proportional.

[0311] (Example 6)

 In this example, multi-point measurement of the frequency shift amount during the water electrolysis operation of ME was performed.

[0312] (Example 6_ 1)

 In this example, prior to actual measurement in Example 6_2 described later, the magnetic field was analyzed when ME A was modeled with a one-dimensional equivalent circuit.

[0313] FIG. 62 is a cross-sectional view showing a schematic configuration of MEA used in the analysis in this example. As shown in Fig. 62, the MEA is an assembly of an electrode and a polymer electrolyte membrane, and the polymer electrolyte membrane (PEM) is sandwiched between upper and lower current collectors.

[0314] In the analysis of this example, the dimensions of ME A are 23 mm X 2 Omm square, thickness 356

; Um. The current collector was a one-pong mesh with a thickness of 30 Om. Pt and Ir were electrolessly attached to one surface of the polymer electrolyte membrane to form an anode-side catalyst layer. In addition, Pt was unelectrolyzed as a catalyst on the other surface of the polymer electrolyte membrane to form a force sword side catalyst layer.

 In the magnetic field analysis of this example, it was assumed that no current flows through the catalyst layer having a large electric resistance, and the current flows through the current collector.

[0315] The RF detection coil had an inner diameter of 0.6 mm and was wound 5 times. There are two coil positions, x = _4.4mm and 2.1mm, with the center of MEA as the origin. The coil spacing is 6.5 mm. The measurement area in the coil depth direction is about one-fifth of the coil diameter, and is within a disc-shaped area about 0.1 mm from the PEM surface.

[0316] In Fig. 62, the RF detection coil is embedded in the lower side (electrolyte membrane side) of the bonbon mesh and is in contact with PEM with catalyst inside ME A. Also The RF detection coil is sandwiched between the upper force mesh and the lower force mesh. The magnetic field created by these two conductors is shown in Fig. 5.

This is different from the case of a single conductor as described in 5.

[0317] That is, in FIG. 62, a state in which the MEA is operated in the water electrolysis operation is considered. Apply a DC voltage and apply a current I with the anode side collector (collector) of ME A as the anode and the force sword side collector as the negative electrode. The current flowing through the anode side collector electrode is I ₂ and the current flowing through the I force source side collector electrode is I ₂ .

[0318] The current flowing through the collector and PEM in this state was assumed as shown in Fig. 63. In other words, it was assumed that current was supplied to the left end of the anode-side collector electrode, and that current flowed out of the anode side and the opposite end on the force sword side.

 In addition, it was assumed that the protons transmitted uniformly through the PEM and had the same resistance value in the entire region within the PEM plane.

 Based on this assumption, the current distribution flowing through the anode-side collector electrode and the cathode-side collector electrode is a distribution that decreases or increases linearly with respect to the X direction. This was assumed to be an equivalent circuit of ME A.

[0319] Based on the above assumptions, in this example, when the current I supplied from the power source is 1.2 A, the magnetic field formed in the PEM is prayed by the method described in Example 5. The frequency shift amount of the N MR signal at the position where the small surface coil was placed was calculated.

 [0320] FIG. 64 is a diagram showing the frequency shift amount Δω analyzed with respect to the positions of the cross section a_a and the cross section b_b shown in FIG. In FIG. 64, the horizontal axis represents the frequency shift (rad / ms), and the vertical axis represents the position z (mm) in the thickness direction of the PEM. The coil is placed at the top end, and the z position on the vertical axis is the top end of the PEM (P t + I r side) at z = _ 1 78 m, and the bottom end (P r side at z = 1 78 m )

[0321] From Fig. 64, the frequency shift Δω (solid line) in the z-axis direction (PEM thickness direction) at the position where the RF detection coil A is placed (x = 2.1 mm) (solid line) and RF detection Frequency shift in the z-axis direction at the position (x = _4.4 mm) where coil Β is placed The amount of rotation Δω (broken line) is known.

 From Fig. 64, the frequency shift amount at the position where the RF detection coil Α is placed, that is, the frequency shift amount at the cross section a _ a (Fig. 63) is about 6 rad / ms from the upper side to the lower side. Decreases from about 1 rads. In the cross section a_a, the current I 1 (at s e c t i o n a _ a) force flowing in the upper collector electrode is larger than the current I 2 (at s e c t i o n a _ a) flowing in the lower collector electrode (I ι>

I 2).

[0322] —On the other hand, at the position where the RF detection coil B is placed, that is, in the cross section b_ b (Fig. 63), the current I ₂ (atsection b_ b) that flows to the lower collector electrode flows to the upper collector electrode Larger than current I, (atsection b_b) (1 ₂ >). As a result, the frequency shift amount in the cross section b_b is opposite to the frequency shift amount in the cross section a_a. Also, the amount of frequency shift at the cross section b_b decreases from about 0 ra 01/3 to about _4 rad / ms from the upper side to the lower side of the PEM.

 [0323] (Example 6 _ 2)

 In this example, the frequency shift distribution when the ME A shown in FIG. 62 was water electrolyzed was measured and compared with the analysis results of Example 6-1.

 (M E A)

 Pt and Ir are electrolessly attached to one side (anode side) of a polymer electrolyte membrane manufactured by Asahi Glass Co., and Pt is electrolessly attached to the other side (force sword side). Produced. The dimensions of ME A are 23 mm x 2 Omm square and thickness 356 m.

[0324] As shown in Fig. 62, the obtained MEA was sandwiched by a force of one meter mesh (manufactured by Japan Gore-Tex) with a thickness of 300 m. When sandwiching the carbon mesh and ME A, the 0.03 mm thick Pt electrode foil was also sandwiched. The current from the power source was changed from the conductor to the Pt electrode foil, Force —Bonmesh, Pt electrode foil, and lead wire flow in this order to return to the power source. The applied voltage during water electrolysis was about 3 V, and the current density was 0.26 A / cm ² . The temperature of ME A during water electrolysis operation was about room temperature. In this embodiment, no steam is supplied to ME A.

[0326] (Small surface coil)

 In this example, a N MR measuring surface coil having an inner diameter of 0.6 mm was used. A small hole was made in the bonbon mesh, and the lead part of the coil was passed there to fix the coil on the bonnet mesh.

 FIG. 65 is a diagram showing a coil used in this example. The small surface coil was made by winding a copper wire with a polyurethane film diameter of 40 m in a spiral shape with an inner diameter of 0.6 mm five times in a flat shape. This coil was manufactured by Star Engineering.

 FIG. 66 is a view showing a coil placed on a carbon mesh in this embodiment.

 [0327] In this example, as in Example 6_1, the center position of coil A is x =

2. 1 mm, coil B center position X = _ 4.4 mm. The origin of position x is the center of ME A.

[0328] (Measurement result)

 The frequency shift amount was determined according to the method described in Example 1. Figure 67 shows the frequency shift measured by coils A and B when the current value is 1.2 A.

 In Fig. 67, the amount of frequency shift in coil A is indicated by a white square (mouth), and the amount of frequency shift in coil B is indicated by a white triangle (△). The measurement area in the depth direction of the coil is about one-fifth of the coil diameter, so the area has a width of about 0.1 mm from the PEM surface. This width is shown as a bar in Figure 67.

[0329] From FIG. 67, the frequency shift amounts measured by coils A and B were the same as the analysis results in Example 6_1. Similarly to Example 6_1, the frequency shift amounts measured by coils A and B are opposite to each other. Thus, the same relationship was obtained between the analysis value and the measurement value.

 According to this example, the frequency shift amount could be measured at multiple points even in the system corresponding to the operating fuel cell. By using this method, it is possible to know the local distribution of the current in the MEA of the fuel cell.

[0330] (Example 7)

 (Measurement of frequency shift during water electrolysis operation)

 The current measurement method based on NMR was applied to ME A (Membrane Electrode Assembly) in which electrodes were joined to PEM, and the frequency shift of the NMR signal was measured by PEM during water electrolysis operation. In addition, in order to confirm its validity, magnetic field analysis was also performed and compared with the measurement results.

[0331] (Equivalent circuit of ME A)

In order to confirm the validity of the measurement compared with the measured value, the magnetic field analysis in ME A was performed, and the frequency shift amount of the N MR signal that increases or decreases due to the magnetic field was calculated. As shown in FIG. 68, the magnetic field analysis was performed with an electrode arrangement in which current was supplied to the end of the MEA and the current flowed from the opposite end of the force sword. In this analysis, it is assumed that the protons pass through the PEM uniformly, have the same resistance value in the entire area of the PEM, and further, current flows through the electrodes on the anode side and cathode side of the PEM. It is assumed that a closed circuit is constructed. Based on this assumption, the current I, flowing through the anode electrode decreases linearly with position X, and conversely, the current I ₂ flowing through the force source electrode increases linearly with position X. The current distribution. The current flowing through the anode and cathode electrodes forms a magnetic field inside the PEM. This magnetic field strength distribution Hi was analyzed using Bio-Savart's law, and the frequency shift amount of the N MR signal was calculated from the magnetic field strength. The origin of position X is the center of PEM.

[0332] Figure 69 shows the z-direction distribution of frequency shifts analyzed at the four sensor positions. Position (x = _7. 5mm) at the anode side of the current I, of the sensor _ A but larger than the current I ₂ forces cathode side, as a result, the frequency shift amount has a positive value. Conversely, at the position of sensor D (x = 7.5 mm), the force sword side The current I2 of the current is also large on the far node side, and the frequency shift amount in that case is a negative value opposite to that of the sensor A. Sensors _B and C located in the middle of both are the frequency shift amount in the middle.

 [0333] (ME A water electrolysis operation and sensor position)

 The ME A used in the water electrolysis operation was manufactured by electrolessly attaching Pt and Ir on the anode side of the polymer electrolyte membrane and Pt on the force sword side. The dimensions of ME A are 23 mm x 20 mm square and 356 m thick.

[0334] As shown in Fig. 70, this MEA is sandwiched as a GDL (Gas Diffusion Layer) by a 300m thick force pon mesh, and current is supplied from the stabilizing power source through the Pt electrode. In order to bring the contact resistance between GDL and MEA closer to uniform, a structure is used in which pressure is applied from both sides of the positive and negative electrodes using a cushioning material so that GDL and MEA are in uniform contact. Applied voltage current density of about 3 V was 0. 26A / cm ^2.

 [0335] The surface coil for NMR measurement was a copper wire with an inner diameter of 0.6 mm and a wire diameter of 0.04 mm, and four coils were placed at 5 mm intervals between the MEA and the force pon mesh. Polyurethane coating is applied to the surface coil wire, and the carbon mesh is in an insulated state. The measurement area in the coil depth direction is about one-fifth of the coil diameter, and is a disk-like area with a depth of about 0.1 mm from the PEM surface. The representative value of the frequency of the NMR signal measured by this coil is considered to be the center of this disk, and its depth is assumed to be 0.05 mm.

[0336] (Measurement result of frequency shift amount)

Figure 71 shows the measured frequency shift amount and the X-direction distribution of the frequency shift amount obtained from the analysis. The horizontal axis in this figure is position X, which corresponds to the position of each sensor. The vertical axis represents the frequency shift amount. In the figure, the country (black square) force <<, sensor _A ~ D of the position X of the frequency shift measured value is shown. The solid line shows the amount of frequency shift calculated using the analysis method shown in the (MEA equivalent circuit) section and considering the thickness of GDL. In this case, the frequency shift The distance from the anode surface position (z = 1 78 Um) to the coil measurement center was the value at the position (z = 1 28 m) where the distance from the coil center was 0.05 mm.

 [0337] From Fig. 71, it can be seen that the amount of frequency shift gradually decreases from sensor A to D, and the analytical value of the solid line and the measured value of the garden almost coincide. It can also be seen that A and D have opposite frequency shift amounts. This indicates that the frequency shift amount of the NMR signal corresponding to the current distribution flowing on the MEA surface can be captured.

 In addition, the experimental values for sensors _C and D are slightly larger than the analytical values, but this is because the force bon mesh and MEA are not joined with a spatially uniform contact resistance, resulting in a slight current distribution. I think this is because of the bias.

 [0338] (Example 8)

 In this example, a small surface coil is inserted between the GDL and PEM in the fuel cell, and the current value when the fuel cell is operated in water electrolysis is measured from the frequency shift amount of N MR. The water content was also measured. If water electrolysis is performed without supplying water, the PEM will gradually dry and the current will gradually decrease. In this example, the spatial distribution of current and water content in this case was measured in time series.

 [0339] As an experiment for alternately measuring the frequency shift amount and the water content, here we used the same ME A and small coil as in Example 7 to determine the frequency shift amount and the water content during water electrolysis operation. Measurement was performed. FIG. 72 is a diagram showing the arrangement of MEAs and small coils in the present embodiment. As shown in Fig. 72, small coils are inserted between PEM and GDL, and the number is 3 on the anode side and 1 on the force side.

[0340] In this example, the direction of the static magnetic field HQ is opposite to that of Example 7. When installing the cell in the magnet, the cell direction was reversed so that the direction of the static magnetic field was reversed. As a result, the sign of the frequency shift amount measured in Example 7 is reversed. For example, the absolute value of the frequency shift amount for sensor A is the same, but the sign is negative. [0341] MEA was immersed in distilled water until immediately before the experiment, and the surface water was wiped off with Kimwipe just before the experiment to obtain an appropriate water content. As described above in Example 2, the water content of ME A when treated in this way is about 1 0 [H ₂ 0 / S 0 ₃ −H +]. Even in this experiment, the water content is considered to be about 10 [H ₂ 0 / S0 ₃ -H +].

 [0342] (Water electrolysis operation and measurement procedure)

 The frequency shift amount was measured by the same PGS E method as in Example 7, and the water content was measured by the CPMG method. FIG. 73 is a diagram showing the measurement timings of PGS E and C PMG in this example when the time when water electrolysis operation was started (voltage was applied) was set to zero.

 [0343] As shown in FIG. 73, in this example, the P G S E measurement and the C P M G measurement were repeated alternately. Specifically, the spin echo signal was acquired once in the PGS E measurement. In C PMG, the same measurement was performed five times to obtain an echo signal. In CPMG measurement, the measurement was performed 5 times, and the second echo signal intensity obtained was averaged for 5 times to obtain the measured value. This set was repeated with one PGS E and five C PMGs as one set. TR is 5 seconds, and the time required to measure this one set is 30 seconds. After applying voltage, 6 sets of measurements were taken. In CPMG measurement, 2 = 20 ms. The time to acquire an echo signal in one measurement is 1 second.

Here, in order to calculate the amount of frequency shift from the echo signal obtained by PGS E, the echo signal obtained before applying a voltage to ME A is required. Therefore, in this example, as shown in FIG. 73, a series of measurements was started before applying a voltage (“O ff” in the figure) and used as a reference echo signal. At this time, the temperature of the magnet increases or decreases with time, the static magnetic field strength increases or decreases, and the frequency of the N MR signal may change. In this measurement, the amount of change in frequency was about 1 OOH z in 1 hour. The frequency that changes in the experiment time of 150 seconds, which was conducted this time, is about a few Hz, which is sufficiently smaller than the frequency shift amount (about 1 kHz) that increases or decreases with the current. For this reason, it is assumed that the reference frequency is the same as the value (echo signal waveform) obtained before applying the voltage. Can do. Here, we calculated the frequency shift amount based on this assumption.

 [0345] Unlike the experimental state described above, when the frequency increases or decreases with time, an NMR detection coil is installed in a place that is not affected by the magnetic field formed by the current flowing through the MEA. By measuring the NMR frequency and using it as the reference frequency, the time change of the static magnetic field strength can be offset.

 [0346] The voltage applied to ME A was 3.4 V at maximum, the maximum current was 1.2 A, and a DC power supply was set. FIG. 74 is a diagram showing the time change of the current flowing through M EA. FIG. 75 is a diagram showing the time change of the voltage applied to MEA.

 [0347] As shown in Fig. 74 and Fig. 75, about immediately after the voltage is applied to M EA

A current of 1.2 A flowed for about 10 seconds. The current density at this time was 0.25 A / cm ² . After about 10 seconds, the voltage applied to ME A reached 3.4 V, and at the same time, the current decreased to about 0.8 A. After that, the current gradually dropped to about 0.5 A.

 [0348] In this example, PGSE measurement is performed six times at a time interval of about 30 seconds, which corresponds to the time indicated by the arrows in FIG. 74 (PGS E # 1 to # 6). In this way, the frequency shift was measured while the current was decreasing. On the other hand, C PMG was measured 5 times during P G S E.

 [0349] (Comparison between measurement results and frequency shift obtained from magnetic field analysis)

FIG. 76 is a diagram showing the frequency shift amounts obtained by the three sensors A, C, and D from the measurements in PGS E # 1 and # 4 (FIG. 74). In FIG. 76, the horizontal axis represents the sensor position x and the vertical axis represents the frequency shift amount, as in FIG. 71 shown in Example 7 above. In Fig. 76, the plot for country (black square) shows the frequency shift measured with PGSE # 1 (current is 1.2 A). The ▲ (black triangle) plot shows the frequency shift measured with PGS E # 4 (current is 0.6 A). In addition, the solid line and the alternate long and short dash line in the figure are analysis values calculated by the same magnetic field analysis as in Example 7. [0350] The result shown in Fig. 76 differs from the result of Fig. 71 in the sign of the frequency shift amount, because the static magnetic field direction is reversed. The direction of the current flowing through the GDL is the same, but if the direction of the static magnetic field is reversed, the sign of the frequency shift amount is reversed. If the measurer knows the direction of the static magnetic field, it will not be a problem.

 [0351] From Fig. 76, it can be seen that the measured and analyzed values are in good agreement for both 1.2A and 0.6A. If the current decreases, the amount of frequency shift decreases accordingly, and it can be seen that it depends on the position of the sensor. In addition, since the measured values and the analyzed values agree with each other, as assumed in the analysis, it can be inferred that in this water electrolysis experiment, a state where a current flows almost uniformly in ME A can be achieved. Monkey.

 [0352] Fig. 77 is a diagram showing the time change of the frequency shift amount obtained by the measurement from PGS E # 1 to # 6. In Fig. 77, the measured and analyzed values of the three sensors A, C, and D are also shown. From this figure, the measured values and the analyzed values are in good agreement. Therefore, even when the current flowing through the MEA decreases transiently from 1.2A to 0.6A, it can be inferred that the current flows uniformly through the MEA.

 [0353] (Measurement result of water content in PEM)

 Next, the results of measuring the water content in PEM by C PMG measurement will be explained.

The time change of the echo signal intensity acquired on the anode side (sensors _A, C, D) is shown in Fig. 78, and the echo acquired on the force sword side (sensor 1E) is shown in Fig. 78. Figure 79 shows the signal strength over time. Here, the echo signal intensity is the intensity of the echo signal that was observed second using the C PMG method, and the average signal intensity when five C PMG measurements were performed. The values on the vertical axis are normalized by the average signal strength of 4 sets measured before applying current.

[0354] As shown in Figure 80, the signal intensity increases as the water content in the PEM increases from another experiment. However, the relationship between the two is not strictly a direct proportional relationship. However, in this embodiment, for the sake of simplicity, it is assumed that the two are in a positive correlation and that the signal intensity is almost directly proportional to the water content.

 [0355] As shown in Fig. 78, when water electrolysis was started by applying current, the moisture content in sensor A on the anode side decreased, and when the current was returned to zero, it returned to the original moisture content. On the other hand, sensors C and D had almost constant water content.

 [0356] On the other hand, the water content once increased on the force source side, and then gradually decreased. Sensors A and E are almost in opposite positions, as shown in Fig. 72.

 [0357] Fig. 8 1 is a diagram for explaining a phenomenon occurring in P E M during water electrolysis operation of M EA. As shown in Fig. 81, water moves to the power lead side by electroosmotic flow in the PEM, and further water is decomposed by electrolysis on the anode side, so the water content on the anode side decreases. To do. On the other hand, on the force sword side, the water content once increases due to electro-osmotic flow. However, if water electrolysis continues, the water content of the entire P EM will decrease due to electrolysis, and eventually the water content will gradually decrease with time even on the power source side.

 [0358] In this example, water electrolysis operation was performed for 150 seconds, but the amount of water decomposed by this amount of current was calculated to be about several percent of the amount of water contained in PEM. For this reason, the increase or decrease in water content on the anode side or the force sword side is considered to be mainly caused by electroosmotic flow. In addition, after water electrolysis is stopped, the water content in PEM will be slightly reduced.

 [0359] From the results of Fig. 7 8 and Fig. 7 9 obtained here, the sensor captures the above phenomenon, decreases on the anode side, temporarily increases on the force sword side, then decreases, and then decreases. It can be said that the results returned to almost the same as the original water content after the electrolysis was stopped.

[0360] In this example, the frequency shift measurement by the PGSE method and the moisture content measurement by the CPMG method are shown alternately. However, the frequency shift measurement and the moisture content measurement are common. You may carry out simultaneously by a pulse sequence. Further, in this embodiment, an example in which the moisture content is measured by the frequency shift measuring device is shown, but the mobility of water can also be measured by the frequency shift measuring device.

Claims

The scope of the claims  [1] A device that locally measures the current at a specific location on a sample using nuclear magnetic resonance.  A static magnetic field application unit that applies a static magnetic field to the sample;  Applying a vibrating magnetic field for excitation to the sample, and obtaining a nuclear magnetic resonance signal generated at a specific location of the sample; a small RF coil smaller than the sample;  Calculating a difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coil and the frequency of the excitation oscillating magnetic field, and calculating a current of the specific portion of the sample from the difference;  A measuring apparatus comprising: [2] In the measuring device according to claim 1,  A detector that detects a real part and an imaginary part of the nuclear magnetic resonance signal; and the current calculation unit uses the real part and the imaginary part detected by the detector to detect the frequency of the nuclear magnetic resonance signal. And a difference between the frequency of the oscillating magnetic field for excitation and a measuring device. [3] In the measuring apparatus according to claim 2,  A plurality of said small RF coils,  The plurality of small RF coils apply the excitation oscillating magnetic field to a plurality of locations of the sample, and acquire the nuclear magnetic resonance signal.  The measuring apparatus configured to calculate the current at the plurality of locations of the sample. 4. The measuring device according to claim 3, wherein the sample is a film. [5] The measuring device according to claim 4,  The small RF coil applies the pulsed oscillating magnetic field and obtains an F I D signal corresponding to the oscillating magnetic field for excitation, The measurement apparatus, wherein the current calculation unit acquires a real part and an imaginary part of the FID signal. [6] In the measuring apparatus according to claim 4,  The small RF coil is  (a) 90 ° pulse, and  (b) Applying the excitation oscillating magnetic field and obtaining an echo signal corresponding to the excitation oscillating magnetic field in a pulse sequence including a 180 ° pulse applied after the elapse of the pulse time of (a). And  The measurement apparatus, wherein the current calculation unit acquires a real part and an imaginary part of the echo signal.  [7] The measuring device according to claim 4,  A solvent amount calculation unit for calculating the amount of the protonic solvent in the sample based on the nuclear magnetic resonance signal acquired by the small RF coil;  A switching unit for switching between a first measurement mode for measuring the current of the sample and a second measurement mode for measuring the amount of the protonic solvent in the sample;  Further comprising  When in the first measurement mode, the current calculation unit is configured to detect the specific portion of the sample based on a difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coil and the frequency of the oscillating magnetic field for excitation. Perform a current calculation,  When in the second measurement mode, the solvent amount calculation unit calculates the amount of the protonic solvent at the specific location in the sample based on the nuclear magnetic resonance signal acquired by the small RF coil. Perform the measuring device. [8] In the measuring device according to claim 7,  In the second measurement mode,  The small RF coil acquires an echo signal corresponding to the excitation oscillating magnetic field, The solvent amount calculation unit calculates a T ₂ relaxation time constant from the intensity of the echo signal, and calculates the amount of the protonic solvent at the specific location in the sample from the calculated ₂ relaxation time constant. Calculate the measuring device. [9] The measuring device according to claim 8, The small RF coil is  (a) 90 ° pulse,  (b) 1 80 ° pulse applied after the time of pulse in (a), and  (c) n 1 80 ° pulses (n is a natural number) that starts at the end of time 2 of the pulse in (b) and is applied at time 2 intervals.  Applying the excitation oscillating magnetic field with a pulse sequence including:  In the first measurement mode, the small RF coil acquires the FID signal corresponding to the pulse (a) or the echo signal corresponding to the pulse (b) or (c), and the current The calculation unit acquires the real part and the imaginary part of the FID signal or the echo signal, In the second measurement mode, the small RF coil acquires a plurality of echo signals corresponding to the pulses of (b) and (c), and the solvent amount calculation unit includes the plurality of solvent signals. A measuring device that calculates the T ₂ relaxation time constant from the intensity of an echo signal.  The measuring apparatus according to claim 4,  A gradient magnetic field application unit for applying a gradient magnetic field to the sample;  Based on the nuclear magnetic resonance signal acquired by the small RF coil, a mobility calculation unit that calculates the mobility of the protonic solvent in the sample,  A switching unit that switches between a first measurement mode for measuring the current of the sample and a third measurement mode for measuring the mobility of the protonic solvent in the sample;  Further comprising  In the third measurement mode,  The small RF coil applies the excitation oscillating magnetic field to the sample, and acquires a nuclear magnetic resonance signal corresponding to the excitation oscillating magnetic field and the gradient magnetic field, The mobility calculation unit calculates the mobility of the specific portion of the sample based on information of the nuclear magnetic resonance signal obtained corresponding to different gradient magnetic fields. , measuring device.  [11] An apparatus for obtaining a current distribution in a plane of a solid polymer electrolyte membrane of a fuel cell using a nuclear magnetic resonance method,  A static magnetic field applying unit that applies a static magnetic field to the solid polymer electrolyte membrane; and an oscillating magnetic field for excitation applied to the solid polymer electrolyte membrane, and generated at a specific location of the solid polymer electrolyte membrane. A plurality of small RF coils that are smaller than the solid polymer electrolyte membrane,  For the plurality of small RF coils, the difference between the frequency of the nuclear magnetic resonance signal acquired by the small RF coils and the frequency of the oscillating magnetic field for excitation is calculated. A current distribution acquisition unit that acquires an in-plane current distribution of the solid polymer electrolyte membrane from the difference; and  A measuring apparatus comprising:  [12] A method for locally measuring the current at a specific location of a sample using nuclear magnetic resonance.  First, an excitation oscillating magnetic field is applied to a specific location of the sample placed in a static magnetic field using a small RF coil smaller than the sample, and a nuclear magnetic resonance signal generated at the specific location is acquired. Steps,  A second step of calculating a difference between the frequency of the nuclear magnetic resonance signal acquired in the first step and the frequency of the excitation magnetic field, and obtaining a current at the specific location of the sample from the difference;  Including the measuring method. [13] In the measurement method according to claim 12,  In the second step, the real part and the imaginary part of the nuclear magnetic resonance signal are detected, and the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the excitation oscillating magnetic field is determined using the real part and the imaginary part. Calculate the measurement method. [14] In the measurement method according to claim 13; In the first step, the small RF coil applies the oscillating magnetic field for excitation in the form of a pulse, and generates an FID signal corresponding to the oscillating magnetic field for excitation. Acquired,  In the second step, the real part and the imaginary part of the FID signal are detected, and the difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation is calculated using the real part and the imaginary part. , Measuring method. [15] In the measurement method according to claim 1 3,  In the first step, the small RF coil is  (a) 90 ° pulse, and  (b) Applying the excitation oscillating magnetic field and obtaining an echo signal corresponding to the excitation oscillating magnetic field in a pulse sequence including a 180 ° pulse applied after the elapse of the pulse time of (a). And  In the second step, a real part and an imaginary part of the echo signal are detected, and a difference between the frequency of the nuclear magnetic resonance signal and the frequency of the oscillating magnetic field for excitation is calculated using the real part and the imaginary part. , Measuring method. [16] In the measurement method according to claim 13,  Applying an oscillating magnetic field for excitation and a gradient magnetic field to a specific location of the sample, and acquiring a nuclear magnetic resonance signal generated at the specific location and a third step of acquiring a nuclear magnetic resonance signal generated at the specific location. A fourth step for calculating the mobility of the specific part of the sample based on the information and the information of the nuclear magnetic resonance signal obtained in the third step;  Further including  In the first step and the third step, a local magnetic field is applied to the specific location of the sample using the small RF coil, and a nuclear magnetic resonance signal is acquired from the specific location,  In the first step, application of a gradient magnetic field to the sample is executed according to a predetermined pulse sequence, In the third step, the application of the gradient magnetic field having a magnitude different from that in the first step is performed according to a predetermined pulse sequence. [17] In the measuring method according to claim 13;  In the first step, the small RF coil is  (a) 90 ° pulse,  (b) 1 80 ° pulse applied after the time of pulse in (a), and  (c) n 1 80 ° pulses (n is a natural number) that starts at the end of time 2 of the pulse in (b) and is applied at time 2 intervals.  And applying the excitation oscillating magnetic field to obtain the F I D signal corresponding to the pulse of (a) or the echo signal corresponding to the pulse of (b) or (c),  In the second step,  Using the real part and imaginary part of the FID signal corresponding to the pulse of (a) or the echo signal corresponding to the pulse of (b) or (c) above, the frequency of the nuclear magnetic resonance signal and the vibration for excitation In addition to calculating the difference from the magnetic field frequency, A T ₂ relaxation time constant is calculated from the intensities of a plurality of echo signals corresponding to the pulses of (b) and (c), and a profile at a specific location in the sample is calculated from the calculated Τ ₂ relaxation time constant. A measurement method for calculating the amount of a tonic solvent. [18] In the measurement method according to claim 17, In the second step, acquisition of an echo signal for calculating a difference between the frequency of the nuclear magnetic resonance signal and the frequency of the excitation magnetic field, and acquisition of an echo signal for calculating Τ ₂ relaxation time constant Measurement method that alternately and. [19] In the measurement method according to claim 17, In the second step, acquisition of an echo signal for calculating a difference between the frequency of the nuclear magnetic resonance signal and the frequency of the excitation magnetic field, and acquisition of an echo signal for calculating Τ ₂ relaxation time constant A measurement method that simultaneously