Force-detected NMR in a homogeneous field: Experiment design, apparatus, and observations

Here I present motivation, experimental progress, and theoretical aspects of the BOOMERANG (b&barbelow;etter o&barbelow;bservation o&barbelow;f m&barbelow;agnetization e&barbelow;nhanced r&barbelow;esolution a&barbelow;nd n&barbelow;o g&barbelow;radient) method of force-detected NMR, a general approach to extending arbitrary NMR experiments to the micron scale and below. Enabling quality of BOOMERANG is that its sensitivity scales much more favorably than traditional inductive detection for small samples. A reduction in the sample size accessible by NMR is strongly motivated by such goals as massively parallel analysis in support of combinatorial chemistry, portability in support of planetary exploration, and the general advantage of highest sensitivity per unit cost.;The key design insight is that the spin-dependent forces are independent of the field homogeneity across the sample. However, throughput is optimized only by providing field homogeneity during detection sufficient to allow coherent control over all target spins in a sample. I present our BOOMERANG design concepts and strategies, which allow detectors with high geometrical efficiency and good prospects for low mechanical dissipation.;The design principles are quantitatively confirmed using a prototype mm-scale spectrometer. Our experimental results, which include proton and fluorine FT-NMR spectra in solids and liquids, heteronuclear J spectra, and liquid-state spin echoes with sub-Hz linewidths, emphasize BOOMERANG's general spectroscopic applicability.;Fabrication of a high-sensitivity spectrometer optimized for 60-micron samples is underway in conjunction with the Microdevices Laboratory (MDL) at the NASA Jet Propulsion Laboratory (JPL). Using state-of-the-art lithography and electrodeposition techniques, we have fabricated magnets and mechanical oscillator structures that show promise for incorporation into spectrometers for in-situ planetary exploration, and for massively parallel analysis.;As the sample size decreases, sensitivity is dominated by quantum-statistical noise in the sample, or spin noise. This fundamental problem is mitigated by the CONQUEST measurement paradigm involving multiple time-correlated measurements on a spin system of interest. This is an essential ingredient in converting polarization fluctuations to coherent time-domain spectroscopy or to images with arbitrary numbers of spins in each pixel.