From Computation to Experiment: Predicting Radical Performance in Overhauser Effect DNP NMR

Dynamic nuclear polarization (DNP) is a technique used to enhance the sensitivity of NMR spectroscopy. DNP can proceed through one of three main mechanisms, although most research relies on the cross-effect mechanism due to the relative ease of designing highly performant radical polarizing agents. Another promising mechanism is the Overhauser effect, which requires only allowed EPR transitions. However, it also depends on the presence of fast molecular motions that modulate electron–nuclear couplings at the correct frequency—an aspect that is difficult to engineer.

In a recent article featured on the cover of the Journal of Physical Chemistry Letters, it was shown that the efficiency of the Overhauser effect, in radicals where it is driven by a methyl rotation, can be quantitatively predicted in silico. DFT calculations, carried out with ADF, were used to determine the strength of hyperfine couplings as a function of the methyl phase, as well as the associated rotation barriers. These values were then employed to predict electron–nuclear cross-relaxation rates, which were subsequently compared with experimental DNP enhancements measured in five CF₃-functionalized Blatter-type radicals. This work demonstrates that computational methods can serve as a predictive tool for evaluating a radical’s Overhauser effect DNP performance before the investment of time and resources into its synthesis.