Laser-induced breakdown spectroscopy (LIBS) offers significant advantages for rapid, remote,
and minimally destructive elemental analysis but suffers from limited accuracy and sensitivity
due to plasma heterogeneity, matrix effects, and instrumental constraints. This thesis addresses
two critical areas to enhance the reliability and analytical capability of LIBS.
First, plasma plume dynamics at a metal boundary was investigated through 2D and 3D imaging,
interferometry, and spectroscopy under varying ambient pressures. The study revealed distinct
plasma asymmetries at material boundaries, especially pronounced at lower pressures, highlighting substantial deviations from the conventional assumption of plasma uniformity. Additionally,
signal intensity was found highly sensitive to optical alignment, emphasizing the importance of
precise alignment of the collection optics during elemental mapping and quantitative analysis
of heterogeneous samples. Most importantly, this study reinforces the need to critically assess
the commonly adopted assumption of plasma homogeneity in LIBS, particularly when dealing
with fine heterogeneous structures, where this assumption may not hold true.
Second, the potential of orthogonal triple-pulse LIBS to enhance the signal of elements with
high excitation energies were systematically studied on soft borosilicate glass and lunar regolith
simulants. Significant signal enhancements, particularly for ionic emission lines with high excitation energies, were demonstrated, outperforming both single- and double-pulse configurations.
Moreover, the relationship between spectral line excitation energies and signal enhancement was
studied in collinear double-pulse configuration, revealing a linear dependence and suggesting
that enhancement predominantly results from increased ablated mass and extended plasma lifetimes rather than higher plasma temperatures.
These findings provide deeper insight into LIBS plasma behavior at complex material boundaries and offer practical optimization strategies for significantly improving the analytical performance of LIBS across diverse matrices