JP2970704B2 - Multidimensional nuclear magnetic resonance measurement method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to multidimensional nuclear magnetic resonance (NM).
More particularly, the present invention relates to a multidimensional NMR measurement method capable of easily performing phase correction for obtaining a pure absorption mode.

[0002]

2. Description of the Related Art In a multidimensional NMR, for example, a two-dimensional NMR method, the NMR signal is displayed as a two-dimensional spectrum, so that the resolution is improved and the analysis of the spectrum is facilitated as compared with the conventional method. It has excellent features such as the ability to elucidate the interaction of

FIG. 2 shows an example of an apparatus for performing such a two-dimensional NMR method. In the figure, a sample coil 2 is arranged in a static magnetic field generated by a magnet 1, and a measurement sample is inserted into a space inside the sample coil 2. High frequency oscillator 3
The high-frequency signal having the resonance frequency of the observed nucleus generated from
After a predetermined phase is given by a variable phase shift circuit 4 capable of selecting an arbitrary phase from 0 ° to 360 °, the phase is supplied to the coil 2 as a high-frequency pulse via an amplifier 5 and a gate 6 and irradiated on a sample. . The resonance signal induced in the coil 2 after the irradiation of the high-frequency pulse is sent to the demodulation circuits 9 and 10 via the gate 7 and the receiving circuit 8. The high-frequency signals from the high-frequency oscillators are sent to the demodulation circuits 9 and 10 as reference signals, and one of them is sent through the 90 ° phase shift circuit 11, so that the two demodulation circuits are 90 °
The two-channel detection systems CHa and CHb having different phases are configured. The free induction decay signal obtained from the two-channel detection system is converted into a digital signal by the A / D converters 12 and 13, sent to the computer 14, and stored in the attached memory 15. 16 is a phase shift circuit 4,
A pulse programmer that controls the gates 6, 7 and the A / D converters 12 and 13, the order of the pulse train applied to the sample, the pulse width, the phase of the high frequency included in each pulse, and the sampling by the A / D converters 12 and 13. Are programmed in advance, and a series of measurements are performed according to the program.

[0004] Such conventional two-dimensional NMR measurement is performed, for example, at 90 ° xt ₁ -90 ° x as shown in FIG.
It will be described with reference to the pulse sequence -t _2.

As shown in FIG. 3A, a general measurement process in the two-dimensional NMR method is a preparation period before the first 90 ° pulse, an expansion period (t ₁ ), and a detection period (t ₂ ).
Of three time domains. The preparation period is necessary to keep the magnetization of the nucleus in a proper initial state, and the preparation pulse (the first 90 ° pulse) brings the magnetization into a non-equilibrium state,
This condition is developed in the development period t _1, the behavior of magnetization in the t ₁ is the detection pulse (second 90 ° pulse) the free induction decay signals detected in the detection period t ₂ after the application (FID signal) Handed over as phase and amplitude information. Therefore, during the FID signal detected during the t ₂ as well as the behavior of magnetization in t _2, also will include information of the magnetization behavior in t _1. Then, t
For example, the number of FID signals (FIDa1 to FIDam and FIDb1 to FIDb1 to FIDb1 to m1) obtained from the two detection systems in the measurement at each stage is changed stepwise by using ₁ as a variable.
FIDbm), set data S ₁ (t ₁ , t ₂ ),
S ₂ (t ₁ , t ₂ ) is added, and the added data is subjected to double Fourier transform for t ₂ and t ₁ , thereby obtaining 2
Dimensional spectrum is obtained.

When the two-dimensional NMR spectrum data obtained in this manner is represented on a plane, diagonal peaks and crossing peaks appear symmetrically, and both the diagonal and crossing peaks have absorption and dispersion waveforms. It has a mixed shape, making the spectrum difficult to analyze.

[0007] As a countermeasure, the present applicant has proposed to perform phase correction on a two-dimensional spectrum to obtain an absorption spectrum.

[0008] This method will be described briefly.
The pulse sequence 90 ° x-t ₁ -90 ° x- in FIG.
At t ₂ , the set data S ₁ (t ₁ , t ₂ ) and S ₂ (t ₁ , t ₂ ) are obtained by the two detection systems as described above, and further, the set data is obtained by the pulse sequence shown in FIG. S ₁ ´ (
t ₁ , t ₂ ) and S ₂ ′ (t ₁ , t ₂ ). And
_{S 1, S 2, S 1} ', S 2' performs a complex Fourier transform on t ₂ respectively, for the data binding S after the Fourier transform _{(t 1, F 2) =} S 1 (t 1, F 2) + iS
_{2 (t 1, F 2)} , S'(t 1, F 2) = S 1 '(
Assuming that t ₁ , F ₂ ) + iS ₂ ′ (t ₁ , F ₂ ), S (t ₁ , F ₂ ) = A ｛−a ₂ (ω ₂ ) cos φ ₂ + d ₂ (ω ₂ ) sin φ ₂ ｝ + i A ｛ a ₂ (ω ₂ ) sin φ ₂ + d ₂ (ω ₂ ) cos φ ₂ … (1) S ′ (t ₁ , F ₂ ) = A ′ ｛− a ₂ (ω ₂ ) cos φ ₂ + d ₂ (ω ₂ ) sinφ ₂ ＋+ iA ′ ｛a ₂ (ω ₂ ) sin φ ₂ + d ₂ (ω ₂ ) cos φ ₂ … (2) where A = M ₀ exp ｛i (ω ₁ t ₁ + φ ₁ )｝ exp ^-t ₁ ^{/ T A} ′ = M ₀ exp ｛i (ω ₁ t ₁ + 90 ° + φ ₁ )｝ exp ^{-t1 / T} (M ₀ is magnetization at thermal equilibrium) a ₂ (ω ₂ ) = T / ｛1 + T ² (ω ₂ + Ω) ² ｝ d ₂ (ω ₂ ) = T ² (ω ₂ + ω) / ｛1 + T ² (ω ₂ + ω) ² ωω ₁ , ω ₂ : Resonance frequency φ ₁ about magnetization t ₁ axis and t ₂ axis, φ ₂ : phase of magnetization with respect to t ₁ axis and t ₂ axis T: lateral relaxation time The phase correction is performed by S expressed by the equations (1) and (2).
Φ ₂ = 0 for (t ₁ , F ₂ ) and S ′ (t ₁ , F ₂ )
The correction is performed so that For example, when the real part of the combined data S (t ₁ , F ₂ ) and S ′ (t ₁ , F ₂ ) is R and the imaginary part is I, −R cos φ + I sin φ
Seeking value, this time phi is equal to _{φ 2, -R cosφ + I sinφ} = a 2 (ω 2) cos 2 φ 2 -d 2 (ω 2) sinφ 2 cosφ 2 + a 2 (ω 2) sin 2 φ _{2 + d 2 (ω 2)} sinφ 2 cosφ 2 = a 2 (ω 2) and turned with the maximum value. Therefore, change φ variously-
If the value of R cos φ + I sin φ is set to be the maximum, the value is a phase-corrected real part.
Is equal to φ ₂ . Further, when obtaining the R sinφ ₂ + I cosφ ₂ by using this phi _2, the imaginary part of the phase-corrected becomes _{R sinφ 2 + I cosφ 2 =} d 2 (ω 2) is obtained. As a result of the phase correction (assuming that φ ₂ = 0), S (t ₁ , F ₂ ) = A {a ₂ (ω ₂ )} + i {Ad ₂ (ω ₂ )} (3) S ′ (t ₁ ) , F ₂ ) = A ′ ｛a ₂ (ω ₂ )｝ + iA ′ ｛d ₂ (ω ₂ )｝ (4) Next, only the real part data of the combined data S (t ₁ , F ₂ ) and S ′ (t ₁ , F ₂ ) expressed by the equations (3) and (4) are taken out and Fourier-transformed, and S (F ₁ ,
F ₂ ) and S ′ (F ₁ , F ₂ ) are obtained and combined, and φ
Performing phase correction seeking phi ₁ in the same manner as for _2.

[0009]

However, in such a conventional phase correction method, after the Fourier transform, φ ₁ ,
φ ₂ is obtained, the operation of phase correction is extremely complicated, and the real part and imaginary part data of both axes of t ₁ and t ₂ must be stored, so a large data capacity is required. In addition, there is a problem that it takes time for calculation, and when the capacity is small, the resolution is deteriorated.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and can obtain a phase correction value by a simple operation, reduce the data capacity, increase the calculation speed, and improve the resolution. An object of the present invention is to provide a three-dimensional nuclear magnetic resonance measurement method.

[0011]

For this purpose, the multidimensional nuclear magnetic resonance measuring method according to the present invention comprises the steps of (a) developing periods t ₁ , t ₂ .
tm _-1 and a detection period t _m (m = 2, 3,...), and the second pulse or pulse train is irradiated after the first pulse or pulse train development period t ₁ has elapsed. , The second pulse or the pulse train is irradiated with the third pulse or the pulse train after the irradiation period t ₂ , and thereafter, the irradiation of the pulse or the pulse train and the developing period are similarly repeated, and the m-th pulse or after irradiation a pulse train, the detection period t _m free induction decay signals from the sample using a sequence to be detected by the two detection systems having different phase by 90 ° across, a plurality of different t
Collective data S ₁ measured for the values of ₁ , t ₂ ... T _{m- 1
(T 1, t 2, ...} t m), S 2 (t 1, t 2, ... t m)
(B) S (t ₁ , t ₂ ,... T _m ) = S
₁ (t ₁ , t ₂ ,... T _m ) + iS ₂ (t ₁ , t ₂ ,.
than _{m), t 1, t 2} ...... when t all values on the time axis but one time axis of the _m is approximately "0", the data S
And performing a one-dimensional Fourier transform as a one-dimensional free induction decay signal for the one time axis to obtain a phase correction value for the time axis. (C) A phase correction value for each time axis according to (b) (D) Performing a multidimensional Fourier transform to obtain S (F ₁ , F ₂ ,... F _m ), and (e) using the phase correction values obtained in (b), (c) It is characterized in that axis phase correction is performed.

[0012]

According to the multidimensional nuclear magnetic resonance measurement method of the present invention, in order to obtain a phase correction value of a certain axis, data obtained when all other values on the time axis are substantially "0" are taken out and a one-dimensional FID is obtained. Data is obtained, one-dimensional Fourier transform is performed on the data, and a phase correction is performed so that an absorption waveform is obtained in a one-dimensional spectrum obtained by Fourier transform to obtain a correction value. Such processing is performed for each time axis. Since the one-dimensional FID data has a small number of data, a phase correction value can be easily obtained, and a phase correction value can be obtained before multidimensional Fourier transform. As a result, the data of the imaginary part is discarded and the absorption spectrum is discarded. Can be obtained, the number of data can be halved, and the resolution can be increased.

[0013]

Embodiments will be described below with reference to the drawings.

An example of two-dimensional NMR using a pulse sequence of 90 ° -t ₁ -90 ° -t ₂ will be described.

FIG. 1 is a flowchart for explaining the phase correction method of the present invention.

As described with reference to FIG. 2, two-dimensional set data S ₁ (t ₁ , t ₂ ) and S
Assuming that ₂ (t ₁ , t ₂ ) is obtained, S (t ₁ , t ₂ ) = S ₁ (t ₁ , t ₂ ) + iS ₂ (t ₁ , t ₂ ) = M ₀ exp ｛i (ω ₁ t ₁ + φ ₁ )｝ exp (-t ₁ / T) × exp
{I (ω ₂ t ₂ + φ ₂ )} exp (−t ₂ / T). Here, M ₀ is the magnetization at the time of thermal equilibrium, ω ₁ and ω ₂ are the resonance frequencies of the magnetization on the t ₁ axis and the t ₂ axis, φ ₁ and φ ₂ are the phases of the magnetization on the t ₁ axis and the t ₂ axis, and T is This is the lateral relaxation time. In this equation, t ₁ ≒ 0
The data at the time of is extracted from the two-dimensional data and is set as a one-dimensional FID as in the following equation.

S (t ₂ ) = M ₀ exp ｛i (ω ₂ t ₂ + φ ₂ )｝ exp (−t ₂ / T) (5) This is one-dimensional data on the time axis t ₂ . Since the number of data is extremely small compared to two-dimensional data, one-dimensional Fourier transform is performed so that an absorption waveform can be obtained.
By performing phase correction with AND error, φ ₂
You can know. When S (t ₂ ) is subjected to one-dimensional Fourier transform, S (F ₂ ) = A ｛−a ₂ (ω ₂ ) cos φ ₂ + d ₂ (ω ₂ ) sin φ ₂ ｝ + iA ｛a ₂ (ω ₂ ) sin φ ₂ + D ₂ (ω ₂ ) cos φ ₂ 、, and, as described in the equations (1) and (2), −
By setting φ so that the value of R cos φ + I sin φ becomes maximum, it is also possible to obtain a phase correction value.

Next, data at the time of t ₂取 り 出 し 0 is extracted from the two-dimensional data, and is set as a one-dimensional FID as in the following equation.

S (t ₁ ) = S ₀ exp ｛i (ω ₁ t ₁ + φ ₁ )｝ exp (−t ₁ / T) (6) As in the case of the time axis t ₂ , φ ₁ Can be obtained.

Thus, the phase correction values φ ₁ and φ ₂ can be accurately known before performing the Fourier transform. If the phase correction value can be known, the absorption spectrum can be obtained by using only the real part data, as can be seen from the above-described equations (3) and (4). Therefore, when performing the two-dimensional Fourier transform, the imaginary part May be discarded as zero data. As a result, the number of data can be halved, so that the calculation speed can be increased, and the resolution can be improved when the data storage capacity is the same.

The magnetization in the multidimensional NMR is S (t ₁ , t ₂ ,..., T _m ) = M ₀ exp {i (ω ₁ t ₁ + φ ₁ )} exp (−t ₁ / T) ×. _{... × exp {i (ω m} t m + φ m)} it is possible to express the exp (-t _m / T), in order to know the phi ₁ is t _2, ... t _m ≒
The one-dimensional FID at the time of 0 may be taken out as one-dimensional data and one-dimensional Fourier transform may be performed. Similarly, for each time axis, the data when the time other than the time axis is almost zero is multidimensional. A phase correction value can be easily obtained by taking out one-dimensional data from the data and performing one-dimensional Fourier transform.

[0022]

As described above, according to the present invention, a phase correction value can be obtained by extracting data when t ≒ 0 from an axis other than the axis for which phase correction is performed, thereby simplifying the operation. In addition, since the phase correction value can be obtained before the Fourier transform, the absorption spectrum can be obtained by discarding the imaginary part data, and the calculation speed is increased by treating the imaginary part as 0 data, and the same data storage capacity is obtained. In this case, the resolution can be improved.

[Brief description of the drawings]

FIG. 1 is a diagram showing a flowchart of multidimensional nuclear magnetic resonance measurement.

FIG. 2 is a diagram showing a configuration of a two-dimensional nuclear magnetic resonance measurement apparatus.

FIG. 3 is a diagram showing a pulse sequence.

[Explanation of symbols]

REFERENCE SIGNS LIST 1 magnet, 2 sample coil, 3 high-frequency oscillator, 4 variable phase shift circuit, 6, 7 gate, 8 receiving circuit, 9, 10
... demodulation circuit, 11 ... 90 degree phase shift circuit, 12, 13 ... A /
D converter, 14 ... computer, 15 ... memory.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) G01N 24/00-24/12

Claims (1)

(57) [Claims] (A) An expansion period t ₁ , t ₂ ... T _m-1 , a detection period t _m (m = 2, 3,...), And expansion after irradiation of the first pulse or pulse train Irradiate a second pulse or pulse train after a period t ₁ , irradiate a third pulse or pulse train after a second pulse or pulse train irradiation period t _2, and similarly apply a pulse, or repeated irradiation with the evolution period of the pulse train, after the irradiation pulses of the m, or a pulse train, with a sequence of detecting a free induction decay signal the two detection systems having different phase by 90 ° of the sample over the detection period t _m, A plurality of different t ₁ , t ₂ ... T
set was measured for _m-1 value data S ₁ (t _1,
t ₂ ,... t _m ), S ₂ (t ₁ , t ₂ ,..., t _m ); (b) S (t ₁ , t ₂ ,..., t _m ) = S ₁ (t ₁ , t ₂ ,
_{... t m) + iS 2 (} t 1, t 2, ... than t _m), t _1,
When the values on all the time axes other than one of the time axes of t ₂ ... t _m are substantially “0”, the data S is extracted, and 1 is set as a one-dimensional free induction decay signal for the one time axis. Performing a four-dimensional Fourier transform to obtain a phase correction value for the time axis; (c) obtaining a phase correction value for each time axis according to (b); (d) performing a multidimensional Fourier transform to obtain S (F ₁ , F ₂ , ...
F _m ); and (e) performing phase correction on each time axis using the phase correction values obtained in (b), (c), and (c).