J. Brondeau

Self-diffusion measurements using a radiofrequency field gradient

Abstract Self-diffusion coefficient measurements have been carried out by means of a novel and efficient NMR technique, which employs the gradient of an RF field produced by a single-turn coil. The sequence involves also pulses of a homogeneous RF field and has the form δ(g) x −( Δ 2 )−180°(h) y −( Δ 2 )−δ(g) x −90°(h) x -Acq., where h refers to the homogeneous RF field whereas g denotes the application of the RF field gradient. The theory of the classical PGSE (pulsed field gradient spin echo) technique (which makes use of a static field gradient) is reconsidered according to a simplified approach. This shows that similar equations prevail for the sequence using RF field gradient pulses. A probe has been designed in order to obtain a linear RF gradient across a 10 mm o.d. sample. A conventional saddle-shaped coil orthogonal to the single-turn coil produces the homogeneous RF field. It is doubly tuned at proton and deuterium frequencies and therefore allows stabilization of the spectrometer in the course of the experiment. The use of an RF field makes it possible to determine directly the magnitude of the gradient by a capillary whose position is varied. The method has been checked on water, benzene, and acetone samples at room temperature, with self-consistent results, in agreement with literature data. The proposed method permits measurement of self-diffusion coefficients with good accuracy (error less than 5%) in a much simpler way than with pulsed B0 gradients.

Superfast T1 determination by inversion-recovery

The one-shot sequence [π/2(S1)-τ-π-τ-π/2(S2)-τ]n is shown to yield longitudinal relaxation times, T1 with fairly good accuracy, provided that r lies in the interval 0.5T1−3T1 and that any transverse magnetization has disappeared prior to each RF pulse. The inverting pulse is a composite pulse, (π/2)x(4π/3)γ(π/2)x. Signals S1 and S2 are acquired and added in separate memory blocks. T1 is deduced from the formula T1 = -τ/ln(1 − S2/S1). A phase-cycling scheme is devised which eliminates spurious effects due either to residual transverse magnetization or to misadjusted π/2 pulses.

Flexible Fourier multinuclear magnetic resonance spectrometer

This paper presents an easy‐to‐build Fourier multinuclear magnetic resonance spectrometer operating, in the present case, below 100 MHz but readily adaptable to any magnet system. Receiver and transmitter are essentially made of commercially available wide‐band components. Some dedicated homemade circuits, necessary to the proper operation of the system, are described. A probe head has been especially designed for the observation of nuclei resonating between 6 and 36 MHz (nitrogen 14 to phosphorus 31), in addition to proton and fluorine (90 MHz and 84.6 MHz), without removing the sample and/or changing the spectrometer configuration. Interchangeable inserts are built in such a way that the measuring channel on the one hand, and the stabilization–decoupler channels on the other hand, are separated through the use of two orthogonal coils. Tuning and matching networks, which involve, for the measuring channel, homemade nonmagnetic capacitors variable in a large range, are described. The flexibility and the ease of operation of the apparatus are outlined.

Selective saturation as an aid for carbon-13 CP/MAS n.m.r. analysis of coals at high frequencies

Abstract A selective saturation experiment is described and applied to selective elimination of 13 C high resolution solid state n.m.r. signals and of their spinning sidebands, as well for sharp lines as for broad bands. A proper adjustment of experimental conditions permits selective saturation of the aromatic carbon signals of coal samples and eliminates unwanted overlapping of aromatic sidebands with the aliphatic region of the spectrum at high working frequencies. The described procedure seems to be especially useful for qualitative and quantitative high frequency n.m.r. investigations of coals having a distribution of short T 2 relaxation times.

The direct experimental determination of a dipole–dipole cross‐correlation spectral density

A novel method for ascertaining certain dipolar cross‐correlation spectral densities is described in detail. This method is based upon the inherent simplicity associated with the perturbation‐initial‐response characteristics of various multiplet or multispin magnetizations. Furthermore, it is demonstrated that suitable multipulse sampling sequences can be devised which effectively sequester conventional nuclear magnetization and, hence, facilitate the study of multispin magnetization. The outlined methodology is applied to determine a three‐spin cross‐correlation spectral density for an adamantyl methylene spin grouping.

Polarization transfer via J coupling by semiselective excitation applied to the high gyromagnetic ratio nucleus

Abstract A novel sequence is proposed for selectively transferring polarization from a given nucleus A to a nucleus X which is J coupled to A. It consists of applying at the A resonance a semiselective pulse of amplitude γ A H 2 = J 2 and of duration t such that γ A H 2 t = π 2 1 2 . This pulse is followed by a nonselective π 2 pulse on A, phase shifted by 90°. A nonselective π 2 pulse on X allows observation of the enhanced X resonances. In practice, instead of the semiselective pulse of low amplitude and long duration, a train of small flip angle pulses (the DANTE sequence of Morris and Freeman) is employed. It is devised to provide the maximum polarization transfer and calculations are given which allow determination of its optimum characteristics. As for other polarization transfer schemes, decoupling of A immediately after the observation pulse is inadequate. The refocusing sequence of Burum and Ernst {( π 2 )(X) − τ − [π(X), π(A)] − τ − Acquisition} was used, which leads to enhanced decoupled spectra. This sequence is further discussed and a procedure is presented for eliminating out-of-phase or unwanted X signals. The proposed polarization transfer sequence is selective regarding both chemical shifts and J coupling constants. This was quantitatively evaluated by means of a complete density matrix calculation. The efficiency and selectivity of the method are thus well characterized: chemical shift selectivity is of the order of 0.15 J if one is able to detect a 10% intensity variation. By taking advantage of this selectivity, correlations can be obtained between A and X chemical shifts. A tripeptide (glutathione) was chosen for providing an example of such a correlation established between proton and carbon-13 chemical shifts.

Analysis of time-domain NMR data by standard non-linear least-squares

Abstract Standard non-linear least-squares are used for deriving spectral characteristics (frequency, amplitude, damping factor and phase) from time-domain NMR data (FID). The method requires initial estimates and involves an iterative process. With normal equations properly recast, the method is seen to be of general applicability within a model of damped sinusoids. The processing time is quite reasonable. Various tests demonstrate the validity of this procedure for quantitative determination purposes, especially for weak signals in the presence of intense resonances or for overlapping resonances with different linewidths.

Analysis of carbon-13 longitudinal relaxation times and nuclear overhauser enhancements in alkyl chains

Abstract A simple procedure is proposed which yields, from carbon-13 longitudinal relaxation times and nuclear Overhauser enhancements, information concerning the local mobility of a given CH vector in an alkyi chain. This method is based on the definition of an effective rotation axis related to internal motions and makes use of Woessners equations. It yields a correlation time τ R characteristic of the overall reorientation. In addition, for each carbon C i a correlation time τ G i , associated with the fast motion of the C i H vector, and an angle θ i , between C i H and an effective rotation axis, are determined. The effect of cross-correlation spectral densities is discussed. This model is used for analysing relaxation data of decylammonium chloride micelles. It is further checked by interpreting experimental results at two different observation frequencies.

Selective pulses applied to multiplets in homonuclear-coupled NMR Spectra; Determination of mutual couplings

Abstract It is shown experimentally and theoretically, using the density matrix formalism, that a selective pulse applied to a given multiplet of a homonuclear-coupled spin system may yield intensity alterations in other multiplets, the NMR spectrum being obtained from the conventional Fourier transform technique used immediately after the selective pulse. These effects occur if J coupling exists between the multiplets involved and if the rf amplitude used for the selective pulse is of the same order of magnitude as the coupling constant. Some examples are given; possible extensions of this method are presented.

A simple and efficient water-suppression scheme based on B1 inhomogeneity

Numerous techniques are currently available for suppressing the strong solvent (generally water) peak in NMR spectra ( 1). Such suppression is mandatory for ensuring proper digitization of free induction decays, meaning that a large suppression factor must be obtained (of the order of several hundreds). Moreover, for many applications, notably for two-dimensional spectroscopy, it is desirable to avoid phase distortions possibly produced by the suppression scheme itself (2-4). Most methods employing hard pulses, which have proven to be efficient as far as suppression is concerned, are known to entail such phase distortions. These methods were in fact developed from the original idea of Plateau and G&on (5)) who devised the socalled jump-return (or l-l) sequence which only includes two 90” pulses. It is the only member of this family of pulse sequences which does not produce any phase distortion. Recently Sklenar and Bax devised a “ 11 echo” scheme which features both suppression efficiency and absence of phase distortions. Unfortunately, this is accompanied by a sin 3 frequency dependence, which leads to a strong signal attenuation in a relatively broad region around the resonance to be suppressed. The other class of water-suppression schemes involves soft pulses. The simplest one, which consists in presaturating the water resonance (6), must however be adapted ( 7) when the experiment considered includes an evolution period (during which water magnetization may recover). Again, Sklenar and Bax (2) proposed a combination of soft and hard pulses which leads to an experiment free of phase distortions with good water suppression. As mentioned by the authors, their method is hardware-demanding and may require careful adjustments. As an alternative to these procedures, we present a scheme based on a property which has been scarcely exploited (3, 8)) namely the inherent inhomogeneity of the radiofrequency field B, . The basic idea is very simple and has already been used in the context of experiments employing surface coils (9). Consider the application of an RF field coinciding with the y axis of the rotating frame. All x and z magnetization components will precess around y at a frequency which will depend on the location of the relevant molecule in the sample, since B, is more or less inhomogeneous (more in the case of saddle-shaped coils used in most probes of superconducting magnets). If applied for a time sufficiently long, B, will produce a complete scatter of magnetization in the x, z plane resulting from the above-mentioned difference in precession