4.1. Sample Preparation

The preparation and characterization of the mesoporous silica materials was done in previous works [36,87]. Briefly, MCM-41 with a pore diameter of d=2.8 nm was synthesized by the Buntkowsky group (Chemistry Department, TU Darmstadt) and SBA-15 with a pore diameter of d=5.4nm was purchased from Sigma-Aldrich. Pore volumes and diameters were determined by nitrogen gas adsorption. Scanning electron microscopy showed that the MCM-41 particles had nearly spherical shapes and particle sizes of ∼400 nm [36,85,87]. Lysozyme from chicken egg white, ficoll (Ficoll PM 70, M=70kDa), H2O, D2O, DMSO-d6, and DMSO-h6 were purchased from Sigma-Aldrich. All chemicals were used as received. However, spurious amounts of water were removed from the silica and lysozyme matrices in high vacuum prior to use [77].

The sample compositions were assured by weighing the components. Mesoporous silica were completely loaded with the aqueous DMSO solutions. In doing so, the knowledge of the specific pore volume was used to determine the amount of solution required for ∼100 % filling. The ficoll- and lysozyme-based materials were prepared by carefully mixing the components. The solvation level of lysozyme was chosen to correspond to a fraction of 1.6 solvent molecules per amino acid, independent of the composition of the solution. This criterion resulted in solvation levels of s=0.68gsol/glys and s=0.97gsol/glys for the studied D2O-DMSO-h6 solutions with water mole fractions of 66 and 33 mol%, respectively. For all samples, we observed shelf-times of several days to assure equilibration prior to the measurements. All NMR samples were sealed in the NMR tubes to avoid solvent losses during the long-lasting NMR measurements.