High-Resolution 19F MAS NMR Spectroscopy: Structural Disorder and ...
High-Resolution 19F MAS NMR Spectroscopy: Structural Disorder and Unusual J Couplings in a Fluorinated Hydroxy-Silicate
John M. Griffin, Jonathan R. Yates, Andrew J. Berry, Stephen Wimperis and Sharon E. Ashbrook*
Supporting Information
1. Solid-state NMR experimental 2. 19F DFT calculations 3. DFT calculations for clinohumite 4. Fast-spinning 19F MAS NMR 5. Full reference 37 from main text 6. References
S1
1. Solid-state NMR experimental
General Solid-state NMR experiments were performed on a Bruker Avance III spectrometer operating at a magnetic field strength, B0, of 14.1 T, corresponding to a
19
F Larmor
frequency ω0/2π = 564.7 MHz. Experiments were performed using either Bruker 2.5-mm or 1.3-mm double resonance H/F X probes, with MAS rates (νR) of 30 kHz or 60 kHz, respectively. Chemical shifts are given relative to CCl3F, referenced using a secondary reference of polytetrafluoroethylene (PTFE) (δ = –122.7 ppm). The spinning angle was accurately adjusted using the using the OD resonance in the 2H MAS NMR spectrum of fully deuterated malonic acid (with an estimated accuracy of better than 0.01°). The 19F longitudinal T1 relaxation times are ~30 s, although there is no significant change in the relative spectral intensities at shorter recycle intervals, and typically a recycle interval of 8 s was used.
Through-space correlation experiments Two-dimensional 19F through-space dipolar correlation experiments were carried S1 out using the pulse sequence shown in Figure S1a, with νR = 30 kHz. One cycle of BABA
recouplingS2 was used for DQ excitation and reconversion, in order to probe only the shortest F-F contacts, between 2 and 3.5 Å). (Experiments were also performed with 2 cycles but showed little difference in appearance). A 16-step phase cycle was used to select a coherence change of ±2 (4 steps) for the DQ excitation pulses, and a coherence change of –1 (4 steps) for the z-filter 90° pulse. The States-TPPI methodS3 was used to obtain sign discrimination in the F1 dimension. For each of the 256 t1 steps, 32 transients were coadded with a recycle interval of 8 s. A rotor-synchronised t1 increment of 33.33 µs was used.
Through-bond correlation experiments Two-dimensional 19F through-bond correlation experiments were performed using the refocussed-INADEQUATE pulse sequence shown in Figure S1b,S4 with νR = 30 kHz and a rotor-synchronised J evolution period τ of 20 ms. A 16-step phase cycle was used to S2
select a coherence change of ±2 (4 steps) over the three pulses preceding t1, and ±2 (4 steps) for the first 180° pulse. The States methodS5 was used to obtain sign discrimination in the F1 dimension. For each of the 256 t1 steps, 16 transients were coadded with a recycle interval of 8 s. A presaturation train of 32 pulses was used to reduce the overall acquisition time of the experiment. A rotor-synchronised t1 increment of 33.33 µs was used.
J-resolved experiments 19
F J-resolved experiments were recorded using the pulse sequence shown in Figure
S1c,S6 with νR = 60 kHz. A z-filter was used prior to acquisition to ensure pure absorptionmode lineshapes. The States methodS5 was used to ensure sign discrimination in the F1 dimension. A 32-step phase cycle was used to select coherence changes of ±1 (2 steps) for the first 90° pulse, ±2 (4 steps) for the 180° pulse, and –1 (4 steps) for the final 90° pulse. In each experiment, rotor-synchronised t1/2 increments of 4 ms were used. For each of the 64 t1 steps 32 transients were coadded with a recycle interval of 8 s. The spinning angle was accurately adjusted to ensure averaging of any residual 19F-19F dipolar couplings between S7
adjacent sites, which may otherwise affect the observed modulation.
This is
demonstrated in Figure S2, where a small deviation from the magic angle of approximately ±0.04° leads to a change of up to ~4 Hz in the observed splitting of the FB resonance in F0.5(OD)0.5-clinohumite.
S3
Figure S1. Pulse sequences and coherence transfer pathway diagrams for (a) a throughspace DQ MAS NMR experiment with one cycle of BABA recoupling, (b) a through-bond DQ MAS refocused-INADEQUATE experiment and (c) a z-filtered J-resolved experiment.
S4
Figure S2. (a) Cross-sections, extracted parallel to F1, for the FB resonance from a twodimensional J-resolved spectrum of F0.5(OD)0.5-clinohumite and (b) 2H MAS NMR spectra of d4-malonic acid, as the spinning angle, θR, is adjusted. In each case, νR = 30 kHz. When the spinning angle is accurately adjusted to 54.74° the splitting observed results from the presence of scalar couplings only.
S5
2. 19F DFT calculations
Calculations of NMR shielding parameters were carried out using the CASTEP DFT code,S8 employing the gauge including projector augmented wave (GIPAW) algorithm,S9 that allows the reconstruction of the all-electron wave function in the presence of a magnetic field. The generalised gradient approximation (GGA) PBE functionalS10 was employed and core-valence interactions were described by ultrasoft pseudopotentials.S11 A planewave energy cut-off of 50 Ry (680 eV) was used and integrals over the Brillouin zone were performed using a k-point spacing of 0.05 Å–1. Prior to the calculation of NMR parameters, full geometry optimisations were performed for all structures (also using a cut-off energy of 50 Ry and k-point spacing of 0.05 Å–1), with variation of both the lattice parameters and internal atomic coordinates. Calculations were performed on the EaStCHEM Research Computing Facility, which consists of 136 AMD Opteron processing cores, partly connected by Infinipath high speed interconnects. Calculation wallclock times ranged from 9 to 75 hours using 4 cores, depending on the size of the model unit cell being calculated.
Calculations generate the absolute shielding tensor (σ σ) in the crystal frame. The isotropic chemical shift, δiso, is related to σ by
δiso = –(σiso – σref) / (1 – σref) • –(σiso – σref) ,
(S1)
where σiso, the isotropic shielding, is (1/3) Tr{σ σ}, and σref (assumed to be
John M. Griffin, Jonathan R. Yates, Andrew J. Berry, Stephen Wimperis and Sharon E. Ashbrook*
Supporting Information
1. Solid-state NMR experimental 2. 19F DFT calculations 3. DFT calculations for clinohumite 4. Fast-spinning 19F MAS NMR 5. Full reference 37 from main text 6. References
S1
1. Solid-state NMR experimental
General Solid-state NMR experiments were performed on a Bruker Avance III spectrometer operating at a magnetic field strength, B0, of 14.1 T, corresponding to a
19
F Larmor
frequency ω0/2π = 564.7 MHz. Experiments were performed using either Bruker 2.5-mm or 1.3-mm double resonance H/F X probes, with MAS rates (νR) of 30 kHz or 60 kHz, respectively. Chemical shifts are given relative to CCl3F, referenced using a secondary reference of polytetrafluoroethylene (PTFE) (δ = –122.7 ppm). The spinning angle was accurately adjusted using the using the OD resonance in the 2H MAS NMR spectrum of fully deuterated malonic acid (with an estimated accuracy of better than 0.01°). The 19F longitudinal T1 relaxation times are ~30 s, although there is no significant change in the relative spectral intensities at shorter recycle intervals, and typically a recycle interval of 8 s was used.
Through-space correlation experiments Two-dimensional 19F through-space dipolar correlation experiments were carried S1 out using the pulse sequence shown in Figure S1a, with νR = 30 kHz. One cycle of BABA
recouplingS2 was used for DQ excitation and reconversion, in order to probe only the shortest F-F contacts, between 2 and 3.5 Å). (Experiments were also performed with 2 cycles but showed little difference in appearance). A 16-step phase cycle was used to select a coherence change of ±2 (4 steps) for the DQ excitation pulses, and a coherence change of –1 (4 steps) for the z-filter 90° pulse. The States-TPPI methodS3 was used to obtain sign discrimination in the F1 dimension. For each of the 256 t1 steps, 32 transients were coadded with a recycle interval of 8 s. A rotor-synchronised t1 increment of 33.33 µs was used.
Through-bond correlation experiments Two-dimensional 19F through-bond correlation experiments were performed using the refocussed-INADEQUATE pulse sequence shown in Figure S1b,S4 with νR = 30 kHz and a rotor-synchronised J evolution period τ of 20 ms. A 16-step phase cycle was used to S2
select a coherence change of ±2 (4 steps) over the three pulses preceding t1, and ±2 (4 steps) for the first 180° pulse. The States methodS5 was used to obtain sign discrimination in the F1 dimension. For each of the 256 t1 steps, 16 transients were coadded with a recycle interval of 8 s. A presaturation train of 32 pulses was used to reduce the overall acquisition time of the experiment. A rotor-synchronised t1 increment of 33.33 µs was used.
J-resolved experiments 19
F J-resolved experiments were recorded using the pulse sequence shown in Figure
S1c,S6 with νR = 60 kHz. A z-filter was used prior to acquisition to ensure pure absorptionmode lineshapes. The States methodS5 was used to ensure sign discrimination in the F1 dimension. A 32-step phase cycle was used to select coherence changes of ±1 (2 steps) for the first 90° pulse, ±2 (4 steps) for the 180° pulse, and –1 (4 steps) for the final 90° pulse. In each experiment, rotor-synchronised t1/2 increments of 4 ms were used. For each of the 64 t1 steps 32 transients were coadded with a recycle interval of 8 s. The spinning angle was accurately adjusted to ensure averaging of any residual 19F-19F dipolar couplings between S7
adjacent sites, which may otherwise affect the observed modulation.
This is
demonstrated in Figure S2, where a small deviation from the magic angle of approximately ±0.04° leads to a change of up to ~4 Hz in the observed splitting of the FB resonance in F0.5(OD)0.5-clinohumite.
S3
Figure S1. Pulse sequences and coherence transfer pathway diagrams for (a) a throughspace DQ MAS NMR experiment with one cycle of BABA recoupling, (b) a through-bond DQ MAS refocused-INADEQUATE experiment and (c) a z-filtered J-resolved experiment.
S4
Figure S2. (a) Cross-sections, extracted parallel to F1, for the FB resonance from a twodimensional J-resolved spectrum of F0.5(OD)0.5-clinohumite and (b) 2H MAS NMR spectra of d4-malonic acid, as the spinning angle, θR, is adjusted. In each case, νR = 30 kHz. When the spinning angle is accurately adjusted to 54.74° the splitting observed results from the presence of scalar couplings only.
S5
2. 19F DFT calculations
Calculations of NMR shielding parameters were carried out using the CASTEP DFT code,S8 employing the gauge including projector augmented wave (GIPAW) algorithm,S9 that allows the reconstruction of the all-electron wave function in the presence of a magnetic field. The generalised gradient approximation (GGA) PBE functionalS10 was employed and core-valence interactions were described by ultrasoft pseudopotentials.S11 A planewave energy cut-off of 50 Ry (680 eV) was used and integrals over the Brillouin zone were performed using a k-point spacing of 0.05 Å–1. Prior to the calculation of NMR parameters, full geometry optimisations were performed for all structures (also using a cut-off energy of 50 Ry and k-point spacing of 0.05 Å–1), with variation of both the lattice parameters and internal atomic coordinates. Calculations were performed on the EaStCHEM Research Computing Facility, which consists of 136 AMD Opteron processing cores, partly connected by Infinipath high speed interconnects. Calculation wallclock times ranged from 9 to 75 hours using 4 cores, depending on the size of the model unit cell being calculated.
Calculations generate the absolute shielding tensor (σ σ) in the crystal frame. The isotropic chemical shift, δiso, is related to σ by
δiso = –(σiso – σref) / (1 – σref) • –(σiso – σref) ,
(S1)
where σiso, the isotropic shielding, is (1/3) Tr{σ σ}, and σref (assumed to be
