Carbohydrates are ubiquitous in nature and exhibit a multitude of roles. Besides nucleic and amino acids, they can be regarded as the third alphabet of life. They are used as energy source to fuel the cells, as structural building blocks and play a key role in cellular recognition processes. Compared to the other two groups of biomacromolecules, carbohydrates display a higher level of structural complexity by virtue of the number of individual monosaccharide building blocks, as well as the greater number of possibilities of connecting them and additional modifications. This renders a high information content and a good understanding of the structure-function relationship of glycans is important, since the presence or absence of specific structures can make the difference between health and disease.

Carbohydrate structures can be characterized and studied by NMR spectroscopy at the atomic level. This process is time-consuming and error-prone, due to the narrow spectral window, in which most carbohydrate resonances are located leading to severe spectral overlap. Computer programs have been developed, aiding this process. This thesis investigates the quality of prediction of NMR chemical shifts of glycopeptides, highly branched oligosaccharide structures and those bearing a non-natural organic aglycone at the reducing end, as well as the automated determination of primary carbohydrate structures from unassigned NMR spectroscopic data thereof. Novel developments of the CASPER program are highlighted.

The three-dimensional structure of carbohydrates plays an important role during carbohydrate-protein interactions. This thesis investigates the conformational preferences and dynamics of glycan structures ranging from di- to tetrasaccharides. A particular focus lies on the measurement of transglycosidic ³J_CH coupling constants by NMR. Furthermore, the experimental spectroscopic data is compared to results from MD simulations.

Synthetic carbohydrate chemistry has a strong focus on stereoselective C−O bond formation for the synthesis of oligo- and polysaccharides. Each glycosylation reaction can produce two stereoisomeric structures. To date, the mechanistic pathway of glycosylation reactions is still not fully understood, since many different parameters influence the stereoselectivity. A combined experimental and computational study exploring the role of the solvent is presented and a linear correlation of the selectivity with a solvatochromic parameter for the polarizability of the solvent was found.