Speaker
Description
Signal Amplification By Reversible Exchange (SABRE) is a hyperpolarization method that generates large, non-equilibrium spin polarizations by transferring spin order from parahydrogen ($|S_H ⟩$) to magnetized states on a target nucleus ($|α_L ⟩$). An iridium-based catalyst simultaneously and reversibly binds parahydrogen and a target ligand with spin-1/2 nuclear target(s) L. Under the right field conditions, couplings then facilitate flow of spin order out of an initially overpopulated singlet hydrogen state into ligand polarization. The most common initial SABRE approach uses low fields, which match energy separations to transfer magnetization at an avoided crossing. However, in the heteronuclear case, the optimal experimental continuous field is ≈-0.5 μT, while the avoided crossing matching condition is ≈-0.05 μT.
While this traditional framework can provide a general understanding of SABRE dynamics, our lab recently introduced several field sequences which never approach a matching condition instantaneously or on average, but still lead to an improvement in the polarization yield. This suggests that an additional degree of freedom, unique to low field experiments might be valuable: the ability to rapidly and fully manipulate fields in three dimensions. However, theoretical assessment is challenging because many terms in the Hamiltonian are comparably sized, and off-diagonal elements are on the same order as the difference in the energy levels. This is particularly confounding in SABRE, where chemical exchange rates are also comparable to J couplings and resonance frequency differences.
We report here two different multiaxial pulse sequence approaches, which improve SABRE spin population transfer to magnetized and singlet spin states. The first approach uses simple circularly polarized transverse irradiation at a field strength equivalent to the leading field to enhance polarization transfer. The second approach, Multi Axis Computer-aided HEteronuclear Transfer Enhancement (MACHETE) moved away from assuming a traditional pulse sequence structure and uses an evolutionary strategy to optimize the pulse shape in three-dimensions. This strategy has yielded a pulse sequence that shows a 7.5-fold experimental improvement in the polarization yield under SABRE hyperpolarization. This gain is highly nonintuitive as transverse fields do not inherently preserve longitudinal magnetization, but it is robust to a wide range of exchange rates and SABRE complex geometries. An average Hamiltonian approach turns out to be insightful. The eigenvectors of the average Hamiltonian show isolation of states such as $|S_H β_L ⟩$ and protection of magnetization on the unbound ligand, but substantial mixing of the $|S_H α_L ⟩$ spin state with states such as $|T_H^+ β_L ⟩$, thus building up spin-down magnetization on the ligand(s). These waveforms, compatible with any 3-channel AWG and a simple multiaxial electromagnet array provide a new strategy for understanding polarization transfer and optimizing population transfer into both magnetized and singlet states.
