Matrix-Assisted NMR

During the last decades, the interest of chemistry toward increasingly sophisticated processes has grown exponentially. As a consequence, the evolution of the systems under investigation has been necessarily paired with the development of modern methodologies capable of handling the enormous amount of data stemming from samples of great complexity. Among the many examples in the literature, one of the biggest ongoing challenges is the analysis of mixtures, from reaction crude extracts to biological fluids like blood and urine. Indeed, chromatography has been - and still remains - one of the primary methods adopted to reduce the complexity of a multi-analyte system. Nonetheless, one intrinsic problem of the chromatographic approach is its inability to identify unknown molecules, and hyphenated techniques (mostly based on mass spectroscopy) have been developed just to overcome this stumbling block. On the other hand, Nuclear Magnetic Resonance (NMR) spectroscopy is one of the most powerful techniques for the investigation of organic compounds. NMR exploits an intrinsic property exhibited by some atomic nuclei -- the spin -- to acquire chemical and structural information through well-established experimental protocols, known as pulse sequences. In particular, solution-state NMR can boast a vast ensemble of procedures aimed at collecting detailed data about through bond connectivities (COSY, TOCSY, HSQC,...) or through space proximities (NOESY, ROESY,...). All these information are nothing less than fundamental for the structure determination of unknown compounds. Even tough this makes NMR spectroscopy largely appealing, the acquisition of such extensive information ultimately translates into detecting many signals at once, so that spectra interpretation can become a very challenging task. This is especially true when observing 1H resonances, which display a small dispersion in the frequency domain (about 12 ppm) and spectral crowding becomes consequently a serious problem. Not surprisingly, the situation becomes almost unmanageable when NMR is applied to the assay of mixtures, where the superposition of signals stemming from different species is virtually assured. Certainly, multidimensional NMR techniques can be useful for the interpretation of crowded single-molecule spectra, but they rapidly loose all their advantages as the number of components in the sample increases. As for chromatography, the advent of hybrid techniques like LC-NMR, where LC stands for Liquid Chromatography, has partly circumvented the aforementioned difficulties, yet at the cost of an expensive and dedicated instrumentation. In the context of mixture analysis, matrix-assisted NMR methodologies stand as an alternative to the various hyphenated techniques. They rely on the combination of NMR spectroscopy and an external agent added to the sample, which can be either a molecular or macromolecular species, or even a mesoscopic matrix. The aim of such matrices is to differentiate the signals of the various components, favouring their detection and characterisation. The present work is divided into three independent parts. The first two are dedicated to different subjects of matrix-assisted NMR. In particular, Part I is aimed at the understanding of the physical phenomena underlying signal broadening when a solid, stationary phase is used in Matrix-Assisted Diffusometry (MAD) NMR measurements. Part II focuses on nanoparticle-assisted NMR chemosensing, a technique where monolayer-protected gold nanoparticles are exploited to transfer magnetization to selected classes of analytes by means of the Nuclear Overhauser Effect. In this second part, different nanoparticle-assisted methodologies are presented and analysed, alongside with some strategies aimed at the enhancement of the sensitivity. Part III concerns the complete 1H-NMR characterisation of the atomically precise Au38(SBut)24 gold nanocluster, which can be considered as a prototypical nanoparticle. The Au38 core features four different symmetry-unique and equally populated binding sites for the grafting of the ligands that constitute the coating monolayer. Each binding site shows a distinct pattern of resonances, so that the overall 1H-NMR spectrum of the cluster is the result of the superposition of four independent subspectra. In this case, the full characterisation of the spectrum has been achieved through a combined NMR-MD (Molecular Dynamics) analysis