Cryogenic and Dissolution DNP NMR on γ-Irradiated Organic Molecules

Nuclear magnetic resonance (NMR) plays a central role in the elucidation of chemical structures but is often limited by low sensitivity. Dissolution dynamic nuclear polarization (dDNP) emerges as a transformative methodology for both solution-state NMR and metabolic NMR imaging, which could overcome this limitation. Typically, dDNP relies on combining a stable radical with the analyte within a uniform glass under cryogenic conditions. The electron polarization is then transferred through microwave irradiation to the nuclei. The present study explores the use of radicals introduced via γ-irradiation, as bearers of the electron spins that will enhance 1H or 13C nuclides. 1H solid-state NMR spectra of γ-irradiated powders at 1–5 K revealed, upon microwave irradiation, signal enhancements that, in general, were higher than those achieved through conventional glass-based DNP. Transfer of these samples to a solution-state NMR spectrometer via a rapid dissolution driven by a superheated water provided significant enhancements of solution-state 1H NMR signals. Enhancements of 13C signals in the γ-irradiated solids were more modest, as a combined consequence of a low radical concentration and of the dilute concentration of 13C in the natural abundant samples examined. Nevertheless, ca. 700–800-fold enhancements in 13C solution NMR spectra of certain sites recorded at 11.7 T could still be achieved. A total disappearance of the radicals upon performing a dDNP-like aqueous dissolution and a high stability of the samples were found. Overall, the study showcases the advantages and limitations of γ-irradiated radicals as candidates for advancing spectroscopic dDNP-enhanced NMR.


Additional solid-state
The solid black and blue lines are the fit to the data using a single exponentials, from which enhancement were calculated as described in the main text.

Additional solid-state 1 H NMR data on the hybrid NMR-EPR spectrometer
. DNP results arising for 150 kGy γ-irradiated glucose from a batch different from that used in Fig. 3. 1 H signals vs buildup times were here measured at 94.98 GHz (glucose) or 94.68 GHz (Ala) using 4 μs (glucose) or 8 μs (Ala) 90˚ pulse in presence and absence of microwaves (black and blue points, respectively).Solid lines are fits to the data using a single exponential component and the reported enhancement calculated as described in the text.

Figure S1 .
Figure S1.Room temperature X-band (a) CW-EPR spectra of 1 mg of two of the irradiated powders sealed under vacuum and irradiated with different doses of γ-irradiation, (b) double integrals of the CW-EPR spectra over time for some of the powders and (c) CW-EPR spectra of 1 mg γ-irradiated powders measured immediately after dissolving them in H2O (blue) vs the spectrum of 1 mg powder (black) under same conditions.CW-EPR conditions: 25 dB in (a) or 30 dB in (c), conversion time 40 ms, modulation amplitude 1 G, modulation frequency 100 kHz, 1 scan.
Figure S2. 1 H signals of empty cup at 1.5 K (a) and 1 H signal vs microwave irradiation times of γirradiated (b) sucrose from batch #1 and (c) salicylic acid from batch #2 at 1.5 K measured on the HS polarizer in presence and absence of microwaves (black and blue points, respectively) using a 4 μs pulse length, following a train of pre-saturation pulses (100 μs pulses, × 100).Microwave conditions: 94.05 GHz, 100 mW in (b) and150 mW power in (a, c).The microwave off data in (b) could not be fitted to a mono-exponential function (as in Fig.2d): it fell on a straight line whose end value was taken as the "off" signal.The inset in (c) gives a representative solid-state 1 H NMR spectrum recorded with 4 μs pulse length and 100 s polarization time in presence of microwaves.

Figure S3 .
Figure S3.(a)1 H signal vs microwave irradiation times of 10 mM 4-aminoTEMPO in H2O/glycerol (3/2), recorded on the HS polarizer (93.99 GHz, 150 mW power) at 1.5 K and a homebuilt 1 H coil tuned at ~143 MHz. 1 H signals were measured in the presence and absence of microwaves (black and blue points, respectively) using a 20 μs pulse length, following a train of pre-saturation pulses (100 μs pulses, × 100).The solid black and blue lines are the fit to the data using a single exponentials, from which enhancement were calculated as described in the main text.

Figure S4. 1
Figure S4.1 H signals vs mw frequency on 150 kGy γ-irradiated powders (indicated samples) in the solid state measured on a HS polarizer at 1.5 K using a home-built 1 H coil tuned at ~143 MHz. 1 H signals were measured as described in Experimental, using mw irradiation times of 4 min each point and 150 mW of power.
Figure S7. 1 H signal decay (single-point FID profiles, shown in black traces) observed upon utilizing 3.5 mL of D2O to dissolve DNP-enhanced samples of γ-irradiated (a) Ala pD 3, (b) AlaAla pD 7, (c) Ala pD 7 and (d) GlyGly pD 7. Shown in (e) are results from water/glycerol co-mixed with 20 mM 4-aminoTEMPO.The data of Ala at pD 7 are affected by radiation damping.AlaAla was fitted by a bi-exponential function whereas the rest of data by a mono-exponential function (fits in red).NMR data recording began ca. 5 s prior to the sample's injections.

Figure S8 .
Figure S8.13 C signal decay (single-point FID point profile) for the γ-irradiated powders indicated on each panel after dissolution with 3.5 mL H 2 O/D 2 O (9/1) (black).Fitting of the data to a monoexponential decay function (red) gave the exponent which reflects 13 C T1 relaxation.NMR data recording began ca. 10 s prior to the sample's injections.

Figure S9 .
Figure S9.13 C data on 13 C1-pyruvic/BDPA after dissolution.(a)13 C signal decay (single-point FID point profile); fitting of the data to a mono-exponential decay function (red) gave the exponent reflecting the joint effects of the pulsing and 13 C T1 relaxation.(b) Hyperpolarized and thermal 13 C solution NMR spectra with the inset focusing on the 150-190 ppm region.40 μl of BDPA dissolved in sulfolane and co-mixed with 13 C1-pyruvic (1/1 v/v) were polarized on a HS polarizer at 1.5 K and at 94.040 GHz using 160 mW power, 2 hrs polarization time.Dissolution was done with 4 mL MeOH and spectra were recorded on a 500 MHz 1 H frequency spectrometer and data recording began ca. 10 s prior to the sample's injections.NMR parameters: 310 K, 1 s acquisition time, 0.05 s relaxation delay, 2 μs (≈13˚) pulse length, 1 scan.The data were recorded as pseudo-2D with the time being the pseudo dimension and a time interval of 1 s from spectrum to spectrum.Hyperpolarized (immediately post-dissolution) vs thermal (FID at 100 s) spectra are shown in blue and red color, respectively.For calculating the enhancements, a thermal spectrum was recorded with 100 scans after the end of the dissolution process and upon sample reaching equilibrium.8.DNP/NMR data on U-13 C6-glucose after 150 kGy γ-irradiation