Assessing the validity of NMR relaxation rates obtained from coarse-grained simulations

NMR relaxometry is a powerful and well-established experimental approach to characterize dynamic processes in soft matter systems. All-atom (AA) resolved simulations are typically employed to gain further microscopic insights while reproducing the relaxation rates R1 . However, such approaches are limited to time and length-scales that hinder modeling of systems like long polymer chains or hydrogels. Coarse-graining (CG) can overcome this barrier at the cost of loosing atomistic details that impede the calculation of NMR relaxation rates. Here, we address this issue by systematic characterization of R1 while performing systematic measurements on a PEG-H2O mixture at two different levels of details: AA and CG. Remarkably, we show that NMR relaxation rates R1 obtained at the CG level obey the same trends when compared to AA calculations, but with a systematic offset. This offset is due to, on the one hand, the lack of an intra-monomer component and, on the other hand, the inexact positioning of the spin carriers. We show that the offset can be corrected for quantitatively by reconstructing a posteriori the atomistic details for the CG trajectories.