The NV defect center in diamond forms a pseudo-atomic quantum system with discreet optically excitable transitions between ground and excited states in the gap between valence and conduction bands, making the NV center a deep-level defect center in diamond. For the negatively charged NV center (NV-), both the ground and excited states are spin triplets (S=1) and coupling between optical and spin states provides unique opportunities for optical detection of magnetic resonance (ODMR) and quantum metrology under ambient conditions. As an atomically sized point defect that is stable in single nanodiamonds as small as a few nanometers, the NV center enables spin based imaging of temperature, electric fields, magnetic fields and mechanical strain with nanoscale spatial resolution. Made of carbon, nanodiamonds are also ideal nanoparticles for use in biological systems, exhibiting extremely low cytotoxicity, no photo bleaching and exceptional contrast in transmission electron and light microscopy. With the ability to act as both a passive, non perturbing sensor (e.g. a diamagnetic material used for sensing magnetic fields) and a connection point for two way, direct coupling of far field optics to NV$^-$ spins and indirectly to the highly localized neighborhood of atoms, electronic states and phonon modes in the local vicinity, NV diamond provides a route to nanoscale field sensing and spin based optical microscopy. These sensors are also stable in nanodiamonds. For nanoparticles, the reduction in particle diameter dramatically enhances the proportion of the material that is surface exposed, enhancing any surface related properties. Nitrogen Vacancy centers in nanodiamonds provide another example of material properties that vary with particle size. In chapter 5, evidence that modulation of nanodiamonds fluorescent emissions by electrical perturbations in their surrouding environment allows transduction of local electrodynamics into a far field optical signal capable of mapping action potentials in cardiomyocytes in presented. The unifying theme for all of the work presented in this dissertation is to enable the use of magnetic resonance signals and interactions for imaging and sensing in previously inaccessible regimes. The specific focus is on methods and techniques to enable nanoscale physical sensing and functional imaging in live tissues, with the ultimate aim to lay the predicate for the combination of both without compromise.

Objective: Three dimensional cell and tissue culture outcomes depend on mechanical forcesin their environment and sufficient delivery of nutrients and removal of waste. Flow and shear influence parenchymal cell distribution and vascular morphogenesis. Herein a bioreactor which can control fluid flow and shear patterns over large areas was developed as a tool for tissue engineering. Methods: A 12-inlet perfusion bioreactor was constructed to be compatible with multiple types of tissue engineering scaffolds including biopolymer hydrogels and macroporous sponges and can support the growth and viability of thick tissue scaffolds. Scaffolds containing mammalian cells cultured under flow were compared to static conditions and other flow patterns. Results: Flow patterns in the scaffolds were obtained using MRI flow maps. Hydrogels and commercial sponges were assessed for 3D culture compatibility and Gel 2 (a specific ECM hydrogel mixture), Collagen, and Gelatin sponges were found to be both optimal for cell growth and compatible with perfusion flow. Culture experiments determined the feasibility of our bioreactor in supporting thick tissue growth and viability and various flow patterns were shown to alter cell distribution, growth, and viability. Conclusion: We demonstrated the ability of the bioreactor to implement macroscopic, tunable flow patterns in a biocompatible environment suitable for cell and tissue culture and found distinct differences in tissue development from different flow patterns. Significance: The bioreactor will enable future studies on different flow patterns and their effects on cell growth and viability for medically relevant tissues and could inform on culture conditions that better mimic what occurs in vivo during tissue development.

The Gd element and other rare-earth metals are notoriously difficult to use in chemical synthesis due to their high reduction potentials and aggressive reactivities with ambient oxygen, which almost always leads to the formation of oxides. The challenge in chemical synthesis limits the range of applications for rare-earth metals. The most important use of Gd at the moment is nanocrystals for technological applications. Herein, we report for the first time the successful production of size-controllable, solid core-shell oxide-free Gd metal nanocrystals. We have solved the long-standing problem of oxidation through a reduction process and appropriate capping. The manuscript describes the procedure and detailed characterizations of the process to ensure the highest quality of the produced particles. In particular, the nanocrystals displayed the largest saturation magnetization observed to date for nanocrystalline Gd metal. This value (206 emu/g at 2K) currently stands as the world record.

Another important application of Gd is contrast agent development for MRI. To this end, we have performed NMR relaxivity measurements to evaluate the performance of the nanoparticles as potential MRI contrast agents. Typically, Gd based nanoconstructs such as FDA approved Gd-DTPA chelates are T1 MRI contrast agents; however, we demonstrate, for the first time, that pure Gd nanoparticles can also be used as state-of-the art T2 contrast agents. World record high values for transverse proton relaxitivity (r2 of 232 mM-1s-1 per Gd atom and per-particle relaxivity (2.9 x 10^8 mM-1s-1 have been obtained, exceeding the current highest per-particle r2 values. These results make our Gd nanocrystals the most promising MRI contrast agents for use in biomedical applications. For the first time, this puts MRI on par with positron emission tomography in terms of sensitivity to detection of a contrast agent.

We further developed the nanoparticles and we demonstrate record high saturation magnetizations for Gd nanoparticles, namely, 226 emu/g at 5 T. This magnetization is substantially higher than anything achieved to date. We have achieved such high magnetizations in a reliable and reproducible manner by controlling the crystallinity of the grown Gd nanofilm. The crystallinity of Gd is found to play an important role in the observed magnetization values. The higher magnetization is observed for nanoparticles that have a lower content of paramagnetic face-centered cubic (fcc) phase and greater content of ferromagnetic hexagonal close-packed (hcp) phase. Control over fcc and hcp content in the lattice was achieved by adjusting the deposition rate of Gd metal during the nanofabrication process. Our results indicate the remarkable influence of nanocrystallinity on the magnetism of Gd and the ability to control it.

Our novel fabrication technique, which overcomes the problems of current synthetic approaches to rare-earth nanoparticle synthesis through the careful optimization of capping and hydrogen reduction techniques, can also be applied to other rare-earth metals and alloys. This opens the door to fundamental studies on these materials at the nanoscale. It will also enable the realization of the full potential of rare-earth metals in industry.

Superconducting transmon qubits are a promising architecture for large-scale fault-tolerantcomputation: transmons possess favorable properties for scalability, and mature microwave control electronics are readily available. Recent advances in coherent control of these devices have produced optimally calibrated microwave control pulses with fidelities approaching fault-tolerance thresholds. However, miscalibrations and drift in the amplitude or frequency of the microwave drive may result in suboptimal gate performance. Compensation schemes employing open-loop optimal control to increase robustness toward these types of frequency detuning and field amplitude errors have been under development in the NMR community since the 1970s. Composite and adiabatic pulses are two such robust pulse design techniques that may benefit coherent control of superconducting architectures.

In this thesis, we demonstrate applications of composite and adiabatic pulses for theimplementation of robust inversion gates on a transmon qubit. We characterized the robustness, fidelities, seepage and leakage rates of the inversion pulses using simulated and experimental transition probabilities and randomized benchmarking methods. Both adiabatic and composite pulses were able to compensate for a broader range of systematic drive amplitude and off-resonance errors compared to standard gates. Several composite pulse schemes also improved on-resonance fidelity.

This thesis aims to improve on the current classification capabilities of deep neural networks on two types of radio-frequency data: radar and OFDM packets. In radar, applying neu- ral networks to Automatic Target Recognition problems is a well-developed field, especially using the MSTAR database. However, existing state-of-the-art methods require precise pre- conditioning of radar data and are unsuited to applications with a large number of radar target classes. Therefore, we asked whether distributed learning can increase the generaliz- ability and scalability of neural networks in these tasks. To test this, we applied distributed learning via Multi-Stage Training and a new network architecture, the Convolutional Multi- Stage Network, to provide a scalable, generalized treatment of radar data for more practical applications. This method was shown to outperform traditional neural network architectures on a new radar dataset. A similar approach was applied to the OFDM data with the goal of identifying specific radio-frequency transmitters for network security purposes. The task of identifying OFDM packet transmitters has previously been performed successfully, though with precise data collection methods. Data collection methods on a live network will likely include imperfect recording times, so we sought to improve network robustness to time- shifted OFDM packets. It was shown that the Convolutional Multi-Stage Network improved robustness to time-shifting of the radio-frequency data over the Multi-Stage Network, which was the previous-best method. Simple preconditioning of the data using variations of the dis- crete wavelet transform further improved robustness to time-shifting of the radio-frequency data using both network architectures. These results are significant, as they provide a new avenue for applying neural networks to radio-frequency in difficult, real-world applications.

Despite recent successes in cancer immunotherapies, challenges including insufficient activation and suppression of effector T cells in the tumor microenvironment, and inadequate ability to expand tumor-specific T cell ex vivo hinder their therapeutic potentials, especially in solid tumors. Some of the common strategies include blocking inhibitory receptors while others include evoking cytotoxic T lymphocyte responses such as chimeric antigen receptor (CAR)-T cell therapies and adoptive cell transfer (ACT) approaches. Despite their revolutionary approaches for hematopoietic cancers, their potency and the need for expanding tumor-specific T cells is a need not yet satisfactorily met for solid tumors. One of the major flaws is their administration route and all the immune-related adverse events associated with it. We have shown that the rate at which cytokines are delivered to CD8+ T cells impacts their differentiation and functionality. To improve ex vivo activation and expansion of tumor-reactive T cells, we studied the impact of particle design and mechanical forces. We designed a series of biocompatible artificial APCs (aAPCs) of various sizes and signal intensities. We confirmed that that absolute amounts of antigen engaged by the T cell are more important for activation than the density of antigen. Moreover, we show that T cells can recognize forces arising from the mechanical rigidity of the microenvironment. We fabricated 3D microporous scaffolds with a broad range of mechanical stiffness. We found that there was an augmentation of T cell activation, proliferation, and migration speed in the context of mechanically stiffer 3D matrices as compared to softer materials. Another major challenge that solid tumors face is the abundance of T regulatory cells (Tregs). Here we developed a platform that offers a solution to the above-mentioned challenges; a “synthetic lymph node” niche implanted proximally to the tumor for supporting locally transferred T cells while enhancing their infiltration and cytotoxic capabilities. Our porous synthetic lymph node serves as a home for the recruitment of endogenous T cells and serves them with the activation clues while providing the sustained release of necessary cytokines. We believe the provided niche offers a home to T cells either for ACT purposes or for localized immunomodulation.

We demonstrate the use of micro-scale nuclear magnetic resonance (NMR) for studying the transfer of spin coherence in non-equilibrium chemical processes, using spatially separated NMR encoding and detection coils. As an example, we provide the map of chemical shift correlations for the amino acid alanine as it transitions from the zwitterionic to the anionic form. Our method is unique in the sense that it allows us to track the chemical migration of encoded nuclear spins during the course of chemical transformations.