Ionization Mechanisms in UV-MALDI

Previous: Desorption/Ablation      Table of Contents      Next: Secondary Ionization

3. The Coupled Physical and Chemical Dynamics (CPCD) Model and Primary Ionization
About the name
It is not a "gas phase" model
Primary Ionization
Table 1 Ionization Potentials

3. The Coupled Physical and Chemical Dynamics (CPCD) Model:

About the name:
The model has previously (before 2013) not had a well-defined name. In earlier versions of this tutorial, and in some papers, it was referred to as the "Two Step" model, but this only refers to aspects of ion formation, and is therefore not complete or precise. As noted below, other names were even less precise. In an effort to end this confusion, papers by the author have begun to refer to the model as the Coupled Physical and Chemical Dynamics, or CPCD, model.

This name is intended to reflect the fact that it explicitly covers everything from the condensed phase, through a very dense fluid, and the expansion of that fluid, to the state of isolated ions. In particular, it includes the way chemical and physical processes interact to determine the final result.

Please do not refer to this as a "gas phase" model!!
Some authors have repeatedly referred to it that way. In some cases, this misunderstanding led to unfair criticism of the model, it was argued that certain processes are not possible in the gas phase. As this tutorial will show, central aspects of the model require a dense sample. All of the primary ionization, and most of the secondary processes take place during the period in which the sample is very dense. The model does include the less dense plume expansion period, during which secondary reactions modulate the products of the dense period, but most of the action is over before the sample reaches such a low density that "gas phase" is a good description.

Introduction
The material is heated in a few nanoseconds, but cannot vaporize this fast. See Fig. 1. This period during and soon after the laser pulse is characterized by energy and material density sufficient for high energy processes like ionization to take place. This is the period of primary ionization, when the first ions are created, the first step of the CPCD UV MALDI ionization model.

As can be seen in Fig. 3 the plume starts as a high pressure, very dense fluid. The primary ions have many opportunities to react with neutral molecules to give new ions, which are finally detected. These are the secondary ionization processes. They start as soon as primary ions exist, but continue much longer than primary ionization. This is the second step of the CPCD UV MALDI ionization model.

3a. Primary Ionization
In UV MALDI, a reasonable hypothesis for primary ionization would be that the matrix is ionized by sequential 2-photon absorption. This phenomenon is quite well known in some areas of physical chemistry, and the matrix absorbs the UV energy very efficiently. However, the ionization potentials of nearly all MALDI matrices lie above the 2-photon energy of the typical N₂ (337 nm, 2 photon=7.4 eV) and tripled Nd:YAG (355 nm, 2-photon=7 eV) lasers. See Table 1.

Even in condensed clusters the IPs have been found to be too high for 2-photon ionization (J. Mass Spectrom., vol. 35, pp. 1237-1245 (2000)). A total of 3 photons of energy is needed, but the peak power of lasers used in MALDI is too low for significant direct absorption of 3 photons by a single matrix molecule. However, this amount of energy can be concentrated on one molecule by two sequential pooling events.

Pooling can occur when two neighboring molecules are close enough that their electronic states interact. In MALDI matrices this involves contact of the aromatic pi-electron systems of the stacked, ordered molecules in the solid. Excited neighbors can can concentrate energy on one molecule, this state has the same total energy as the original delocalized excitation.

This would not be very efficient if only directly excited pairs were involved. However, single excitations are mobile in the matrix solid. This is also because they interact with their neighbors. Energetically, it doesnt matter on which molecule the excitation "sits," so it has some probability of hopping about. Consequently wandering excitations collide relatively frequently at typical MALDI laser intensities.

Figure 4. Schematic of a pooling and hopping processes involving neighboring matrix molecules. S₀ is the electronic ground state, S₁ the first singlet excited state, and S_n a higher excited state at twice the energy of the S₁.

Evidence for S₁-S₁ pooling may be found in fluorescence quenching experiments, such as seen in Fig. 5. The steep drop in efficiency was found to be well described by pooling of mobile, hopping excitations. In addition to time resolved fluorescence, traps for the fluorescence gave further verification and validation of the effect. The hop time in DHB is about 50 ps, and the range before decay is typically about 15 molecules. These values are for very pure matrix.

Figure 5. Fluorescence quenching vs. laser intensity in the MALDI matrix 2,5 dihydroxybenzoic acid (DHB). The fluorescence is normalized to the laser intensity, so no quenching corresponds to a value of 1. Quenching at high intensity is consistent with S₁-S₁ pooling. Adapted from Setz and Knochenmuss, J. Phys. Chem. A, vol. 109, p. 4030 (2005).

Evidence for S₁-S_n pooling was found in time-delayed 2-pulse MALDI experiments. A delayed maximum in matrix ion formation efficiency demonstrated that higher excited states are involved in ionization. See Knochenmuss and Vertes, J. Phys. Chem. B, v. 104, p. 5406 (2000). This was later reproduced by the differential equation model presented below. This, finally, is the process which is believed to produce the bulk of all MALDI primary ions!

Figure 6. Schematic of a pooling processes involving neighboring S₁ and S_n matrix molecules. This process leads to ions by combining three photons of energy on one matrix molecule.

Three types of matrix ions generally appear in the mass spectra: radicals (m⁺, m^-), protonation/deprotonation products (mH⁺, (m-H)^-) and cation adducts (e.g. mNa⁺). There is strong evidence that they can be interconverted in the plume. One set of species (m⁺, m^-) is thought to be the true primary ions, the others are derived from it.

Note that in this picture, all primary ions are derived from the matrix. This is not to say that some analytes can not be ionized by the laser, but matrix is so much more abundant that nearly all ions will come from it. Also many analytes are not believed to undergo efficient laser ionization, especially if they do not have strong UV chromophores.