Thermodynamic control of final ion distributions in MALDI: in-plume proton transfer reactions

Dedicated to Prof. Franz Hillenkamp on the occasion of his 65th birthday.
https://doi.org/10.1016/S1387-3806(02)00965-XGet rights and content

Abstract

Gas-phase thermochemical data on matrix-assisted laser desorption/ionization (MALDI) matrix species were used to calculate the energetics of possible proton transfer reactions. These were investigated for the MALDI matrices 2,4,6-trihydroxyacetophenone and sinapic acid, with the tripeptide glycyl–glycyl–histidine as analyte. MALDI proton transfer product ion distributions were found to be predicted by the energetics of possible secondary ion–molecule reactions, at all laser fluences sufficient to generate a dense plume. Near the ion generation fluence threshold, the mass spectra deviate from the thermodynamic predictions. This shows that the MALDI plume exhibits both thermodynamically and kinetically controlled regimes, depending on desorption conditions. Possible thermal ground state proton transfer primary ionization pathways were also considered, and found to be inconsistent with the data.

Introduction

Matrix-assisted laser desorption/ionization (MALDI) [1], [2], [3] has become an indispensable tool for biomolecular analysis, and the mechanisms by which both the primary and the final, observed ions form are currently of much interest [4], [5], [6], [7], [8], [9]. The hope is that, with deeper understanding, MALDI can be rationally and systematically developed, and that the outcome of MALDI experiments can be predicted and planned. Practical consequences could be greater sensitivity, wider applicability, better reproducibility, and possibly quantitative measurements.

As recently described [10], MALDI ionization mechanisms can be divided into primary and secondary ionization steps. By primary steps we mean those processes converting neutral reactants into charged products by action of the laser pulse on the sample. These include multi-photon ionization, energy pooling mechanisms, disproportionation reactions, excited-state proton transfer, thermal ionization, desorption of preformed ions, and break-up of the sample into charged chunks and clusters [10]; these are not the subject of the present study. The primary steps may differ significantly for infrared and ultraviolet desorption/ionization.

Secondary steps are any subsequent ion–molecule reactions that convert the initial charged species to the ions that are observed at the detector. These may involve proton transfer, electron transfer, and cation attachment or transfer. As proposed in [7], primary ionization may be partially or completely masked by such secondary reactions. While this complicates elucidation of the primary ionization steps, it explains the similarity of MALDI spectra recorded under different conditions, and provides a framework for predicting MALDI mass spectra. Because thermochemical data for matrix and analytes are becoming more available [7], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25] these hypotheses are increasingly subject to experimental test.

A more comprehensive discussion of secondary ion–molecule reactions in MALDI can be found in a recent article [7]; here we present a case study supporting the concepts presented therein, for systems in which proton transfer reactions are dominant. The present study is concerned only with proton transfer reactions. Other ions are often observed in MALDI, particularly metal adduct ions and matrix radical species. The minor quantities of such ions which can be found in the systems studied here are not treated in detail because not enough thermodynamic information is known, and because they are clearly far less important than the protonated or deprotonated species.

Systems were chosen for this study in which Gibbs free energies of matrix–matrix and matrix–analyte proton transfer reactions are known or were determined experimentally. This allows us to test the hypothesis that thermochemistry determines the final ion distributions as a result of efficient secondary reactions in the plume. Experimental conditions were also chosen so as to most clearly test the hypothesis. For investigating matrix–analyte reactions, samples were prepared by mixing matrix and analyte in molar ratios of as high as 1:1. Such high analyte concentrations are not typically used in MALDI experiments, but were chosen here to ensure that reactions are not limited by quantities of reactants. Under these conditions, complete suppression of all matrix ions in the mass spectrum can be observed. This “matrix suppression effect,” has been studied in some detail [7], [26], [27], [28]. Because MALDI plume density and therefore the number of collisions can be expected to increase with increasing laser fluence, the effect of fluence on relative ion yields was also investigated.

Section snippets

Experimental

Experiments were performed on a Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer with a 4.69 T superconducting magnet (Bruker, Fällanden, Switzerland). The rf electronics and Odyssey data acquisition system were from Finnigan (Finnigan FT–MS, Madison, WI, USA). The laboratory-built vacuum system was comprised of a closed cylindrical ion cell of unit aspect ratio and a sample transfer device for insertion of solid material. The instrument operating pressure was below 10−8 mbar.

Matrix suppression effect with THAP matrix and GGH analyte

Fig. 1 shows the mass spectra of THAP alone and with GGH analyte, in both positive and negative modes. A matrix suppression effect [7], [26], [27], [28] is observed in positive mode, analogous to that found for numerous other matrix–analyte pairs [27], [28]. As is often the case with the matrix suppression effect, the matrix–analyte ion intensity ratio shows a step-like approach to suppression at high analyte concentration. GGH being a relatively low molecular weight analyte, the suppression

Conclusions

Matrix–analyte systems exhibiting different proton transfer energetics were examined to probe the role of thermodynamics in determining the final ion distributions in MALDI mass spectra. Relative yields of matrix species and analyte ions strongly depend on the laser fluence used in the MALDI experiment, and ion distributions at fluences not too near threshold are consistent with ground state proton transfer thermochemistry, even for the complex system of ions observed with sinapic acid matrix.

Acknowledgements

The authors thank the Kommission für Technologie und Innovation (KTI, grant no. 3165.1) and the Austrian Science Fund (FWF, project P15767) for financial support.

References (41)

  • M. Karas et al.

    Int. J. Mass Spectrom.

    (1987)
  • K. Breuker et al.

    J. Am. Soc. Mass Spectrom.

    (1999)
  • K. Breuker et al.

    Int. J. Mass Spectrom.

    (1999)
  • S.R. Carr et al.

    J. Am. Soc. Mass Spectrom.

    (1996)
  • J.W. McKiernan et al.

    J. Am. Soc. Mass Spectrom.

    (1994)
  • R.A.J. O’Hair et al.

    Int. J. Mass Spectrom.

    (1992)
  • Z. Wu et al.

    Tetrahedron

    (1993)
  • K. Dreisewerd et al.

    Int. J. Mass Spectrom.

    (1995)
  • M. Karas et al.

    Anal. Chem.

    (1988)
  • M. Karas et al.

    Angew. Chem. Int. Ed. Engl.

    (1989)
  • M. Glückmann et al.

    J. Mass Spectrom.

    (1999)
  • M. Karas et al.

    J. Mass Spectrom.

    (2000)
  • G.R. Kinsel et al.

    J. Mass Spectrom.

    (1999)
  • R. Knochenmuss et al.

    J. Mass Spectrom.

    (2000)
  • V.L. Talrose et al.

    Rapid Commun. Mass Spectrom.

    (1999)
  • S. Alimpiev et al.

    J. Chem. Phys.

    (2001)
  • R. Zenobi et al.

    Mass Spectrom. Rev.

    (1998)
  • T.J.D. Jørgensen et al.

    Eur. Mass Spectrom.

    (1998)
  • R.D. Burton et al.

    Rapid Commun. Mass Spectrom.

    (1997)
  • R.J.J.M. Steenvoorden et al.

    Eur. Mass Spectrom.

    (1997)
  • Cited by (60)

    • Organic matrices, ionic liquids, and organic matrices@nanoparticles assisted laser desorption/ionization mass spectrometry

      2017, TrAC - Trends in Analytical Chemistry
      Citation Excerpt :

      The performance of different matrices for the analysis of polyhexamethylene guanidine (PHMG) polymers was in the order of CHCA ∼2,5-DHB > 5-NSA > DHAP, THAP > ATT > IAA ∼ super-DHB ∼ HABA [366]. The ion-to-neutral ratio under conditions outside MALDI showed values of 10−3–10−4 [367–370]. The ion-to-neutral ratio was determined for different organic matrices and laser wavelength [371–374].

    • Formation of gas-phase metal fluorides in reactions of fluorinated fullerenes at activated metal surfaces

      2017, Journal of Fluorine Chemistry
      Citation Excerpt :

      For instance, the observed intensity ratio of I(AlF4−)/I(F−) was approximately 2.40 ± 0.06 and the ratio I(AlF4−)/I(Al2F7−) gave a value of 6.00 ± 0.09. Recent experiments by Zenobi et al. [45] showed that the MALDI plume exhibits both thermodynamically as well as kinetically controlled regimes. For the matrix-assisted protonation of peptides using laser fluences that were clearly above the threshold value, it was found that the ion distributions were consistent with the ground state proton transfer thermochemistry.

    • Ionic liquids for mass spectrometry: Matrices, separation and microextraction

      2016, TrAC - Trends in Analytical Chemistry
      Citation Excerpt :

      The analysis was investigated without the need of internal standard [89]. The quantitative analysis was also extended to Desorption Corona Beam Ionization (DCBI) [116]. The applicability of ILMs for 21 small low-polar molecules indicated the promising future of these materials for mass spectrometry.

    View all citing articles on Scopus
    1

    Present address: Department of Chemistry, University of Innsbruck, Innsbruck, Austria.

    2

    Co-corresponding author. Present address: Novartis Pharma AG, WSJ-503.1104, CH-4002 Basel, Switzerland.

    3

    Present address: Department of Analytical Chemistry and Applied Spectroscopy (ACAS), Faculty of Sciences, Vrije Universiteit Amsterdam, de Boelelaan 1083, 1081 HV Amsterdam, The Netherlands.

    View full text