Nanomaterials have become increasingly important to study as their applications have expanded from heterogeneous catalysis, medicinal therapeutics, optoelectronics, and more. Understanding how these materials function is of utmost important to developing the next generation of functional nanoparticles. While there are various methods of characterizing nanoparticles using spectroscopic techniques such as infrared and Raman, solution state nuclear magnetic resonance has not seen the same type of methodological advancement.

Large molecular systems are difficult to study by NMR. Signal relaxation occurs faster as the molecule of interest has a slower solution tumbling rate. The result for the relatively large nanoparticle system are spectral lines that are broadened to the point of invisibility. Direct observation of spins is not helpful in these cases. Instead, surface contrast NMR experiments are proposed for the indirect detection of ligands that adsorb and desorb from a nanoparticle surface. The effect of signal broadening is used as an amplification factor to measure lowly populated surface bound ligand states.

An important aspect of performing NMR experiments on a sample that contains nanoparticles is that the nanoparticles must remain in suspension for the duration of the experiment. Nanoparticles typically are not soluble and tend to aggregate and sediment quickly. The use of the hydrogel, agarose, is proposed to trap nanoparticles within the pores of the gel to keep these materials homogeneously distributed for extended periods of time. We show that agarose is an inert matrix that does not participate in the ligand exchange process and does not interfere with data quality. The exchange processes of cholic acid and phenol onto the surface of ceria nanocubes were used as a model system. 1H dark state exchange saturation transfer (DEST) and relaxation dispersion experiments were performed and data analyzed using the Bloch- McConnell equations. Cholic acid was found to be involved in a two site exchange process comprised of a surface bound and unbound state. Phenol was found to participate in a three site exchange; unbound, a weakly associated state characterized by a high degree of rotational freedom, and a tightly associated surface bound state characterized by dramatically decreased rotational freedom. Accessibility to the tightly associated state was found to be only allowed by first passing through the weakly associated intermediate state.

The hydrogenation of phenol can be achieved using H2 and Pd supported on ceria under mild conditions with high conversion rates. However, the role of the ceria support, required in this catalytic system, is not fully understood. Ceria possesses exceptional redox properties but a full description of the interactions between phenol and ceria remains unknown. Here, we further our surface contrast NMR methodologies with the addition of 13C relaxation (R1 and R2) and DEST measurements which allows for ligand dynamics modelling. We found that phenol interacts with the ceria surface in two binding modes consistent with hydrogen bonding and a covalent interaction with an oxygen defect site in the crystal lattice. As alluded to in our previous study, the tighter covalent interaction is only accessible by passing through the weak hydrogen bonding intermediate first. Additionally, phenol can bind to Pd in a flat conformation again passing through a hydrogen bonding to ceria intermediate. Adding phosphate into mixture perturbs the interactions between phenol hydrogen bonding to ceria but not to Pd. The corresponding decrease to the catalytic conversion suggests that the ability for phenol to hydrogen bond to the ceria surface plays an important role in this catalytic reaction.

The use of agarose gels in the previous studies described here have been extremely important for the successful, practical application of NMR experimentation to ligand- nanoparticle systems. However, agarose will only gel with an aqueous solvent which severely limits the scope of this methodology to systems that use a non-aqueous medium. Several gels that are compatible with organic solvents were characterized for their use in trapping nanoparticles from sedimentation. These included agarose, polystyrene, polydimethylsilicone, ((4,6-O-Benzylidene)-Methyl-𝛼-D-Glucopyranoside), methylcellulose, and polystearylacrylate. Six basic criteria were used to qualify important characteristics of the gels, namely low residual NMR signal, low internal viscosity, large pore size, macroscopic integrity, solvent compatibility, and ideal preparatory conditions. Overall, polystearylacrylate appears to be the most suitable for application to organic solvent based surface contrast NMR studies as it adequately satisfies all diagnostic criteria. Methylcellulose is another strong candidate, especially for its ease of preparation, but its application is limited as it will only gel using DMF as a solvent.

In addition to the project described above, I have collaborated on a number of other projects applying NMR to solve research questions. I have developed a software tool for fitting alignment tensors of proteins members in an ensemble to residual dipolar coupling data. This was used for defining a structural ensemble for AlkBH5 to better understand the origin of its catalytic activity and selectivity. I have characterized sugar composition of batches of alginate to aid in determining the chemical basis for its optical properties. This investigation was used to solve a reliability issue in the production of transparent soil. I have also contributed to characterizing a new ionic liquid formed from hydrogen sulfide and oleylamine. This ionic liquid was used as a ‘green’ sulfur precursor in the synthesis of metal sulfide nanoparticles.