Solid-state NMR (SSNMR) spectroscopy is an excellent technique for investigating chemical structure and molecular-level dynamics for a diverse range of systems.1,2 Much of the recent success of SSNMR spectroscopy has resulted from the development and application of acquisition techniques for collecting high-quality NMR spectra of quadrupolar nuclei (i.e., nuclei having nuclear spin quantum numbers S > 1/2), which comprise ca. 73% of all spin- active nuclei on the periodic table.3–5 SSNMR experiments on spin-1/2 nuclei are generally facile to conduct, and their spectra easy to interpret. SSNMR
experiments on quadrupolar nuclei, by comparison, are generally more complex, and their spectra are more challenging to collect and interpret, owing to the anisotropic line-broadening effects of the quadrupolar interaction (QI), which results from the coupling between the nuclear electric quadrupole moment (NQM) and surrounding electric field gradients (EFGs).6,7
Severe anisotropic line broadening often results in the NMR spectra of quadrupolar nuclei that occupy chemical sites having low spherical (i.e., Platonic) symmetry,8,9 which spreads the integrated signal intensity over large frequency regions, reducing both the signal-to-noise ratio (S/N) and the spectral resolution. SSNMR powder patterns of half-integer quadrupolar nuclei (i.e., S = 3/2, 5/2, 7/2, 9/2) affected by large QIs generally span several hundreds of kHz, but spectral breadths spanning many MHz is not uncommon. For this reason, SSNMR spectra of half-integer quadrupolar nuclei normally focus in on the central transition (i.e., CT, m = −1/2 ↔︎ 1/2), which is unaffected by the large first-order
quadrupolar interaction (FOQI). However, a sizeable second-order quadrupolar interaction (SOQI) can severely broaden CT NMR spectra, which are also affected by the other anisotropic NMR interactions (e.g., the chemical shift anisotropy, dipolar couplings). Routine acquisition techniques such as magic- angle spinning (MAS),10 multiple-quantum MAS (MQMAS),11 satellite-transition MAS (STMAS),12 and other hardware-related methods,13,14 which enhance the S/N and/or spectral resolution in NMR spectra of half-integer quadrupolar nuclei, generally work best on nuclides that have small QIs. Therefore, these CT NMR patterns that are affected by large SOQIs are commonly referred to as ultra- widelineNMR (UW NMR) patterns and require specialized acquisition techniques for their collection, some of which are briefly reviewed below.15,16
The CPMG (Carr/Purcell17-Meiboom/Gill18) pulse sequence (Scheme
3.1a) is routinely used to enhance the inherently low S/N that is commonly
encountered in the UW NMR spectra of quadrupolar nuclei.19–24 A series of short, high-power radio-frequency (rf) pulses, which have constant phase and
amplitude, are used to excite and then repeatedly refocus dephasing transverse spin polarization, leading to the formation of multiple spin echoes in the time domain.3 This so-called CPMG echo train can be processed and Fourier transformed “as is” to give a series of discrete and sharp spikelets in the frequency domain, whose outer manifold traces out the total NMR powder
pattern. The CPMG echo train can also be processed by adding the spin echoes
3 The effective T
2 constant (i.e., T2eff), which is the transverse relaxation time constant that is measured when heteronuclear decoupling is used to partially or wholly eliminate the contribution of heteronuclear dipolar coupling to transverse relaxation, and governs how many spin echoes can be collected.
together in the time domain, producing a more traditional-looking NMR spectrum upon Fourier transformation.25,26
Scheme 3.1. Schematic representations of the (a) CPMG, (b) WURST/CPMG (WCPMG), (c) CP/CPMG, and (d) BRAIN-CP/WCPMG (BCP) pulse sequences. See Supporting Information for information regarding the pulse sequence variables and nomenclature (Table A.15).
In many instances, UW NMR powder patterns of half-integer quadrupolar nuclei exceed both the excitation bandwidth of conventional rf pulses and the detection bandwidth of the NMR probe. In these cases, the powder pattern can be mapped out by collecting multiple sub-spectra at different transmitter
frequencies using the variable offset cumulative spectroscopy (VOCS)
method,27,28 which can then be co-added together or skyline-projected to give the total NMR pattern. Frequency-swept pulses are commonly used to excite broad frequency regions, since the excitation bandwidth for these pulses is generally not governed by their pulse width and the applied rf field strength.29 In particular, the Wideband, Uniform-Rate, Smooth-Truncation (WURST) pulse,30–32 which features combined modulation of both its phase and amplitude, is used extensively for collecting high-quality UW NMR spectra for spin-1/2 and quadrupolar nuclei in both diamagnetic and paramagnetic materials.33–40 The
quadratic phase modulation of the WURST pulse gives a corresponding linear sweep of its rf frequency (i.e., νRF(t)), which can be used for the purposes of exciting, refocusing, and transferring spin polarization (vide infra). The WURST pulse can be utilized in the framework of a CPMG-sequence (i.e., the WURST- CPMG (or WCPMG) pulse sequence, Scheme 3.1b),33,34 in which the WURST-B pulse provides broadband excitation of spin polarization and the WURST-C pulse provides broadband refocusing.
Cross polarization (CP) is extensively employed for enhancing weak NMR signals, which involves the transfer of abundant spin polarization from spins with large gyromagnetic ratios (γ) to proximate, dipolar-coupled spins with smaller
values of γ, which are commonly referred to as I and S, respectively.41
Conventional CP experiments (i.e., those using rectangular and monochromatic spin-locking rf pulses with constant phase and amplitude, Scheme 3.1c) often have very narrow excitation bandwidths, which is problematic for efficiently collecting high-quality UW NMR spectra. The BroadbandAdiabaticInversion CrossPolarization (BRAIN-CP) pulse sequence (Scheme 3.1d)42 is a modified CP sequence that uses a WURST pulse as the S-channel spin-locking pulse, which effectively addresses the limited excitation bandwidth. The WURST pulse generates an effective magnetic field (i.e.,Beff(t)), which over the course of the spin-locking period, traverses from +z′ to –z′ (i.e., Beff(t) performs an inversion). In the case of two dipolar-coupled spin-1/2 nuclei (e.g., I and S represent 1H and 13C, respectively), the time-dependent Hartmann-Hahn matching conditions are described by:42
νRF(t)=ΩS
2π ± ΩI
2π+ ν1(1H)2− ν1Amax(S)2 (3.1.1)
This equation demonstrates that up to two Hartmann-Hahn matching conditions can be fulfilled for a large range of S-spin isochromats, depending on the values of the 1H and S spin-locking rf fields used (denoted ν1(1H) and ν1A(S),
respectively), the I- and S-spin resonance offsets (denoted ΩS/2π and ΩI/2π, respectively), and the time-dependent WURST-A rf frequency. Under the right conditions for adiabatic passage, the subsequent S-spin polarization follows
coherence. A broadband conversionpulse then excites the stored S-spin polarization into the transverse plane for detection.
The BRAIN-CP pulse sequence can also be implemented within a CPMG framework as the BRAIN-CP/WCPMG pulse sequence (or BCP for brevity), which can substantially increase the S/N in UW NMR spectra through the combined action of broadband CP and broadband refocusing of spin echoes. This sequence has been previously used to collect high-quality UW NMR spectra for both spin-1/2 and spin-1 nuclei, in variety of organic and inorganic
materials.42–45
Herein, we demonstrate that the BCP pulse sequence can be used to transfer abundant proton spin-polarization to the CT of half-integer quadrupolar nuclides, resulting in broad and uniform excitation of their CT NMR spectra. Four organometallic complexes were investigated (Scheme 3.2), with each featuring a half-integer quadrupolar nuclide having a distinct spin-quantum number. NMR spectra acquired with the BCP pulse sequence are compared with those
acquired using conventional methods, and the technicalities and performance of each technique are evaluated in each case. The potential for the broad
application of these techniques to the routine study of quadrupolar nuclei from across the periodic table is discussed. In addition, detailed discussions are provided of both the optimization procedures for these UW NMR pulse
sequences, as well as general experimental guidelines, in order to encourage researchers to use these techniques to obtain chemical information that is otherwise unattainable using conventional NMR pulse sequences.