Compendium

NMR Spectroscopy Deep Dive

23 min read

NMR Spectroscopy Deep Dive

A Tier 2 deep-dive into nuclear magnetic resonance — the dominant analytical tool for organic structure elucidation, protein structure determination, dynamics, and materials characterization. Covers one-dimensional pulse experiments (¹H, ¹³C, DEPT, APT), the homonuclear and heteronuclear 2D toolkit (COSY, TOCSY, NOESY, ROESY, EXSY, HSQC, HMBC, INADEQUATE, J-resolved), beyond-¹H/¹³C nuclei (¹⁵N, ¹⁹F, ³¹P, ¹¹B, ²⁹Si, metal NMR), diffusion experiments (DOSY), solid-state methods (MAS, CP, DNP), protein NMR 3D-4D experiments (HNCA, HNCO, TROSY), structure determination workflows (NOE distances, RDCs, CSA, dihedrals), dynamics (T1, T2, NOE, Lipari-Szabo, CPMG relaxation dispersion, CEST), hardware (cryoprobes, 1.2 GHz, JEOL, Spinsolve), databases (NMRShiftDB, BMRB), and processing software (Mestrenova, TopSpin, NMRPipe, Sparky, CCPN Analysis, mnova). Modern NMR is a tightly choreographed pulse-sequence + RF + gradient + magnetic-field-shaping discipline; this note maps the territory rather than teaching pulse mechanics from scratch.

See also

Fundamentals (recap from analytical-chemistry-methods)

Nuclei with non-zero spin (I ≠ 0) precess in a static magnetic field B₀ at the Larmor frequency ω₀ = γB₀, where γ is the gyromagnetic ratio (in MHz/T). Reference table for key NMR-active nuclei:

NucleusINatural abundanceγ (MHz/T)Receptivity (vs ¹H)
¹H1/299.985%42.5771.0
²H10.015%6.5361.1×10⁻⁶
¹³C1/21.07%10.7081.8×10⁻⁴
¹⁴N199.63%3.0781.0×10⁻³
¹⁵N1/20.37%-4.3163.9×10⁻⁶
¹⁹F1/2100%40.0780.83
³¹P1/2100%17.2356.6×10⁻²
¹¹B3/280.1%13.6600.13
²⁹Si1/24.68%-8.4653.7×10⁻⁴
¹⁹⁵Pt1/233.8%9.1533.4×10⁻³
¹¹⁹Sn1/28.59%-15.8664.5×10⁻³
²⁰⁷Pb1/222.1%8.9182.0×10⁻³

On a 600-MHz spectrometer (B₀ = 14.1 T), the Larmor frequencies are: ¹H 600 MHz, ¹³C 150.9, ¹⁹F 564.7, ³¹P 242.9, ¹⁵N 60.8, ²⁹Si 119.2, ¹¹⁹Sn 223.8.

1D experiments

¹H NMR

The 90° pulse-and-acquire workhorse. Chemical shifts 0-12 ppm; reference TMS at 0 ppm in CDCl₃, residual solvent signals (CDCl₃ 7.26, DMSO-d₆ 2.50, CD₃OD 3.31, D₂O 4.79, C₆D₆ 7.16, acetone-d₆ 2.05) as practical referencing in routine work. Multiplicity from scalar (J) coupling: n+1 rule for first-order, second-order patterns (AB, ABX, AA’BB’, AA’XX’) when chemical shifts close to coupling magnitudes. Integration directly proportional to proton count.

Typical acquisition: 30°-90° pulse, 1-3 s acquisition + relaxation delay, 16-64 scans for ~mg sample at 400 MHz.

¹³C NMR

Chemical shift range 0-220 ppm — far better resolution than ¹H. Complications: low natural abundance (1.07%); long T1 (1-100 s for quaternary carbons → need long relaxation delays or paramagnetic relaxation agent Cr(acac)₃); proton coupling makes spectra unreadable unless decoupled.

Standard ¹³C{¹H} — composite-pulse decoupling (WALTZ-16, GARP, SUSAN) on ¹H channel collapses C-H multiplets and gives NOE enhancement (η_max = 0.5 × γ_H/γ_C ≈ 2 → 3× intensity boost). Acquisition: 30° pulse, 1.5 s acquisition + 2 s delay; 1024-16384 scans typical.

DEPT (Distortionless Enhancement by Polarization Transfer)

Doddrell-Pegg-Bendall 1982. Transfers ¹H polarization to ¹³C via INEPT, edits by τ delay = 1/(2J_CH). DEPT-90 shows only CH; DEPT-135 shows CH and CH₃ up, CH₂ down, quaternary absent. ~4× sensitivity gain over standard ¹³C (factor γ_H/γ_C from polarization transfer). Standard tool for C-H multiplicity assignment.

APT (Attached Proton Test)

Patt-Shoolery 1982; modification of standard ¹³C with refocusing delay; CH/CH₃ vs CH₂/quaternary phase distinction. Less sensitive than DEPT but shows quaternary carbons too.

Quantitative ¹³C

Inverse-gated decoupling (decouple only during acquisition, not during relaxation delay) → preserves quantitative integration. Add Cr(acac)₃ 50-100 mM as paramagnetic relaxation agent; pulse delay 5×T1_longest (10-30 s typical). 30° pulse for best Ernst-angle sensitivity. Quantitative ¹³C is the standard for polymer composition analysis, complex-mixture quantitation, residual-solvent quantification in API.

Selective experiments

1D NOESY/ROESY/TOCSY (Gaussian or eburp shaped pulse on chosen resonance, mix, observe target). Used to confirm specific connectivities, simplify overlapped 2D, validate stereochemistry.

Homonuclear 2D

COSY (Correlation Spectroscopy)

Jeener 1971 — first 2D NMR experiment. 90°-t1-90°-acquire pulse sequence shows cross-peaks between scalar-coupled (³J typically, sometimes ²J or ⁴J) protons. Phase-sensitive COSY (DQF-COSY — double-quantum filtered, Piantini-Sørensen 1982; or States-TPPI) → narrower diagonal, anti-phase cross-peaks revealing coupling constants. Decisive for spin-system identification in complex molecules.

TOCSY (Total Correlation Spectroscopy)

Braunschweiler-Ernst 1983 (HOHAHA — Homonuclear Hartmann-Hahn). MLEV-17, DIPSI-2, FLOPSY-8 spin-lock sequences propagate magnetization along entire J-coupling network within mixing time τ_m (typically 60-150 ms). Each cross-peak connects every proton in a spin system → ideal for sugar rings, amino-acid sidechains, peptide spin systems. Selective 1D TOCSY (Kessler) targets specific resonance.

NOESY (Nuclear Overhauser Effect Spectroscopy)

Anet-Bourn 1965 NOE; Jeener-Meier-Bachmann-Ernst 1979 2D NOESY. Cross-peaks intensities encode through-space proximity (~5-6 Å limit; ¹/r⁶ distance dependence). For molecules tumbling slowly (large; τ_c × ω₀ > 1) NOE is negative (same phase as diagonal); for fast tumbling small molecules (τ_c × ω₀ < 1) NOE is positive (opposite phase). At ω₀τ_c ≈ 1.1 (~MW 1000 at 600 MHz) NOE vanishes — use ROESY instead.

ROESY (Rotating-frame Overhauser Effect Spectroscopy)

Bothner-By 1984. Spin-lock during mixing time → NOE always negative regardless of τ_c. Standard for medium-sized molecules (peptides, oligosaccharides 500-3000 Da) where NOESY fails. T-ROESY (Hwang-Shaka) suppresses TOCSY-NOESY mixing artifacts.

EXSY (Exchange Spectroscopy)

Same pulse sequence as NOESY but cross-peaks now indicate chemical exchange on the ms-s time scale. Rotamers, tautomers, conformational exchange, ligand-bound vs free states.

INADEQUATE (Incredible Natural Abundance Double-Quantum Transfer)

Bax-Freeman-Frenkiel 1980. ¹³C-¹³C connectivities at natural abundance — requires both adjacent atoms to be ¹³C (probability ~10⁻⁴ for two adjacent carbons). Extreme sensitivity penalty; needs ~100 mg sample on standard 400 MHz, ~10 mg on 600 MHz cryoprobe, weekend acquisition. Unambiguous carbon skeleton; the “gold standard” for natural product carbon skeleton verification.

J-resolved

Aue-Karhan-Ernst 1976. Separates chemical shift (F2) from coupling (F1) — homonuclear J-resolved cleans up overlapped ¹H multiplets. Useful for complex mixtures and metabolomics where overlap obscures.

Pure shift NMR

Modern technique (Zangger-Sterk; Pure SHift Yielded Anisotropic-data PSYCHE — Foroozandeh, Adams, Morris 2014) eliminates ¹H-¹H J-coupling — gives a single singlet per chemical environment, dramatically simplifying spectra. Used in metabolomics, mixtures, complex natural products.

Heteronuclear 2D

HSQC (Heteronuclear Single Quantum Coherence)

Bodenhausen-Ruben 1980. Inverse-detection (¹H → ¹³C polarization transfer + back to ¹H detection) → sensitivity ~γ_H³/γ_C × natural abundance gain. Each cross-peak ↔ one ¹³C–¹H bond. Edited HSQC (multiplicity-edited; like DEPT-135) — CH/CH₃ vs CH₂ phase distinction in 2D.

Acquisition: 256-1024 t1 increments, 16-128 scans each on a 100-MHz indirect dimension. Sensitivity ~1 mg in 30 minutes at 600 MHz cryoprobe.

HMBC (Heteronuclear Multiple Bond Correlation)

Bax-Summers 1986. ²J and ³J ¹H-¹³C correlations (typically 7-10 Hz coupling) — through-bond connectivity across heteroatoms, quaternary carbons, methylene groups. Long-range delay τ = 1/(2J_long) ~62 ms.

Cross-peak presence vs absence is sometimes ambiguous (anti-phase nature; missing correlations); confirm with low-pass J-filter (suppresses 1-bond correlations). HMBC is the workhorse for organic structure elucidation; combines with HSQC + COSY + NOESY for complete connectivity in moderately sized natural products.

HMQC (Heteronuclear Multiple Quantum Coherence)

Müller 1979; precedes HSQC. Multi-quantum coherence during t1 → broader cross-peaks than HSQC (loss of F1 resolution from H-H couplings). Largely superseded by HSQC except in very large molecules where multi-quantum offers relaxation advantages.

HSQC-TOCSY

Adds TOCSY mixing after HSQC → cross-peaks at each carbon resonance for every proton in its TOCSY-connected spin system. Powerful for crowded mixtures, sugar oligomers.

HOESY (Heteronuclear Overhauser Effect Spectroscopy)

Heteronuclear through-space; ¹H to ¹³C, ¹H to ¹⁹F, etc. Used in supramolecular complexes (host-guest H-F contacts, F-F ion pairing).

Sensitivity-enhanced and PEP

Pulsed-field-gradient coherence selection + Preservation of Equivalent Pathways → √2 sensitivity gain over conventional HSQC/HMBC. Now standard on modern spectrometers.

Cross-polarization-INEPT pseudo-2D and pure-shift HSQC

Modern variants combine pure-shift and HSQC for ultra-narrow cross-peaks (Foroozandeh, Morris).

Heteronuclear nuclei beyond ¹H and ¹³C

¹⁵N NMR

Natural abundance 0.37% prohibitive for direct detection except enriched samples (protein expression in ¹⁵NH₄Cl-supplemented M9 medium → 99% ¹⁵N labeling). Chemical shift range 0-1000 ppm; reference liquid NH₃ at 0 ppm or CH₃NO₂ at 380 ppm. Indirect detection via ¹H{¹⁵N}-HSQC is the foundation of all protein NMR.

¹⁹F NMR

I = 1/2, 100% natural abundance, γ similar to ¹H. Chemical shift range -200 to +400 ppm (referenced to CFCl₃ at 0 ppm). High sensitivity; no overlap with ¹H spectra → ideal probe in F-labeled drugs, F-containing materials, ⁵F-substituted amino acids, fluorinated polymers, fluorinated solvents. ¹⁹F-¹⁹F couplings (³J 5-15 Hz, ⁴J 10-40 Hz) — informative. ¹⁹F-¹H couplings (³J 45-80 Hz on CFH; ²J 50-90 Hz on CF₂H) — large.

¹⁹F PET (fluorine-19 magnetic resonance imaging) — emerging in vivo imaging modality using ¹⁹F probes (perfluorocarbon emulsions, ⁵-fluorocytosine for fungal infection, ¹⁹F protein labels).

³¹P NMR

I = 1/2, 100%, γ = 17.2 MHz/T. Chemical shift -300 to +500 ppm vs 85% H₃PO₄ at 0 ppm. Phosphine ligands (PPh₃ -5, PCy₃ +10, dppe +30), phosphonates, phosphates, ATP/ADP/AMP triphosphate pattern (γ -5.6, α -10.5, β -19.2 ppm). In-cell metabolic ³¹P MRS — clinical (Phys 1-3 T) — measures ATP/PCr ratios in muscle, brain, tumor.

¹¹B NMR

I = 3/2 quadrupolar; 80% abundance. Chemical shift range -100 to +90 ppm vs BF₃·OEt₂ at 0 ppm. Boranes, boronic acids/esters, boronate clusters (C₂B₁₀H₁₂), boron-doped graphene. Quadrupolar broadening manageable in low-symmetry environments.

²⁹Si NMR

I = 1/2, 4.68% abundance, γ negative. Chemical shift -350 to +50 ppm vs TMS at 0 ppm. Q⁰/Q¹/Q²/Q³/Q⁴ groups (numbered by Si-O-Si bridges) characterize silicates and silica networks. Cross-polarization (CP) MAS ²⁹Si is the workhorse for amorphous-silica characterization, zeolites, silsesquioxanes, sol-gel materials.

Metal NMR

  • ¹⁹⁵Pt (33.8%, I = 1/2, γ = 9.15). Range >10,000 ppm. Cisplatin and carboplatin pharmaceutical analysis; ligand binding; Pt(II)/Pt(IV) discrimination.
  • ¹¹⁹Sn (8.6%, I = 1/2, γ = -15.9). Range -2000 to +2500 ppm. Organotin compounds (Bu₃SnH δ -94, Me₄Sn 0 ppm).
  • ²⁰⁷Pb (22%, I = 1/2). Range >16,000 ppm. Pb(IV)/Pb(II) discrimination in organolead, perovskite materials.
  • ⁵⁹Co, ⁵¹V, ⁹³Nb, ⁹⁵Mo, ¹⁸³W, ¹⁰³Rh — transition metal NMR ladder; quadrupolar except ¹⁰³Rh and others I=1/2.

DOSY — Diffusion-Ordered Spectroscopy

Morris-Johnson 1992. Pulsed-field-gradient (PFG) experiment encodes molecular diffusion coefficient D in F1 dimension; chemical shift in F2. Stokes-Einstein D = k_B T / (6πηr_h) → estimate hydrodynamic radius r_h (and thus MW for globular species). Used to:

  • Distinguish components of mixtures by MW.
  • Quantify host-guest binding (free vs bound diffusion coefficient).
  • Polymer dispersity analysis.
  • Aggregation studies (monomer vs dimer vs oligomer).

Standard pulse sequence: stimulated-echo with bipolar gradients (BPP-LED — Wu-Chen-Johnson 1995); 8-32 gradient strengths from 5-95% max gradient. Magritek and Bruker spectrometers offer optimized DOSY packages.

Caveat: DOSY assumes single-population diffusion; multimodal distributions (e.g., polymers with multiple MW components) require CONTIN-style or biexponential fitting (DECRA, ITAMeD).

Solid-state NMR (ssNMR)

Powder NMR shows broad, anisotropic line shapes from chemical-shift anisotropy (CSA), dipolar coupling (³P_2(cosθ) angular dependence), and quadrupolar interactions. Narrowing the lines opens solid-state to organic structure determination.

Magic Angle Spinning (MAS)

Andrew-Bradbury-Eades 1958. Spin the sample at the magic angle θ_m = 54.74° (cos²θ_m = 1/3) at frequencies matching the anisotropy magnitude (>10 kHz for ¹H CSA, >20 kHz for ¹³C in organics, 60-110 kHz for proton-detected biomolecular ssNMR). Modern probes: 0.7-1.3 mm rotors at 100-160 kHz MAS (Bruker, JEOL, Phoenix-NMR).

Cross-polarization (CP)

Pines-Gibby-Waugh 1973. Match Hartmann-Hahn condition γ_H B_1H = γ_C B_1C in spin-locked frame → polarization transfer from abundant ¹H to dilute ¹³C. ~4× sensitivity gain, faster repetition (use ¹H T1 not ¹³C T1). Standard ¹³C ssNMR.

CP-MAS

Combine CP + MAS — the workhorse organic ssNMR experiment.

Heteronuclear decoupling

TPPM (two-pulse phase modulation; Bennett-Rienstra 1995), SPINAL-64, XiX, swfTPPM. Modern ¹H decoupling at fast MAS reduces residual line width to <1 ppm for ¹³C in organic solids.

CRAMPS (Combined Rotation and Multi-Pulse Spectroscopy)

For ¹H solid-state — multi-pulse homonuclear decoupling (WHH-4, MREV-8, BR-24, BLEW-12) + slow MAS combine to narrow ¹H lines.

FROSTY

Frequency-switched Lee-Goldburg in homonuclear decoupling; used in modern proton-detected ssNMR sequences.

Dynamic Nuclear Polarization (DNP)

Griffin (MIT) 1993 — irradiate stable polarized electrons (TOTAPOL, AMUPol, TEKPol biradicals) at electron Larmor frequency under microwave illumination → transfer polarization to coupled nuclear spins via solid effect, cross-effect, or thermal mixing. Signal enhancements 50-200× → ssNMR of dilute species (drug formulations at trace concentrations; surface species on catalysts; metabolites in cell pellets). Bruker DNP systems at 9.4-14.1 T with gyrotron μ-wave sources at 263 GHz / 395 GHz / 527 GHz.

Cryogenic operation (~100 K) standard. ³⁹⁵ GHz 600-MHz commercial systems (Bruker BioSpin) since 2013; Hyperpolarization Solutions (Karlsruhe spin-out) and Bruker Daltonics dominate.

Surface and porous-materials NMR

Catalyst surface species (zeolite Brønsted sites, MOF defects, supported metal nanoparticle ligand shells) — DNP enhanced ssNMR (Emsley Lausanne, Lesage Lyon) routinely visualizes surface chemistry.

NMR crystallography

Combine high-resolution ssNMR with DFT chemical-shift prediction (GIPAW — gauge-including projector-augmented wave; Pickard-Mauri 2001; CASTEP-NMR, Quantum ESPRESSO QE-GIPAW) to assign and refine polymorphic crystal structures. Routinely applied to APIs to characterize amorphous dispersions and polymorphs.

Protein NMR

Wüthrich 1980s (Nobel 2002) — first solution-state protein structure (BPTI) by NMR. Now standard for ≤30 kDa proteins in solution; complementary to crystallography and cryo-EM. Cross-link structural-biology and cryo-em-and-structural-determination.

Isotope labeling

  • U-¹⁵N — backbone amide HN-N correlations. M9 medium with ¹⁵NH₄Cl (Cambridge Isotope Labs CIL CNLM-467; CAS 39466-62-1).
  • U-¹⁵N/¹³C — backbone + sidechain. ¹³C₆-glucose CIL CLM-1396 (CAS 110187-42-3).
  • Per-deuteration (²H) — relaxes ¹³C TROSY at high MW. CIL DLM-2062 D-glucose.
  • Selective labels — ILV (isoleucine, leucine, valine) methyl-protonation in otherwise deuterated background (Kay-Tugarinov methyl TROSY) for proteins >50 kDa.
  • Segmental labeling — sortase-mediated or expressed protein ligation; specific labeling of one domain in multidomain protein.
  • Site-specific ¹⁹F-amino acids (5-fluoro-Trp, 4-fluoro-Phe, 4-CF₃-Phe) via amber-suppressor systems (Schultz Scripps).

Backbone assignment experiments (triple-resonance)

All proton-detected; ¹⁵N + ¹³C(α/β/CO) coordinates for sequential resonance assignment:

ExperimentMagnetization pathwayInformation
HNCAHN→N→Cα(i,i-1)→HNCα of own residue + previous
HN(CO)CAHN→N→CO→Cα(i-1)Cα of previous only
HNCOHN→N→CO(i-1)CO of previous
HN(CA)COHN→N→Cα→CO(i,i-1)CO own + previous
CBCA(CO)NHCα/β→CO→N→HNCα and Cβ of previous
HNCACBHN→N→Cα,Cβ(i,i-1)Cα, Cβ own + previous

Combine: assign each Cα/Cβ to a residue type (Cβ chemical shifts distinguish A, S, T, etc.) → walk along chain matching i and i-1 chemical shifts.

Sidechain assignment

  • HCCH-TOCSY, HCCH-COSY — sidechain carbon-proton networks.
  • CC(CO)NH-TOCSY — sidechain ¹³C correlations to backbone amide.
  • HC(C)H-TOCSY — sidechain ¹H to ¹H via ¹³C.

Protein structure determination

  1. NOE-derived distance restraints. ¹⁵N-edited and ¹³C-edited NOESY-HSQC, ¹³C-¹³C/¹H-¹H NOESY at 80-150 ms mixing → distance bins (≤2.5 Å strong, ≤3.5 Å medium, ≤5.5 Å weak).
  2. Dihedral angles from chemical shifts. TALOS-N or TALOS+ (Cornilescu-Bax) predict φ/ψ from chemical shifts.
  3. Hydrogen bonds. From slow H/D exchange (H/D-protected amides → buried in core; H-bonded to acceptor).
  4. RDCs (residual dipolar couplings). Weakly align protein in liquid-crystal (filamentous phage Pf1, n-alkyl-PEG, polyacrylamide gel) → residual one-bond ¹H-¹⁵N or ¹H-¹³C couplings encode orientation of inter-nuclear vectors relative to alignment tensor. Long-range orientational restraints.
  5. PRE (paramagnetic relaxation enhancement). Attach MTSL spin label or paramagnetic metal at engineered Cys → enhance T2 of nearby ¹H within ~25 Å. Distance restraints to remote points.
  6. PCS (pseudocontact shifts). Lanthanide-binding tag → through-space contact shift contains distance + orientation information.

Structure calculation: XPLOR-NIH (Schwieters-Kuszewski-Clore), CYANA (Güntert; automated NOE peak assignment + structure simulated-annealing), ARIA (Linge-Nilges; ambiguous-distance handling), Rosetta-NMR.

TROSY (Transverse Relaxation-Optimized Spectroscopy)

Pervushin-Riek-Wider-Wüthrich 1997 PNAS 94:12366. Select the slower-relaxing component of the ¹⁵N-¹H multiplet → much narrower lines for high-MW proteins. Combined with per-deuteration, enables protein NMR on >100 kDa systems (Kay GroEL, Wüthrich 23S ribosome subunits, Membrane proteins in micelles/bicelles/nanodiscs).

Membrane protein NMR

Detergent micelles (DPC, DDM, LDAO) for small membrane proteins; bicelles (DMPC/DHPC) for size match; nanodiscs (MSP-bounded lipid disks; Sligar Illinois) for native-like membrane; lipodisks; SMA nanodiscs (styrene-maleic-acid copolymer; Knowles, Dafforn 2009).

Solid-state ssNMR of membrane proteins (Baldus, McDermott, Reif, Polenova) — proton-detected MAS at 100-110 kHz on per-deuterated samples. M2 ion channel, KcsA, ATP-binding cassette transporters.

IDPs (intrinsically disordered proteins)

Narrow chemical-shift dispersion (no secondary structure) → 3D-4D NMR mandatory. ¹⁵N CEST and CPMG dispersion → invisible transient structure (Kay, Mittag, Tjandra). Used in α-synuclein, tau, p53 transactivation domain, FUS, eIF4G LCR.

Dynamics

  • Fast (ps-ns) — T1, T2, ¹⁵N-{¹H} NOE. Lipari-Szabo model-free analysis (S² order parameter, τ_e internal correlation time). Local backbone flexibility.
  • µs-ms — CPMG relaxation dispersion (Loria-Rance-Palmer 1999). Measure R₂_eff vs CPMG frequency; extract k_ex, p_A, Δω of exchange between ground and minor excited state.
  • µs-ms — CEST (chemical exchange saturation transfer). Long, weak RF saturation at variable offsets → dip when saturation hits invisible minor state; characterizes minor state populations 0.5-10% and chemical shifts. Vallurupalli-Bouvignies-Kay 2012 J Am Chem Soc.
  • ms-s — EXSY, lineshape analysis. Slow conformational/chemical exchange.

Instrumentation

Magnets

  • Resistive (electromagnet) — 0.5-1.5 T legacy; superseded.
  • Permanent magnet — Spinsolve (Magritek; Aachen + Wellington) benchtop 1.0-1.4 T (43-62 MHz); Nanalysis (Calgary), Anasazi (now obsolete), Oxford Instruments PulsarMR (now retired). Used in process monitoring, QC, education.
  • Superconducting NbTi/Nb₃Sn. Liquid-He bath; persistent mode. Standard 400-1000 MHz.
  • High-temperature superconducting (HTS). YBa₂Cu₃O₇ at ~30-50 K. Bruker Avance NEO 1.0/1.1/1.2 GHz Ascend platforms — current state-of-the-art commercial. Bruker 1.2 GHz installed at >20 sites globally (CERM Florence first 2020; CEMHTI Orléans, EPFL Lausanne, Yale, MPI Heidelberg, IIIT Hyderabad, Boğaziçi Istanbul). 1.3 GHz under development.

Console

Modern Bruker Avance NEO, JEOL ECZL, Varian/Agilent VNMRS (Agilent NMR product line discontinued 2014 — service continues via Resonant Spec, Aspect Imaging). Multi-receiver consoles enable parallel acquisition of different nuclei.

Probes

  • Inverse-detect ¹H{X} — standard for organic chemistry.
  • TXI / TXO / TBI broadband observe — multi-nuclear ¹H, ¹³C, ¹⁵N, ³¹P, ¹⁹F.
  • Cryoprobe (CryoProbe, ColdProbe, CRYOPlatform). Cooled RF coils (~20 K) reduce thermal noise → 4× sensitivity (1/√S/N gain → 16× faster acquisition). Bruker QCI-F (¹H ¹³C ¹⁵N ¹⁹F cold), TCI (¹H ¹³C ¹⁵N + ²H lock). JEOL Royal cryo. Indispensable on natural products, metabolomics, protein NMR.
  • MicroCoil and capillary probes — Protasis, Magritek; nL-mL volumes. Used in mass-limited samples (single peptide synthesis, hyphenated LC-NMR).
  • HR-MAS — High-resolution magic-angle-spinning probe for swollen-gel and tissue samples. Resin-bound peptide synthesis, intact tissue metabolomics (Hatcher-Becerra ex vivo brain biopsy).
  • MAS solid-state probes — 0.7-3.2 mm rotors at 100-160 kHz MAS, 60-110 kHz typical.

Field-shift / ambient field

Spectrometer fields drift; field-frequency lock (²H lock on deuterated solvent — CDCl₃, DMSO-d₆) compensates. Modern Bruker Spinpack lock channel uses ¹⁹F or ⁷Li as alternate lock nuclei when deuterium not feasible.

Shimming

Gradient shimming (Hurd; Bruker TopShim, JEOL AutoShim) iteratively optimizes shim currents based on gradient-echo profile. Achieves <0.5 Hz line width on ¹H of CHCl₃ on 600 MHz routinely.

Sample tubes

  • 5 mm (180 µL min) — standard organic. Norell 502-PP-7, Wilmad 528-PP-7.
  • 3 mm — reduced volume (160 µL) for limited samples; cryoprobe compatible.
  • 1.7 mm microtube — Bruker SampleJet capable; ~50 µL volume.
  • Shigemi tubes — magnetic-susceptibility-matched plug + glass for limited samples in 5 mm.

Software ecosystem

Acquisition

  • TopSpin (Bruker) — Bruker proprietary; current TopSpin 4.x. Pulse program editor (Pulse Program Language); IconNMR for automation.
  • Delta (JEOL) — JEOL spectrometers.
  • VnmrJ (Varian/Agilent legacy) — Open-source successor OpenVnmrJ (UCSF).
  • NMRSpec, SpinAPI — research/educational.

Processing

  • MestreNova (Mestrelab Research, Spain) — commercial; dominant in organic chemistry labs; auto-phase, baseline, integration, multiplet analysis, advanced 2D, structure verification (mnova MNVal). Cross-format reader.
  • TopSpin — also processes.
  • NMRPipe (Delaglio Frank NIST) — Unix-style pipeline; protein NMR standard. Free.
  • NMRViewJ / NMRFx (One Moon Scientific, Bruce Johnson) — protein NMR analysis successor.
  • Sparky (Goddard UCSF; now NMRFAM-SPARKY by Lee-Markley Madison) — peak picking, assignment, integration. Free; protein NMR standard tool.
  • CCPN Analysis (CCPN consortium; Vranken-Boucher) — protein NMR analysis; integrates assignment + structure calc. Free.
  • POKY (Lee-Yang-Markley NMRFAM 2021) — modern SPARKY replacement.
  • DAMNA, NUTS — academic/legacy tools.

Database and chemical-shift prediction

  • NMRShiftDB (Steinbeck-Krause-Kuhn) — open-access ¹H, ¹³C database; ~50,000 spectra.
  • BMRB (Biological Magnetic Resonance Data Bank; Madison) — protein/nucleic-acid chemical shift database; ~14,000 entries 2024.
  • SDBS (Spectral Database for Organic Compounds; AIST Japan) — ¹H, ¹³C, IR, MS, Raman; ~30,000 compounds.
  • CAS SciFinder NMR — chemical-shift queries within CAS substances.
  • ChemDraw + ChemNMR (PerkinElmer) — chemical-shift prediction by additivity rules + ANN trained model.
  • ACD/NMR Predictor (Advanced Chemistry Development, Toronto) — most accurate commercial prediction; HOSE-code-trained.
  • Mnova Predict — Mestrelab.
  • NMR-AI (deep learning predictors) — emerging; CASCADE (Bagdasarian-Lee), DimoNet, IMPRESSION.

Quantum-chemistry shift prediction

DFT (GIAO — Gauge-Including Atomic Orbitals; mPW1PW91/6-311+G(2d,p) or B3LYP/6-31+G(d,p) standards) + scaling with empirical correction. CHESHIRE (Lodewyk-Siebert-Tantillo) calibration. NMR crystallography for solids via GIPAW (Pickard-Mauri 2001). Cross-link computational-chemistry-deep for method detail.

Workflow — organic structure elucidation

Standard package on unknown natural product (~10 mg)

  1. ¹H 1D — count protons, identify functional groups by chemical shift.
  2. ¹³C 1D — count carbons, identify quaternaries.
  3. DEPT-135 — assign C, CH, CH₂, CH₃.
  4. COSY — proton spin systems.
  5. HSQC — direct C-H assignments + multiplicity edit.
  6. HMBC — long-range connectivity through heteroatoms and quaternaries.
  7. NOESY (or ROESY for medium-sized) — stereochemistry, conformational preferences.
  8. Selective 1D experiments — confirm specific couplings, NOEs.
  9. Optional: ¹H J-resolved, INADEQUATE — for fully assigned skeletal verification.
  10. DOSY — mixtures, aggregation, dispersity.

Total time on 600-MHz cryoprobe: 4-12 hours for full set on 5-10 mg sample.

Structure verification via Mnova MNVal

Computed structure → predicted ¹H and ¹³C → match against experimental. Confidence score from automated peak assignment + shift agreement. Used in pharma QC to verify lot-to-lot identity at scale.

CASE (Computer-Assisted Structure Elucidation)

ACD/Structure Elucidator, Mnova Structure Elucidator, COCON — input 2D NMR cross-peak lists + MS molecular formula → enumerate consistent structures. Used in fully blinded natural product elucidation.

Applications across chemistry

Pharmaceutical analysis

  • Identity confirmation of API and intermediates.
  • Quantification (qNMR; ERETIC reference) — primary metrological standard for pharmaceutical reference materials (USP, EDQM, PMDA).
  • Stereochemistry — enantiomeric excess via chiral shift reagents (Eu(hfc)₃, Eu(tfc)₃), chiral solvating agents (Pirkle’s alcohol), in-cell ¹⁹F labels with chiral CD.
  • Polymorph identification — solid-state ¹³C CP-MAS distinguishes API polymorphs.
  • Amorphous solid dispersion (ASD) characterization — ssNMR + DNP measures API molecular dispersion in polymer matrix.

Metabolomics

¹H NMR of urine, plasma, CSF, tissue extracts; integration of >100 metabolites per sample. Chenomx, Mnova Metabolomics, BMRB Metabolomics, Bruker Avance IVDr Profiler. Quantitative; reproducible across labs and instruments — major advantage over LC-MS metabolomics for population studies. Examples: COMET (NCI cohort), UK Biobank ¹H NMR profile (~120,000 participants by 2023).

Polymer characterization

  • Composition of copolymers (mol %).
  • End-group analysis for low-MW polymers.
  • Tacticity (isotactic, syndiotactic, atactic) from ¹³C splitting.
  • Sequence distribution (Bernoullian vs first-order Markov).
  • Branching quantification (LDPE, LCB-PE) by ¹³C.
  • MW by DOSY at low MW where SEC fails.

Natural products

Total elucidation of complex marine natural products (palytoxin C129H223N3O54 — Moore-Bartlett 1981), terpenoids, alkaloids, peptide secondary metabolites. NMR remains the technique for natural product structures (mass spec gives formula, NMR gives structure, X-ray for crystals when available).

Inorganic and organometallic

  • ³¹P for phosphine ligands; ¹⁹⁵Pt for Pt complexes (cisplatin, carboplatin); ⁵⁹Co for cobalt complexes; ¹¹B for boranes; ²⁹Si for silanes/siloxanes; ¹¹⁹Sn for organotin; ⁵¹V for vanadates; ¹⁰³Rh for Rh catalysts.

Reaction monitoring

Online flow-NMR — Magritek Spinsolve Benchtop in continuous flow reactors. Bruker InsightMR for high-field online monitoring. Reaction kinetics, intermediate identification, optimization. Cross-link green-chemistry-and-process-intensification for flow context.

Hyperpolarized in vivo NMR/MRI

¹³C-pyruvate dynamic-nuclear-polarization (Ardenkjær-Larsen 2003 PNAS 100:10158) → ~10,000× signal enhancement; clinical trials in prostate, glioma, breast cancer (Vigneron UCSF, Wilson UCSF, Spielman Stanford, Brindle Cambridge, Lin Princeton). General Electric/MIT Hypersense, Bruker SpinAligner clinical platforms.

Sample preparation

Solvent choice for solution NMR

Standard deuterated solvents (Cambridge Isotope Labs CIL, Sigma-Aldrich, Eurisotop):

SolventResidual ¹H¹³CWater peakNotes
CDCl₃7.2677.161.56Workhorse; mildly acidic; degrades pH-sensitive compounds
DMSO-d₆2.5039.523.33High dissolving power; very hygroscopic
CD₃OD3.3149.004.87Hydroxyl + acidic protons exchange to OD
D₂O4.79——Aqueous; exchangeable protons disappear
C₆D₆7.16128.060.40Aromatic shielding; π-stack-sensitive shifts
Acetone-d₆2.0529.84 (CD₃), 206.26 (C=O)2.84Volatile; intermediate polarity
CD₃CN1.941.32, 118.262.13Polar aprotic; good for charged species
THF-d₈1.72, 3.5825.31, 67.212.46Ethereal; air-sensitive compounds
Toluene-d₈2.08, 6.97-7.0920.43, 125-1380.43Higher-T VT
Pyridine-d₅7.19, 7.55, 8.71123.5-149.94.96H-bond-acceptor; sugars; alkaloids

Sample concentration

  • Routine ¹H: 5-20 mg in 0.6 mL for 5 mm tube → ~10-50 mM.
  • ¹³C: 30-100 mg routinely (legacy); 5-30 mg with cryoprobe.
  • 2D HSQC: 1-5 mg sufficient on 600 MHz cryoprobe in overnight.
  • 2D HMBC: 3-10 mg typical for clean cross-peaks.
  • Protein 3D: 0.3-1 mM in 280 µL Shigemi tube; isotope labeling mandatory.

Tube fill height

5-mm tube: 600 µL fills the active region (~40 mm depth) and provides shim volume; 450 µL minimum for cryoprobe; 600 µL standard. Shigemi tube (susceptibility-matched plug + restricted volume): 280-320 µL.

Sample purity considerations

  • Paramagnetic impurities (Fe, Cu, Mn) drastically shorten T2 → broad lines. Use Chelex 100 cleanup or EDTA spike for biological samples.
  • Acid/base contaminants from previous chromatography → shift sensitive resonances. Wash through silica or basic alumina.
  • Volatile impurities (residual solvent from synthesis) → may appear as extra peaks; consult spectroscopy-reference-tables for residual-solvent table (Gottlieb-Kotlyar-Nudelman 1997 J Org Chem 62:7512).
  • Water/moisture in non-aqueous deuterated solvents → integrates as massive peak; activate molecular sieves or use freshly opened ampoules.

Cryoprobe-specific considerations

Higher salt sensitivity → measurement of biological samples in >150 mM ionic-strength buffer requires precooling or longer pulse calibration. Salt-tolerant probes (Bruker QCI Salt-Tolerant, JEOL Royal MIcro-Salt-Tolerant) help.

Troubleshooting

Broad lines

  • Poor shim → re-shim with TopShim or manual Z₁/Z₂.
  • Paramagnetic contamination → Chelex cleanup, switch glassware, check sample-prep tools (some stainless steel needles release Fe).
  • Concentration too high / aggregation → dilute 2-10×; observe line-width change.
  • Exchange broadening (slow exchange near coalescence) → cool or heat sample to push off coalescence (Δν ≈ k_ex/π at coalescence).
  • High viscosity (large macromolecule) → heat sample within stability limits or accept the broadening.

Phase distortion

  • First-order phase wrong → adjust pivot-point phase; recheck zero-order at peak of interest.
  • Bloch-Siegert in selective experiments → use composite pulses (BURP family).
  • Baseline-rolling 2D → linear prediction in indirect dimension; better t1 sampling (NUS — non-uniform sampling, Wagner-Hyberts 2007).

Integration unreliable

  • Long T1 + short relaxation delay → quaternary or aromatic CH undercounted; extend delay or use Ernst-angle pulse.
  • Strong coupling distortions → reduce concentration, change solvent, or measure at higher field where Δν ≫ J.
  • Overlapping multiplets → 2D pure-shift NMR (PSYCHE) or fit deconvolution in MestreNova.

Cross-peak missing in HMBC

  • Coupling constant ³J ≪ 1/(2 τ): adjust τ to match suspected J (62 ms for 8 Hz; 100 ms for 5 Hz).
  • Active resonance suppressed by ¹³C-decoupling — switch to non-decoupled HMBC variant.

Excessive ¹³C run time

  • Switch to cryoprobe (4× sensitivity → 16× faster).
  • Use indirect-detect 2D (HSQC + HMBC) to extract carbon shifts via ¹H detection.
  • Use DEPT instead of full ¹³C if quaternaries not needed.
  • Use SOFAST-HMQC / band-selective methods for protein NMR.

NUS (non-uniform sampling)

Wagner-Hyberts Harvard 2007; sample only some t1 points (random or Poisson-distributed schedule) → faster 2D/3D/4D at cost of post-processing (MDD — multidimensional decomposition; SMILE — sparse multidimensional iterative lineshape enhanced; IST — iterative soft thresholding). Standard in protein 4D experiments; growing in small-molecule 2D.

Adjacent

Further reading

  • Keeler, J. — Understanding NMR Spectroscopy, 2nd ed. Wiley 2010 — best modern pedagogical treatment of pulse-NMR.
  • Levitt, M.H. — Spin Dynamics: Basics of Nuclear Magnetic Resonance, 2nd ed. Wiley 2008 — rigorous quantum-mechanical foundations.
  • Cavanagh, J., Fairbrother, W.J., Palmer, A.G., Rance, M., Skelton, N.J. — Protein NMR Spectroscopy: Principles and Practice, 2nd ed. Academic 2007 — protein NMR bible.
  • Claridge, T.D.W. — High-Resolution NMR Techniques in Organic Chemistry, 3rd ed. Elsevier 2016 — practical 1D/2D NMR for organic chemists.
  • Friebolin, H. — Basic One- and Two-Dimensional NMR Spectroscopy, 5th ed. Wiley-VCH 2010.
  • Silverstein, R.M., Webster, F.X., Kiemle, D.J., Bryce, D.L. — Spectrometric Identification of Organic Compounds, 8th ed. Wiley 2014 — combined NMR/IR/MS/UV-Vis interpretation.
  • Pretsch, E., Bühlmann, P., Badertscher, M. — Structure Determination of Organic Compounds: Tables of Spectral Data, 5th ed. Springer 2020.
  • Wüthrich, K. — NMR of Proteins and Nucleic Acids. Wiley 1986 — Nobel-winning protein NMR foundation.
  • Duer, M.J., ed. — Solid-State NMR Spectroscopy: Principles and Applications. Blackwell 2002.
  • Anslyn, E.V., Dougherty, D.A. — Modern Physical Organic Chemistry. University Science Books 2006 — physical organic context for chemical-shift and coupling interpretation.
  • Smith, M.B. — March’s Advanced Organic Chemistry, 8th ed. Wiley 2020 — reaction mechanisms whose intermediates NMR identifies.
  • Crabtree, R.H. — The Organometallic Chemistry of the Transition Metals, 7th ed. Wiley 2019 — for ³¹P, ¹⁹⁵Pt, ¹⁰³Rh organometallic context.
  • Frenking, G., Shaik, S. — The Chemical Bond + The Chemical Bond Across the Periodic Table. Wiley-VCH 2014 — bonding context for chemical-shift interpretation.

Graph View

Backlinks