JPH0350218B2 - - Google Patents

Description

[Detailed description of the invention]

[Industrial Application Field] The present invention relates to a nuclear magnetic resonance (NMR) measurement method, and in particular to a method for selectively detecting multiple quantum transitions.
This relates to NMR measurement methods. [Conventional technology] Recently, two-dimensional measurement has become a new NMR measurement method.
NMR method is attracting attention. Since the NMR signal is developed as a two-dimensional spectrum, this method provides higher resolution (separation of multiplets) than conventional methods, making it easier to analyze the spectrum, and its range of applications will expand significantly in the future. It is expected that Furthermore, recently, a measurement method using this two-dimensional NMR measurement method (Multiple
Quantum NMR (hereinafter abbreviated as MQNMR)
is proposed, for example, in Japanese Patent Application Laid-Open No. 54-83890.
This proposed method uses a pulse train consisting of three 90° pulses P1 (non-selected pulse), P2 (non-selected pulse), and P3 (mixed pulse), as shown in Figure 1, for example, and the period τ is fixed. So, pulse P1
There is a phase difference of φ between the high frequency signals included in the pulse P3 and the high frequency signal contained in the pulse P3, and the free induction decay signal FID is detected and stored during the period t2. By the irradiation of the pulses P1 and P2, a non-equilibrium statistical state of the set of magnetic rotational resonators contained in the sample is pre-created, and then the phase of the vibrations characterizing the pre-created state of the resonators is Mixed pulse P3 whose phase is shifted by the phase angle φ
is applied. Then, the measurement is repeated while changing t1 in steps, and the measurement is further repeated while changing the value of φ. A two-dimensional spectrum is obtained by creating a linear combination of the value of t1 obtained in this way and the free induction attenuation signal stored corresponding to the various values of the phase shift, and double Fourier transforming it into the frequency domain. can get. However, in the conventionally proposed MQNMR method, the free induction attenuation signal is detected in one channel, and in such a one-channel detection, the high-frequency side or the low-frequency side is detected based on the frequency of the irradiated high-frequency pulse. Since only one of these is measured, a decrease in the S/N ratio is unavoidable,
There is a problem in that a large capacity memory is required. Also, regarding the spectral peak located at the irradiation frequency, even if the pulse is a 90° pulse,
There is also the problem that spectral peaks far from the irradiation frequency do not become 90° pulses, and therefore the phase inevitably changes as the distance from the irradiation frequency increases. Therefore, as a way to solve these problems,
It is conceivable to adopt quadrature detection (QD), which is used in conventional NMR measurement methods rather than two-dimensional NMR measurement methods. This allows the detection of free induction damped signals.
It is performed using two channels with a 90° phase difference. If this QD method is used, the irradiation frequency can be set at the center of the measurement area and the high frequency side and low frequency side of the irradiation frequency can be distinguished and measured. If the memory capacity is high, the resolution can be improved, and if the resolution is the same, the memory capacity can be reduced, which is an excellent effect. Moreover, assuming the same measurement area width as in the case of one channel, the deviation from the irradiation frequency at the end of the measurement area is half of that in the case of one channel, and therefore the phase deviation due to deviation from the irradiation frequency is also reduced. It turns out. However, using such a QD method,
Even if MQNMR is performed, the phase shift due to the deviation from the irradiation frequency will decrease, but it will not become zero, and the phase difference between the two-channel detection system may not be exactly 90 degrees, or the two channels may Due to a hardware imbalance in which the gains are not exactly equal, spectral peaks due to aliasing and ghost signals appear at positions where there is no original signal in the resulting two-dimensional spectrum. , which becomes an obstacle to analysis. [Object of the invention] The present invention has been made in view of this point,
The object of the present invention is to provide an MQNMR measurement method that uses the QD method but eliminates the above-mentioned disadvantages. [Characteristics of the Invention] The present invention is an NMR measurement method for selectively detecting multiple quantum transitions of a selected order in a sample containing a magnetic rotational resonator system, comprising: , applying a high-frequency pulse train consisting of a plurality of high-frequency pulses, in which the phase of at least one of the previous pulses is shifted by φ with respect to the last pulse; (b) after application of the high-frequency pulse train, a step called t2; (c) detecting and storing the free induction decay signal of the resonator over time with a two-channel detection system whose phases differ by 90 degrees from each other; (c) an evolution time t1 given as a specific pulse interval in the high-frequency pulse train; (d) sequentially setting the amount of phase shift to a predetermined series of values of φ and repeating steps (a) to (c) by one increment; (e) creating a set of linear combinations of free induction decay signals stored corresponding to the value of t1 and various values of the phase shift; (f) each of the series of values of φ; (g) creating a set of linear combinations of three free induction damping signals by repeating the above (a) to (e) for φ, which is obtained by adding 90°, 180°, and 270° to φ, respectively; (h) adding or subtracting the set of linear combinations of induced damped signals to form a set of linear combinations of free induced damped signals; and (h) the set of linear combinations of free induced damped signals produced in step (g). The method is characterized in that it consists of a step of performing a double Fourier transform on t1 and t2 into the frequency domain. Hereinafter, the present invention will be explained in detail using the drawings. [Example] FIG. 2 is a block diagram showing an example of an NMR apparatus for carrying out the present invention. In the figure, magnet 1
A sample coil 2 is placed within the static magnetic field generated by the sample coil 2, and a measurement sample is inserted into the space inside the sample coil 2. A high frequency signal having the resonant frequency of the observation nucleus generated from the high frequency oscillator 3 is given a predetermined phase by the variable phase shift circuit 4 which can select any phase from 0° to 360°, and then sent to the amplifier 5 and the gate. 6 to the coil 2 as a high-frequency pulse, and irradiates the sample. After the high-frequency pulse irradiation, the resonance signal induced in the coil 2 is transmitted to the gate 7.
and is sent to demodulation circuits 9 and 10 via reception circuit 8. The high-frequency signals from the high-frequency oscillator are sent as reference signals to the demodulation circuits 9 and 10, and one of them is sent through the 90° phase shift circuit 11, so the two demodulation circuits have a 90° phase shift. It constitutes a detection system with two different channels. The free induction attenuation signal obtained from this two-channel detection system is converted into a digital signal by A-D converters 12 and 13, sent to a computer 14, and stored in an attached memory 15. 16 is a pulse programmer that controls the phase shift circuit 4, gates 6, 7, and A-D converters 12, 13, and is used to control the order of the pulse train irradiated to the sample, the pulse width, the phase of the high frequency included in each pulse, A- The timing of sampling by the D converters 12 and 13 is programmed in advance, and a series of measurements are performed according to the program. MQNMR using a device with the above configuration
Triple Quantum Filter
The present invention will be explained by taking as an example a case where an experiment is performed. Figure 3 shows the three-quantum filter experiment reported by Piantini et al. (Journal of American
The pulse train used in Chemical Society Vol. 104, p. 6800-) is composed of three 90° pulses P1, P2, and P3, similar to the pulse train in Figure 1.
This pulse train differs from the pulse train in Figure 1 in that t1 and τ are reversed. The period τ is fixed, and if the phase of the high frequency included in pulse P3 is 0°, the phase of the high frequency included in pulse P1 is φ, and the phase of the high frequency included in pulse P2 is φ+90°. The free induction decay signal FID is detected and stored during period t2. Table A shows the combinations of phases given to P1, P2, and P3 in measurements using the pulse train of FIG.

[Table] In Table A, measurement 1 is a measurement with φ = 0°, and with the phases of P1, P2, and P3 set to 0°, 90°, and 0°, respectively, t1 is changed from 0 sec, for example.
512 FID signals can be obtained by repeating measurements while sequentially changing to 512 steps with equal steps of 2 msec.
Obtain and store FIDa0 to FIDa511. The next measurement 2 is a measurement set at φ = 60°,
With the phases of P1, P2, and P3 set to 60°, 150°, and 0°, respectively, change t1 from 0 sec in exactly the same way as measurement 1.
By repeating the measurement while sequentially changing to 512 steps in 2 msec steps, 512 FID signals FIDa0~
FIDa511 is acquired, and this FID signal is obtained from FIDa0 to FIDa511 previously obtained and stored in measurement 1.
are added (subtracted) with their signs reversed and linearly combined. Thereafter, measurements 3 to 6 are performed in exactly the same way, and the FID signals FIDa0 to FIDa511 obtained in each measurement are added or subtracted according to the ± sign described in the first column, and the linearly combined FID signals are The set Sa is obtained. The set Sa of the linearly combined FID signals is t1
By performing double Fourier transform on and t2, a two-dimensional spectrum can be obtained, for example, as shown in FIG. FIG. 4 schematically shows all four Fourier components obtained by double Fourier transform using contour lines. Measurement according to the above-mentioned procedure is introduced in the report by Piantini et al., and the apparatus for implementing it is the demodulation circuit 10 shown in FIG.
(and 90° phase shift circuit 11) is used, which has only one channel detection system.
Therefore, there are some drawbacks as mentioned above. Therefore, the present invention employs a QD system in which a two-channel detection system with a 90° phase difference is provided as shown in FIG. When the measurements shown in Table A are performed using the configuration shown in Figure 2, the Sa (FIDa0 to FIDa51) described above is obtained.
In addition to 1), from another channel, a set of FID signals Sa′ (FIDa0′ to FIDa51
1') is obtained. And this Sa(FIDa0～
FIDa511) and Sa′ (FIDa0′ to FIDa511′)
When a double Fourier transform (of which the Fourier transform regarding t2 is a complex Fourier transform) is performed using , a two-dimensional spectrum as shown in FIG. 5, for example, is obtained. The spectrum in Figure 5 is obtained when a high-frequency pulse is irradiated at the center of the measurement area, and because of the QD method, the measurement area is contained within 1/4 of the memory area, so It can be seen that memory capacity is saved. Resolution can be improved using the same memory capacity. However, if we simply set the detection system to two channels and perform the measurements shown in Table A, what we are detecting is only the FID signal during period t2, so t2
Regarding the axis F2 obtained by complex Fourier transformation of As can be seen, aliasing occurs, and in an actual spectrum in which complex peaks exist in a jumbled manner, this aliasing regarding axis F1 becomes an obstacle to analysis. Therefore, as a way to eliminate this folding, Table A
In addition to the measurements shown in Table B, the F1 axis can also be simulated by performing the measurements shown in Table B as shown below.
It is possible to apply the QD method.

[Table] The measurements in Table B are those in which the phases of P2 and P3 in the measurements in Table A are shifted by 90°, and the Sa(FIDa0~
FIDa511) and Sa′ (FIDa0′ to FIDa511′)
is the FID signal obtained from one detection system regarding t1, then Sb (FIDb0 to FIDb511) and Sb' (FIDb
0'~FIDb511') does not actually exist, but
This should correspond to the FID signal obtained from the detection system of another channel with a 90° phase difference with respect to t1. Therefore, the Sa obtained in the measurements in Table A
(FIDa0~FIDa511), Sa'(FIDa0'~FIDa5
11') and Sb (FIDb0~
FIDb511) and Sb'(FIDb0' to FIDb511')
By linearly combining (for example, addition or subtraction) and performing complex Fourier transform on both t1 and t2, the resulting two-dimensional spectrum is F as shown in Figure 6a.
It should be correct with the folds removed for both axes 1 and F2. However, in reality, due to the unbalance of the two-channel detection system mentioned earlier and the phase shift, which is reduced by the QD method but does not go to zero, the aliasing cannot be completely eliminated, and the resulting two-dimensional spectrum is distorted. As shown by the circle in Figure 6b,
It is inevitable that ghost signals will appear. Therefore, in the present invention, ghosts are removed by further adding measurements shown in Tables C and D shown below.

【table】

[Table] The measurements in Table C are obtained by further shifting the phases of P2 and P3 in the measurements in Table B described above by 90°, and the measurements in Table D are the phases of P2 and P3 in the measurements in Table C. Sc (FIDc0 to FIDc51
1) and Sc'(FIDc0' to FIDc511'), but in the measurement in Table D, Sd (FIDd0 to FIDd511) and Sd' (FIDd
0' to FIDd511') are obtained. Comparing Tables A and C, the phases of P2 and P3 are different by 180°, so Sa, Sa' obtained by the measurements in Table A and Sa, Sa' obtained by the measurements in Table C
If Sc and Sc' are added to obtain Sa+Sc and Sa'+Sc', the ghost component whose sign is reversed between the measurement in Table A and the measurement in Table C becomes zero. Similarly, when comparing Tables B and D, the phases of P2 and P3 are also different by 180 degrees, and therefore, Sb, Sb' obtained by the measurements in Table B and Sd, Sd obtained by the measurements in Table D ′ are added, and Sb+Sd is obtained.
If Sb'+Sd' is calculated, the ghost component whose sign is reversed between the measurement in Table B and the measurement in Table D becomes zero. Therefore, Sa+Sc, Sa′+Sc′ and Sb+Sd, Sb′+
Linearly combine (e.g. add or subtract) Sd′, t1,
If complex Fourier transform is performed on both t2, a two-dimensional spectrum from which ghosts are removed, as shown in FIG. 6a, is obtained, making it possible to perform correct analysis. In addition, even if the measurement and processing are performed according to the above procedure,
Ghost signals may appear due to imbalance in the two-channel detection system. In that case, for the measurements in Tables A to D above, measurements with 90° added to all phases and measurements with 180° added,
Ghosts can be removed by further performing measurements at an angle of 270° and linearly combining the data obtained from each phase measurement. 7 and 8 show two-dimensional spectra for explaining the effects of the present invention, and FIGS. 7a and 8a are spectra in which folding and ghosts appear, corresponding to FIG. Figure 7b and Figure 8b
corresponds to FIG. 6a, which is a spectrum in which both aliasing and ghosting have been removed by the present invention. The above explanation was about three quantum filters, but
The same applies to two-quantum filters and four-quantum filters. In addition, in the above explanation, the second and third pulses P2 and P3
Table B by changing the phase of 90° with respect to the measurement in Table A.
~D were measured, but it is common knowledge that the phase relationship between the pulses included in one pulse train is relative and does not need to be fixedly considered. For example, for the phase of each pulse in Table B,
By adding 270°, Table B' is obtained.Similarly, by adding 180° to the phase of each pulse in Table C and 90° to the phase of each pulse in Table D, Tables C' and D' are obtained, respectively. , respectively, are substantially equivalent to the measurements in the original Tables B-C.

【table】

【table】

[Table] Therefore, even if the four measurements of Table A, Table B', Table C', and Table D' are performed, the same effect as described above can be obtained.
For measurements in Tables B', C', and D', the phases of pulses P2 and P3 are set to be the same as those in Table A, and the phases of pulse P1 are set to φ+270°, φ+180°, and φ+180° in Table A.
They were each set at φ+90°. In the above explanation, we adopted the procedure of measuring t1 in 512 steps under one phase φ and repeating this measurement while changing φ. Even if the measurement is performed by changing φ to several steps, and then repeating this measurement by changing t1 to 512 steps, the number of data obtained is exactly the same and is equivalent. In that case, the measurement procedure is as follows. (a) Applying to a magnetic rotational resonator in the sample a high-frequency pulse train consisting of a plurality of high-frequency pulses, in which the phase of at least one of the preceding pulses is shifted by φ with respect to the last pulse; , (b) detecting the free induction decay signal of the resonator for a time t2 after application of the high-frequency pulse train with a two-channel detection system whose phases differ by 90 degrees from each other; (c) detecting a predetermined series of φ (d) repeating steps (a) and (b) by sequentially setting the amount of said phase shift to a value; (e) creating three more free induction damping signals by repeating the steps (a) to (d) for φ which is obtained by adding 90°, 180°, and 270° to each of the series of φ values; (f) adding or subtracting a set of linear combinations of the four free induction damped signals to create a set of linear combinations of the free induction damped signals; and (g) the frequency (h) repeating steps (a) and (f) by varying the evolution time t1 given as a particular pulse interval in the pulse train by one increment; (h) the free induction decay signal obtained in step (g); double Fourier transform the set of linear combinations of into the frequency domain with respect to t1 and t2. Furthermore, the present invention can also be applied to multi-quantum coherence. FIG. 9 shows an example of a pulse train used in a three-quantum coherence experiment, which is one type of multi-quantum coherence. This pulse train is a 90° pulse,
It consists of four pulse trains: 180° pulse, 90° pulse, and 90° pulse, and the phase of each pulse is set according to the sequence shown in Table E.

[Table] However, in the case of this measurement, P3 in Table E
It is also necessary to carry out the measurements in Table F with only the phase of +30°.

[Table] Then, add or subtract the data obtained from the measurements in Table E and the data obtained from the measurements in Table F, and further change the phase of P4 in Tables E and F to 90°, 180°, and 270°. Measurements according to Tables E-2 to E-4 and F-2 to F-4 below may be performed, and the obtained data may be linearly combined and subjected to double Fourier transform in the same manner as described above.

【table】

【table】

【table】

【table】

【table】

【table】

[Table] In addition, the above Tables E-2 to E-4 and Tables F-2 to F-
The values in the table for the phase of 4 are not absolute, for example, from the phase of each pulse in Table E-2 and Table F-2.
Table E-2' and Table F-2' are obtained by subtracting 90°, and Table E-3' and Table F are obtained by subtracting 180° from the phase of each pulse in Table E-3 and Table F-3.
-3' is obtained, and by subtracting 180° from the phase of each pulse in Table E-4 and Table F-4, Table E-4' and Table F-4' are obtained, both of which are the measurements of the original table. is substantially equivalent to

【table】

【table】

【table】

【table】

【table】

【table】

[Table] Therefore, Tables E, F, E-2', F-2', E-3',
The same effect as described above can be obtained even if measurements are made based on F-3', E-4', and F-4'.

[Brief explanation of drawings]

Figure 1 is a diagram showing a pulse train used in the proposed method, Figure 2 is a block diagram showing an example of an NMR apparatus for carrying out the present invention, and Figure 3 is a diagram showing a pulse train used in a three-quantum filter experiment. , Fig. 4 is a diagram schematically showing a two-dimensional spectrum obtained according to the conventionally reported procedure using contour lines, and Fig. 5 is a diagram showing a two-dimensional spectrum obtained by adopting the QD method. Figure 6a shows a two-dimensional spectrum from which ghosts have been removed according to the present invention, Figure 6b
is a diagram showing a two-dimensional spectrum in which a ghost appears, FIGS. 7 and 8 are diagrams for explaining the effects of the present invention, and FIG. 9 is a diagram showing a pulse train used in a three-quantum coherence experiment. . 1: Magnet, 2: Sample coil, 3: High frequency oscillator, 4: Variable phase shift circuit, 6, 7: Gate, 9, 1
0: demodulation circuit, 11: 90° phase shift circuit, 12,1
3: A-D converter, 14: Computer, 15:
Memory, 16: Pulse programmer.

Claims (1)

[Scope of Claims] 1. A nuclear magnetic resonance measurement method for selectively detecting multiple quantum transitions of a selected order in a sample containing a magnetic rotational resonator, comprising: (a) the resonator, applying a high-frequency pulse train consisting of a plurality of high-frequency pulses, in which the phase of at least one of the preceding pulses is shifted by φ with respect to the last pulse; (b) a time t2 after application of the high-frequency pulse train; (c) detecting and storing the free induction decay signal of the resonator with a two-channel detection system whose phases are different from each other by 90 degrees; (d) sequentially setting the amount of phase shift to a predetermined series of values of φ and repeating steps (a) to (c) by one increment; (e) creating a set of linear combinations of free induction decay signals stored corresponding to different values of said t1 and said phase shift; (f) for each of said series of values of φ; (g) creating a set of linear combinations of the three free induction damping signals by repeating the above (a) to (e) for φ with 90°, 180°, and 270° added, respectively; (g) the four free induction signals; (h) adding or subtracting the set of linear combinations of damped signals to form a set of linear combinations of free induction damped signals; and (h) adding or subtracting the set of linear combinations of free induction damped signals produced in step (g). A nuclear magnetic resonance measurement method comprising the steps of double Fourier transform to the frequency domain for t1 and t2. 2. The nuclear magnetic resonance measurement method according to claim 1, wherein the high-frequency pulse train consists of three 90° pulses, and the interval between two specific 90° pulses is selected to be t1. 3. A nuclear magnetic resonance measurement method for selectively detecting multiple quantum transitions of a selected order in a sample containing a magnetic rotational resonator, comprising: (a) applying a plurality of high-frequency pulses to the resonator; , applying a high-frequency pulse train in which the phase of at least one of the preceding pulses is shifted by φ with respect to the last pulse; (b) applying the high-frequency pulse train to the resonator for a time t2 after application of the high-frequency pulse train; (c) sequentially setting the amount of phase shift to a series of predetermined values of φ; (d) creating a set of linear combinations of the detected free induction decay signals corresponding to different values of said phase shift; (e) 90° for each of said series of values of φ; , 180°, and 270°, respectively, by repeating the above (a) to (d), respectively, to create a set of linear combinations of three free induction damping signals, (f) creating a set of linear combinations of the four free induction damping signals; (g) adding or subtracting a set of linear combinations of signals to form a set of linear combinations of free induction decay signals; (h) double Fourier transforming the set of linear combinations of free induction damped signals obtained in step (g) to the frequency domain with respect to t1 and t2; A nuclear magnetic resonance measurement method consisting of steps. 4. The nuclear magnetic resonance measurement method according to claim 3, wherein the high-frequency pulse train consists of three 90° pulses, and the interval between two specific 90° pulses is selected to be t1.