Multi-nuclear magnetic resonance study of Na3AlF6-AlPO4 molten and solidified mixtures

Phosphorus is one of the predominant impurities in the Hall-Heroult process for industrial aluminium production. The nature of the dissolved phosphorus species in the Na(3)AlF(6)-AlPO(4) system has been investigated by in situ high-temperature (HT) (19)F, (23)Na, (27)Al, (17)O, and (31)P NMR. The combination of these experiments enables to define the presence of PO(4)(3-), AlF(5)(2-) and (AlF(4)-O-PO(3))(4-) anions in the melt, and then the formation of Al-O-P bonding. Melts solidified at different cooling rates were characterised using various solid-state NMR techniques including multiple quantum magic angle spinning (MQMAS), rotational echo double resonance (REDOR) and heteronuclear single quantum correlation (HSQC). The glass obtained by the rapid quenching of the hypereutectic melt has been carefully described in order to better understand the structure of the melt.     

Analysis of the REDOR Signal and Inversion

An inversion of the REDOR signal to recover the dipolar couplings has been recently proposed [K. T. Muelleret al., Chem. Phys. Lett.242, 535 (1995)]: The corresponding integral transform was performed by tabulation of the kernel followed by numerical integration. After explicit determination of the inverse REDOR kernel by the Mellin transform method, we propose an alternative inversion method based on Fourier transforms. Representation of the inverse REDOR kernel by its asymptotic expansion reveals that the inverse REDOR operator is essentially a weighted sum of a cosine transform and of its derivative. Consequently, known properties of Fourier transforms can easily be transposed to the REDOR inversion, allowing for a precise discussion of the value of the method. Moreover, the first term of the asymptotic expansion leading to a derivative of a cosine transform, the REDOR inversion is found to be extremely sensitive to noise, thus considerably reducing the useful part of the theoretical dipolar window.     

A Comparison between REDOR and θ-REDOR for Measuring C–D Dipolar Interactions in Solids

¹³C-observe REDOR and θ-REDOR experiments for recovering the ¹³C–²D dipolar interaction during MAS NMR are compared. It is found that limited ²D RF power may severely compromise the performance of the REDOR experiment whereas the θ-REDOR experiment can be designed to work well. Results are presented for an isolated ¹³C–²D spin pair with a large deuterium quadrupolar coupling constant and for a ¹³C coupled to three methyl deuterons undergoing fast methyl group rotation.     

Local order in polycarbonate glasses by 13C{19F} rotational‐echo double‐resonance NMR

Nearest‐neighbor chain packing in a homogeneous blend of carbonate ¹³C‐labeled bisphenol A polycarbonate and CF₃‐labeled bisphenol A polycarbonate has been characterized using a shifted‐pulse version of magic‐angle spinning ¹³C{¹⁹F} rotational‐echo double‐resonance (REDOR) NMR. Complementary NMR experiments have also been performed on a polycarbonate homopolymer containing the same ¹³C and ¹⁹F labels. In the blend, the ¹³C observed spin was at high concentration, and the ¹⁹F dephasing or probe spin was at low concentration. In this situation, an analysis in terms of a distribution of isolated heteronuclear pairs of spins is valid. A comparison of the results for the blend and homopolymer defines the NMR conditions under which higher concentrations of probe labels can be used and a simple analysis of the REDOR results is still valid. The nearest neighbors of a CF₃ on one chain generally include a carbonate group on an adjacent chain. A direct interpretation of the REDOR total dephasing for the polycarbonate blend indicates that at least 75% of carbonate‐carbon ¹³C ··· F₃ nearest neighbors are separated by a narrow distribution of distances 4.7 ± 0.3 Å. In addition, analysis of the variations in REDOR spinning‐sideband dephasing shows that most of the ¹³C ··· F₃ dipolar vectors have a preferred orientation relative to the polycarbonate mainchain axis. This combination of distance and orientational constraints is interpreted in terms of local order in the packing of the carbonate group of one polycarbonate chain relative to the isopropylidene moiety in a neighboring chain. © 2006 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 44: 2760–2775, 2006     

Compensation for pulse imperfections in rotational-echo double-resonance NMR by composite pulses and EXORCYCLE

We describe a simple method to compensate for pulse-angle errors in rotational-echo double resonance (REDOR) experiments for determining heteronuclear distances in solids. By using composite 180 degrees pulses on the unobserved dephasing spin and EXORCYCLE for the single pi pulse on the observed channel, the REDOR curve becomes much less sensitive to pulse-angle errors. Both improvements are demonstrated by experiments on the model compound, (15)N, (13)Calpha -labeled N-t-BOC-glycine, and are confirmed by numerical simulations. The advantage of EXORCYCLE is also shown analytically using the product operator formalism. The proposed simple schemes compensate for unavoidable pulse-angle errors that arise, for example, from radiofrequency field inhomogeneity. They also make REDOR experiments more accurate and robust for low-sensitivity samples where direct pulse-length calibration is difficult.     

Natural abundance C REDOR coupled to a singly N-labeled nucleus: simultaneous determination of interatomic distances in crystalline ammonium [N] -glutamate monohydrate

REDOR technique was applied to natural abundance ¹³C nuclei coupled to a singly labeled ¹⁵N nucleus to determine the ¹³C, ¹⁵N interatomic distances simultaneously in crystalline ammonium [¹⁵N] -glutamate monohydrate (1). Consequently, the interatomic C–N distances between ¹⁵N and ¹³C=O, ¹³C_α, ¹³C_β, ¹³C_γ, and ¹³C_δ carbon nuclei for 1 were determined with a precision of ±0.15 Å, after the experimental conditions such as the location of samples in the rotor, length of π pulse etc. were carefully optimized. ¹³C-REDOR factors for three spin system, (ΔS/S₀)_CN1N2, and the sum of two isolated 2-spin system, (ΔS/S₀)^*=(ΔS/S₀)_CN1+(ΔS/S₀)_CN2, were further evaluated by the REDOR measurements on isotopically diluted 1 in a controlled manner. Subsequently, the intra- and intermolecular C–N distances were separated by searching the minima in the contour map of root mean square deviation (RMSD) between the theoretically and experimentally obtained (ΔS/S₀)^* values against two interatomic distances, r_C–N1 and r_C–N2. When the intramolecular C–N distance (r_C–N1) of the particular carbon nucleus is substantially shorter than the intermolecular one (r_C–N2), C–N distances within a single molecule were obtained with an accuracy of ±0.06 Å as in the cases of C=O, C_α and C_β carbon nuclei. C–N distances between the molecule in question and the nearest neighboring molecules can be also obtained, although accuracy was lower. On the contrary, it was difficult to determine the interatomic distances in the same molecule when the intermolecular dipolar contribution is larger than the intramolecular one as in the case of C_δ carbon nucleus.     

Recoupled Polarization-Transfer Methods for Solid-State H–C Heteronuclear Correlation in the Limit of Fast MAS

An in-depth account of the effects of homonuclear couplings and multiple heteronuclear couplings is given for a recently published technique for ¹H–¹³C dipolar correlation in solids under very fast MAS, where the heteronuclear dipolar coupling is recoupled by means of REDOR π-pulse trains. The method bears similarities to well-known solution-state NMR techniques, which form the framework of a heteronuclear multiple-quantum experiment. The so-called recoupled polarization-transfer (REPT) technique is versatile in that rotor-synchronized ¹H–¹³C shift correlation spectra can be recorded. In addition, weak heteronuclear dipolar coupling constants can be extracted by means of spinning sideband analysis in the indirect dimension of the experiment. These sidebands are generated by rotor encoding of the reconversion Hamiltonian. We present generalized variants of the initially described heteronuclear multiple-quantum correlation (HMQC) experiment, which are better suited for certain applications. Using these techniques, measurements on model compounds with ¹³C in natural abundance, as well as simulations, confirm the very weak effect of ¹H–¹H homonuclear couplings on the spectra recorded with spinning frequencies of 25–30 kHz. The effect of remote heteronuclear couplings on the spinning-sideband patterns of CH_n groups is discussed, and ¹³C spectral editing of rigid organic solids is shown to be practicable with these techniques.     

1H‐19F REDOR‐filtered NMR spin diffusion measurements of domain size in heterogeneous polymers

Solid state NMR spectroscopy is inherently sensitive to chemical structure and composition and thus makes an ideal method to probe the heterogeneity of multicomponent polymers. Specifically, NMR spin diffusion experiments can be used to extract reliable information about spatial domain sizes on multiple length scales, provided that magnetization selection of one domain can be achieved. In this paper, we demonstrate the preferential filtering of protons in fluorinated domains during NMR spin diffusion experiments using ¹H‐¹⁹F heteronuclear dipolar dephasing based on rotational echo double resonance (REDOR) MAS NMR techniques. Three pulse sequence variations are demonstrated based on the different nuclei detected: direct ¹H detection, plus both ¹H?¹³C cross polarization and ¹H?¹⁹F cross polarization detection schemes. This ¹H‐¹⁹F REDOR‐filtered spin diffusion method was used to measure fluorinated domain sizes for a complex polymer blend. The efficacy of the REDOR‐based spin filter does not rely on spin relaxation behavior or chemical shift differences and thus is applicable for performing NMR spin diffusion experiments in samples where traditional magnetization filters may prove unsuccessful. This REDOR‐filtered NMR spin diffusion method can also be extended to other samples where a heteronuclear spin pair exists that is unique to the domain of interest.     

Dipolar Coupling Information in Multispin Systems: Application of a Compensated REDOR NMR Approach to Inorganic Phosphates

Anexperimental strategy has been developed for measuring multiple dipole–dipole interactions in inorganic compounds using the technique of rotational echo double resonance (REDOR) NMR. Geometry-independent information about the dipole couplings between the observe nuclear species S (arbitrary quantum number) and the heteronuclear species I (spin- ) can be conveniently obtained from the experimental curve of ΔS/S₀ versus dipolar evolution time by limiting the analysis to the initial data range 0 < ΔS/S₀ < 0.30. Numerical simulations have been carried out on a three-spin system of type SI₂ in order to assess the effect of the I–I homonuclear dipole–dipole coupling and the influence of experimental imperfections such as finite pulse length and misadjustments of the 180° pulses applied to the I-spin species. The simulations show further that within the initial data range the effects of such misadjustments can be internally compensated by a modified sequence having an additional 180° pulse on the I channel in the middle of the dipolar evolution periods. Experimental ²⁷Al{³¹P} REDOR results on the multispin systems Al(PO₃)₃, AlPO₄, [AlPO₄]₁₂(C₃H₇)₄NF, and Na₃PO₄ confirm the general utility of this approach. Thus, for applications to unknown systems the compensation strategy obviates calibration procedures with model compounds.     

Relaxation-induced dipolar exchange with recoupling-An MAS NMR method for determining heteronuclear distances without irradiating the second spin

A new magic-angle spinning NMR method for distance determination between unlike spins, where one of the two spins in question is not irradiated at all, is introduced. Relaxation-induced dipolar exchange with recoupling (RIDER) experiments can be performed with conventional double-resonance equipment and utilize the familiar π-pulse trains to recouple the heteronuclear dipolar interaction under magic-angle spinning conditions. Longitudinal relaxation of the passive spin during a delay between two recoupling periods results in a dephasing of the heteronuclear coherence and consequently a dephasing of the magnetization detected after the second recoupling period. The information about the dipolar coupling is obtained by recording normalized dephasing curves in a fashion similar to the REDOR experiment. At intermediate mixing times, the dephasing curves also depend on the relaxation properties of the passive spin, i.e., on single- and double-quantum longitudinal relaxation times for the case of I = 1 nuclei, and these relaxation times can be estimated with this new method. To a good approximation, the experiment does not depend on possible quadrupolar interactions of the passive spin, which makes RIDER an attractive method when distances to quadrupolar nuclei are to be determined. The new method is demonstrated experimentally with ¹⁴N and ²H as heteronuclei and observation of ¹³C in natural abundance.