Ordered silicon nanocavity arrays were prepared with e-beam lithography to yield systematically varied pore features and porosity (4-92%). These substrates were used to investigate the effects of substrate morphology on desorption ionization on porous silicon-mass spectrometry (DIOS-MS). Five benzylpyridinium salts, 1,2dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), and angiotensin III were used as the model molecules in the study. For substrates of the same pore depth, MS results suggested that the pore size and the interpore spacing had little impact on the laser irradiation threshold required for ionization. Instead, the laser threshold was found to be highly dependent on the overall porosity for all substrates investigatedsthe higher the porosity the lower the threshold. Moreover, the substrates with deeper pores but of similar porosity showed significantly reduced laser thresholds. This close relationship between laser threshold and substrate morphology was attributed to the thermal confinement property of porous structures. Benzylpyridinium salts were used to study molecular fragmentation tendency during desorption and ionization (D/I). The results suggested the presence of two competing D/I processes: direct laser desorption ionization (LDI) dominated for the substrates of low porosities where analytes desorbed directly from hot silicon surfaces; for highly porous substrates, the retained solvent molecules behaved as the "pseudo" matrix-assisted laser desorption/ionization (pseudo-MALDI) matrix that facilitated analyte desorption and ionization in a MALDI mode.

Toward cracking the problem of understanding, characterizing, and predicting “solvent‐effect” while the world awaits an effective explicit solvent model, we introduce and justify herein a novel set of atomic radii to be used within the most commonly used continuum reaction field, the polarizable continuum model (PCM). The radial values emerge from a quantitative description of the elemental electronic density distribution and are shown to be accurate in such a self‐consistent reaction field (SCRF); labeled accordingly as isodensity‐based SCRF (IDSCRF) radii. Transition row elements with dynamic oxidation states are addressed through an averaging of the electronic properties from all states in the determination of their effective radii. All results for nonmetal elements have been verified with Guthrie's SAMPLE1 test set and are in quantitative agreement with experimental values from the literature and self‐consistent isodensity polarizable continuum model (SCIPCM) calculations. For the compounds with transition metal elements, our IDSCRF results have been verified with SCIPCM results as there are rarely experimental results available. Finally, explicit solvent particles “solvating” Pd‐ and Ni‐containing homogeneous catalysts are also shown to be in close agreement with the IDSCRF radii calculations. © 2012 Wiley Periodicals, Inc.