Mechanisms of energy deposition in infrared matrix-assisted laser desorption/ionization mass spectrometry

The mechanisms of energy deposition in matrix-assisted laser desorption/ionization mass spectrometry with infrared lasers (IR-MALDI-MS) have been evaluated in a series of experiments. In a first part, the threshold fluences for the generation of IR-MALDI spectra were determined between 2.7 and 4.0 mum wavelength with an optical parametric oscillator as a tunable laser source for nine solid state and two liquid matrices of different chemical structure and compared to the infrared absorption spectra of the compounds. Preliminary spectra of IR-MALDI in the wavelength range of 1.45-1.75 mum are also presented using the overtone vibrations of a glycerol matrix. Matrices were chosen with regard to their IR-MALDI performance and to allow conclusions on the IR-absorption mechanisms. Whereas the wavelength dependence of the threshold fluence for non-hydrogen-bound C-H vibrations essentially follows the absorption spectrum of this functional group, strong discrepancies between the spectral dependence of threshold fluences and IR-absorption spectra were found for the vibrations of O-H and N-H groups around 3-mum wavelength that form strong intermolecular and intramolecular hydrogen bonds. In a second part, experiments are described that interrogate the two most probable mechanisms for the observed deviation of the threshold fluence behavior from the wavelength course of the IR-absorption spectra, that is, absorption by either free or weakened O-H and N-H stretching modes or by residual free water. All investigations were performed with glycerol and succinic acid as examples for common liquid and solid state matrices for IR-MALDI. For glycerol, a fluence-dependent, dynamic change in absorption during the laser pulse was revealed by laser transmission measurements on thin glycerol layers. This effect, characterized by a significant blue shift of the O-H stretch absorption, can be attributed to a weakening of intermolecular hydrogen bonds caused by the transient laser heating of the sample. Taking this effect into account, a good correspondence of the wavelength dependence of the threshold fluence with the infrared absorption under IR-MALDI conditions is derived for glycerol. For succinic acid, in contrast, the identification of the predominant absorption mechanism in the 3-mum wavelength range appears more difficult. A fluence-dependent absorption was not detected in laser transmission experiments with succinic acid single crystals. A change in analyte-to-matrix ratio, with the intention of inducing free absorbers near crystal defects, also did not influence the wavelength dependence of the threshold fluences. However, an influence of the surface-to-volume ratio on the wavelength-dependent threshold fluences was found by a comparison of three different preparation techniques for succinic acid, indicating a putative influence of weakly hydrogen-bound surface absorbers. In combination with the detailed analysis of the wavelength dependence of the threshold fluence given from the first part, a determination of the IR-MALDI process for succinic acid based on the absorption by weakly hydrogen-bound hydroxyl groups is suggested. No evidence for a significant contribution of residual free water absorption to the low-threshold fluences around 3-mum wavelength was found by monitoring a possible change in threshold fluence at the phase transition from water to ice and by reducing the analyte hydration and varying the writer content in glycerol preparations. Also, in preparations with frozen hydrated proteins without organic matrices, the wavelength dependence of the threshold fluence did not reflect the spectral absorption of ice, supporting the assumption of a rather minor role of the absorption by residual water in IR-MALDI. (C) 2001 Elsevier Science B.V.