Improving the Hyperpolarization of 31P Nuclei by Synthetic Design

Traditional 31P NMR or MRI measurements suffer from low sensitivity relative to 1H detection and consequently require longer scan times. We show here that hyperpolarization of 31P nuclei through reversible interactions with parahydrogen can deliver substantial signal enhancements in a range of regioisomeric phosphonate esters containing a heteroaromatic motif which were synthesized in order to identify the optimum molecular scaffold for polarization transfer. A 3588-fold 31P signal enhancement (2.34% polarization) was returned for a partially deuterated pyridyl substituted phosphonate ester. This hyperpolarization level is sufficient to allow single scan 31P MR images of a phantom to be recorded at a 9.4 T observation field in seconds that have signal-to-noise ratios of up to 94.4 when the analyte concentration is 10 mM. In contrast, a 12 h 2048 scan measurement under standard conditions yields a signal-to-noise ratio of just 11.4. 31P-hyperpolarized images are also reported from a 7 T preclinical scanner.

[PdCl 2 (dppf)] (39 mg, 0.05 mmol) was added and the reaction heated to reflux for 16 h. The reaction was allowed to cool and concentrated in vacuo, then purified by flash column chromatography (0-5 % MeOH in EtOAc) to afford the title compound as a dark orange oil (638 mg, 59 %

N-(Diethoxyphosphoryl)-N-methylpyridin-4-amine (L 10 )
A solution of 4-(methylamino)pyridine (216 mg, 2 mmol) and triethylamine (222 mg, 2.2 mmol) and DCM (15 mL) was cooled to 0 °C and diethyl chlorophosphate (345 mg, 2 mmol) added carefully over a period of 5 minutes. The reaction was allowed to warm to ambient temperature and stirred for 2 hours. The reaction was quenched by addition of sat. NaHCO 3(aq) (10 mL) and the product extracted into DCM (3 x 10 mL). The combined extracts were concentrated in vacuo, then purified by flash column chromatography (0-5 % MeOH in EtOAc) to afford the title compound as a pale yellow oil.  The polarization transfer experiments that are reported were conducted in either an NMR tube that was equipped with a Young's Tap (Method A) or using an automated polarizer (Method B).

Method A
Samples for these polarization transfer experiments were based on a 5 mM solution of [IrCl(COD)(IMes)] (1) and substrate (L 1 -L 12 ) in methanol-d 4 (0.6 mL). The samples were degassed prior to the introduction of parahydrogen at a pressure of 3 bar. Samples were then shaken for 10 s in the specified fringe field of an NMR spectrometer before being rapidly transported into the magnet for subsequent interrogation by NMR spectroscopy.
A schematic representation of the polarizer used for flow measurements is shown in Figure s1. The Mixing Chamber (MC) is housed within a tuneable copper coil (0.5 to ±150 G). The coil was situated in a magnetic field which has the components x 4.9 -5.1 G, y 3.3 -3.6 G and z 1.5 -2.1 G. All magnitudes of the magnetic fields in which polarization transfer occurs (PTF) are stated without correction for this local field). The MC houses the solvent, catalyst and substrate. Liquid and gas flow is computer-controlled via the pulse program. As such, the system is entirely automated.
Parahydrogen is introduced into the MC first to activate the catalyst. Nitrogen gas is used to shuttle the hyperpolarized solution from the MC to the NMR probehead for measurement. The transportation time was calibrated to 2.9 s. A further delay of 0.5 s was allowed for settling of the sample prior to signal acquisition (1 s).

Figure s1
: Schematic representation of the polarizer, the hyperpolarization process and its subsequent NMR analysis.

-1 H and 31 P Enhancement Factors
For calculation of the enhancement of 1 H and 31 P NMR signals the following formula was used:

E = SI(pol) SI(unpol)
Where, E = enhancement, SI(pol) = signal of polarized sample, SI(unpol) = signal of unpolarized (reference) sample. Experimentally, both spectra were recorded on the same sample using using identical acquisition parameters, including the receiver gain. The raw integrals of the relevant resonances in the polarized and unpolarized spectra were then used to determine the enhancement levels. For all of the 31 P measurements, the FID was processed under magnitude calculation mode to aid with the integration process.

-Determination of ligand exchange rates
The ligand exchange pathways were evaluated by 1D-EXSY spectroscopy using a reported method. [11][12] We follow magnetization transfer from the ligand resonance position as a function of the sequences evolution time. This results in the collection of data that encodes the reaction which is then analysed by a differential procedure to extract the listed rate constants. 11-12

-1 H and 31 P OPSY spectra
The generation of double quantum (DQ) coherence between 1 H and 31 P nuclei after polarization transfer via SABRE was confirmed using the OPSY protocol. Gradient ratios of 22.9 : 80 and 57.1 : 80 were used to selectively detect 31 P and 1 H signals arising from a DQ coherence pathway ( Figure S6). Figure s6: NMR spectra of a sample consisting of 3,5-diphosphorylpyridine (5 eq.) and IrCl(COD)(IMes) (1.92 mg) in methanol-d 4 (0.6 mL) using the OPSY sequence to select the double quantum coherence between 1 H and 31 P on either A) the 31 P channel or B) the 1 H channel. Figure s6A possesses only a single peak, which is the 31 P resonance of free 3,5-diphosphorylpyridine. Resonances for the bound 3,5-diphosphorylpyridine are not observed. The 1 H spectrum ( Figure S6B), is more complicated. The free and bound aromatic resonances are observed; the free signals are significantly more intense than those of the bound peaks. Two smaller peaks are observed for the free 1 H resonances of the ethyl group. These peaks have approximately 1 % of the signal area of the aromatic resonance at δ 9.13 ppm. Thus, significant DQ coherence is formed between the 31 P nucleus and a 1 H nucleus of the aromatic ring. There is only a minor contribution to observed DQ coherence between the 1 H nuclei in the ethyl chains and the 31 P nucleus.
Further measurements were conducted at the Earth's magnetic field to investigate any triple quantum (TQ) coherence that was formed between either 2 1 H and 1 31 P nuclei or 2 31 P and 1 1 H nucleus. 1 H and 31 P NMR spectra were collected using gradient ratios to probe both these combinations. In each case, despite the 1 H NMR spectra possessing a signal, the corresponding 31 P NMR spectrum proved sufficiently insensitive to allow detection. The gradient ratios used were 33.3 : 80 to select 2 1 H and 1 31 P nuclei and 44.4 : 80 to select 1 1 H and 2 31 P nuclei.

Kinetic behaviour of 2a-L 4
We monitored the hydride ligand exchange processes shown by 2a-L 4 at 283 K. 2a-L 4 proved to exchange into a methanol-adduct that yielded hydride signals at δ −24.98 and δ −23.14. A role for such complexes in the H 2 exchange pathway of [Ir(H) 2 (IMes)(py) 3 ] + has been documented. 13 The rate of H 2 loss from 2a-L 4 was determined to be 0.028 ± 0.007 s -1 at 283 K. The rate at which 2a-L 4 converted into the corresponding methanol adduct 2b-L 4 was estimated to be 0.44 ± 0.014 s -1 and this species proved to undergo H 2 loss at a rate of 0.887 ± 0.075 s -1 . We also followed the transfer of magnetization from bound diethyl 4-pyridylphosphonate (L 4 ), trans to hydride, into the free substrate. The corresponding rate was 0.522 ± 0.007 s -1 at 283 K with magnetization transfer proceeding into the methanol adduct at a rate of 0.125 ± 0.003 s -1 . We note that axial diethyl 4pyridylphosphonate (L 4 ) loss is also evident at a rate of 0.645 ± 0.010 s -1 with in-cage recombination suggested because of the equilibrium position evident in Figure s9. The raw data for these processes and the corresponding fitting curves are presented in Tables s3 -s5 and Figures s7 -s9 respectively.

-Reaction with 3,5-bis(diethoxyphosphoryl)pyridine (L 7 ).
A sample containing 3,5-bis(diethoxyphosphoryl)pyridine (5 eq.) and IrCl(COD)(IMes) (1.92 mg) in methanol-d 4 (0.6 mL) at 253 K was used to obtain the NMR information shown in Table s6. We  Kinetic behaviour of 2c-L 7 2c-L 7 was also shown to undergo both H 2 loss and hydride site interchange by EXSY. Tables s7 and s8 and figures s10 and s11 present the corresponding raw and fitted data. The rate of H 2 loss from 2c-L 7 at 273 K is now just 0.053 ± 0.009 s -1 and the rate of hydride site interchange is 0.24 ± 0.07 s -1 . We note that the high error associate with the hydride interchange pathway arises from the differential line-widths and hence relaxation times of the hydride resonances.

-Spin-Lattice Relaxation Times
Appropriate T 1 values for specific materials are listed in Table s8.

-Polarization Transfer methods
Samples for these polarization transfer experiments were conducted using a 5 mM solution of [IrCl(COD)(IMes)] (1) and 5 or 6 eq. of substrate (L 4 -L 7 ) in methanol-d 4 (3.0 mL) and placed in a 10 mm NMR tube fitted with a J. Young's tap. The samples were degassed prior to the introduction of parahydrogen at a pressure of 3 bar. Samples were then shaken for 10 s in a specified fringe field of an MRI scanner before being rapidly transported to the magnet for interrogation. The MRI experiments were completed using a RARE pulse sequence with an echo train length of 32 and a matrix size of 32 x 32 that was zero filled to 128 x 128. The hyperpolarised images were acquired as a single shot experiment whereas the thermal images took 2048 averages to get a S: