Since the last 15 years MAS-DNP has developed into an important research tool in the structural investigation of materials and surface science. The sensitivity gain is provided by highly polarized electron spins transferring polarization to nuclear spins upon microwave irradiation. The polarizing agent (PA) is represented typically by exogenous radicals with the gold-standard bis-nitroxides AMUPol  and TEKPol  leading to 1H enhancements of around 250 at 9.4 T and 100 K.
Here we show that the, so far overlooked, local conformation around the N-O● region can dramatically impact the MAS-DNP performance, with the new synthesised HydrOPol returning a 1H enhancement of 330 at 9.4 T and 100 K: about 30% better enhancement than the current standard in the field.
However, a major limitation is that nitroxides are not always compatible with sample formulations, loosing for example the capability to yield any enhancement in reducing environments as encountered in cells or in many catalysts. We show as an example that AMUPol fails to yield any enhancement on a solution of ascorbic acid. A possible way to overcome the problem is using alternative Gd(III)-based PAs.
By selecting the chelating ligand in [Gd(tpatcn)] we can polarize ascorbic acid and achieve a maximum 1H enhancement of 37 at 9.4 T and 100 K (a factor 2 better than [Gd(dota)(H2O)]-. Moreover, by using a simplified theoretical model tested on different Gd(III) complexes we show that the reduction of zero-field splitting impacts quadratically the expected MAS-DNP signal enhancement, therefore establishing an important design parameter.
Overall, by rational design of either the local geometry around the N-O● region in bis-nitroxides or of the chelating ligand in Gd(III)-systems, we show that HydrOPol and [Gd(tpatcn)] currently yield the best reported enhancements at 9.4 T and 100 K in their own class of PAs.