Degree Type


Date of Award


Degree Name

Doctor of Philosophy





First Advisor

Aaron J Rossini


This dissertation describes the development of sensitivity-enhanced solid-state nuclear magnetic resonance (NMR) spectroscopy techniques for the characterization of inorganic semiconductors. Inorganic semiconductors will play a vital role in the development of next generation energy materials and understanding the structure of these materials will enable the development of better performing devices. Inorganic semiconductors synthesized as nanomaterials are of interest because of increased tuneability of the band gap through modifications in the size, surface ligands or composition of the particles. Ultimately, the surface structure of nanomaterials will greatly affect their optoelectronic properties, making it vital to obtain a molecular-level understanding of the surface structure. Solid-state NMR spectroscopy is an ideal characterization method for determining the surface structure of nanomaterials because it probes the local chemical environments of the surface. However, surface solid-state NMR experiments are often challenging to obtain because of the low concentration of surface sites and the use sensitivity enhanced NMR methods such as proton detection or dynamic nuclear polarization (DNP) is needed.

Chapter 2 focuses on the development of fast magic angle spinning (MAS) and indirect detection methods to characterize the surface structure of silicon nanocrystals (NCs). Using 2D dipolar and scalar 1H{29Si} heteronuclear correlations (HETCOR) solid-state NMR spectra were used to reveal the surface structure of Si NCs and identify the different surface silicon environments. Using the 1H{29Si} refocused insensitive nuclei enhanced by polarization transfer (INEPT) experiments it is possible to differentiate the NMR signals associated with different surface hydride species. Varying the 29Si evolution time in the INEPT experiment and fitting the oscillation of the NMR signal allows for the determination of the relative populations of the different surface hydride species. From the refocused INEPT experiments, it was determined that the monohydride species is the predominate surface hydride in the hydride terminated Si NCs. When the hydride terminated Si NCs were functionalized with dodecyl groups a reduction in the relative populations of di- and trihydrides was observed, which is consistent with the hypothesis that the trihydride group is primarily responsible for initiating the surface functionalization process.

The work in Chapter 3 builds off of the work presented in chapter 2 in order to probe the location of phosphorus doped into Si NCs (P:Si NCs). When the phosphorus is doped into the Si NCs, surface phosphine groups (*PH and *PH2) are formed. These groups make up 20% of the surface functional groups in the hydride terminated P:Si NCs (P:Si-H NCs). Through analysis of the dipolar and scalar 1H{31P} HETCOR interactions it was determined that the majority of the phosphorus that was doped into the Si NCs is located at the surface of the P:Si NCs. 29Si{31P} scalar correlations were used to confirm that all of the phosphorus present within the P:Si NCs is chemically bonded to the silicon.

Chapter 4 focuses on improving the sample preparation of nanoparticles (NPs) for DNP surface enhanced NMR spectroscopy (SENS) experiments. Through modifications of the established of NPs for DNP SENS experiments, the overall NMR sensitivity substantially improved. This was achieved by changing the support material from mesoporous silica to hexagonal boron nitride, the latter of which has more favorable dielectric properties for the propagation of the microwaves through the sample. In addition to changing the support material the concentration of the NPs was increased by using a precipitated powder of the NPs instead of a colloidal solution of NPs. The new DNP NP sample preparation method enabled the acquisition of challenging homonuclear and heteronuclear correlations of CdS, Si and Cd3P2 NPs. Including a 2D 13C-29Si HETCOR of functionalized Si NPs at natural abundance.

In Chapter 5 the improved DNP NP sample preparation method was used to perform DNP SENS experiments on cadmium-treated indium phosphide quantum dots (Cd-InP QDs), in order to ascertain the location of the Cd in the InP QDs. To determine the location of Cd in Cd-InP QDs, DNP SENS experiments on InP QDs, Cd3P2 QDs and InP magic size clusters (MSCs) with various amounts of cadmium. Using 1D DNP-enhanced 31P cross-polarization magic angle spinning (CPMAS), 113Cd CPMAS NMR spectra, 113Cd{31P} rotational echo double resonance (REDOR) and 31P{113Cd} dipolar heteronuclear multiple quantum correlation (D-HMQC) experiments it was determined that the majority of the Cd atoms are located at the surface of the Cd-InP QDs, and Cd is primarily coordinated to the surface phosphate groups.

In Chapter 6 the sensitivity of 207Pb solid-state NMR of organolead halide perovskites was examined. The sensitivities of spectra obtained using fast MAS and DNP techniques were quantitatively compared to determine which NMR method provided the optimal 207Pb sensitivity. Using fast MAS and proton detection, 1H{207Pb} HETCOR spectra were obtain in less than a half-hour with less than 5 μL of the perovskite. This was achieved by modifying the double CP HETCOR experiment to one where the 207Pb spins were directly excited and then the magnetization was transferred to the protons for detection. In the organolead halide perovskites, the spin lattice relaxation time (T1) of the 207Pb spins at 50 kHz was several orders of magnitude smaller than the 1H T1, which allows for more efficient relaxation using the modified method. Alternatively, relayed-DNP experiments with modest DNP enhancements (1–20) and cryogenic temperatures (<110 K) provide a large boost in the sensitivity, allowing for the 207Pb NMR spectra of the organolead halide perovskites to be obtained in several minutes. Consequently, with the boost in the NMR sensitivity it was possible to obtain a 207Pb solid-state NMR spectrum of a 300 nm model thin film of the CH3NH3PbI3 perovskite in 34 hours. This spectrum would be unattainable through conventional NMR spectroscopic methods.


Copyright Owner

Michael Patrick Hanrahan



File Format


File Size

308 pages

Available for download on Friday, January 07, 2022