Description
The selectivity and function of biologically active molecules is closely intertwined with their three-dimensional structure, which in turn depends on the skeletal shape resulting from covalent and non-covalent interactions, such as hydrogen bonds between neighbouring parts of the biomolecule. In addition to function, structure also controls molecular transport, molecular recognition and selective binding of ligands, such as adenosine triphosphate (ATP) association at the active site of ATPases to initiate enzymatic activity. A delicate balance between intra- and intermolecular hydrogen-bonding forces and hydrophobic interactions controls the resulting biomolecular conformation, and is key to understanding the function of biological molecules. Not surprisingly, much effort has been devoted to the elucidation of biomolecular structure and the associated interactions. Well-known techniques such as X-ray crystallography, nuclear magnetic resonance (NMR), single molecule spectroscopy, mass spectrometry and computational simulations have been employed to clarify the structure-to-function relation of biomolecules. In addition, molecular spectroscopic techniques have been developed to study biolog-ical molecules under isolated conditions, removing inhomogeneous line broadening induced by environmental influences. Two parallel pathways, both using molecular spectroscopy to probe either neutral or ionic isolated species, have led to the development of state-of-the-art techniques, which are described in Sects. 1 and 2 of this chapter, respectively.
To employ the rich toolbox of gas-phase optical spectroscopy techniques in the study of this class of molecules, the biomolecules need to be brought into the gas phase and cooled to their lowest energy conformations. Pioneering experiments on the amino acid tryptophan were performed by Levy et al. using thermospray evaporation combined with a supersonic beam [1]. This allowed them to record the electronic spectrum of tryptophan and tryptophan analogs under the cold and isolated conditions of a supersonic molecular beam. For the first time, the existence of different conformations was inferred from the power dependence of the signals resulting from electronic transitions [1–5]. Later, a more elegant version of con-formation selection was used, IR–UV double resonance spectroscopy [6–8], to identify spectral features originating from distinct conformers [9–11]. Its full potential was shown when it was combined with quantum-chemical calculations to assign vibrational bands and reveal conformational structures [11]. Since then, this method and a number of its variants have been applied to numerous biomolec-ular systems. In Sect. 1 of this chapter we focus on the methods currently employed to obtain these IR spectra of mass-selected, conformation-selected neutral bio-molecules in the gas phase.
Neutral Biomolecules in the Gas Phase
As virtually all biological molecules possess low vapour pressures gas-phase molecular spectroscopy methods to investigate isolated neutral biomolecules require volatilization methods other than thermal evaporation. The combination of laser desorption with a supersonic molecular beam expansion together with the selectivity of IR and UV double resonance methods opened up the possibility of characterizing isolated, neutral biomolecules and their clusters with the biological environment.
Transfer into the Gas Phase: Laser Desorption
The simplest method to transfer molecules as neutrals into the gas phase is by thermal heating. Here, a temperature-controlled oven is located either before or after the nozzle in the pre-expansion region. This method has been used for a limited number of small biomolecules including some nucleobases [12–15], amino acids [1–5], anaesthetics [16–23] and several neurotransmitters [24–26]. However, this method often requires very high temperatures to achieve significant vapour pressures, resulting in impractical experimental conditions and unacceptable high levels of thermal decomposition.
Thermal degradation can be avoided by instant heating of the biomolecule of interest using laser desorption. Laser desorption in combination with supersonic expansion has been widely used to bring intact biomolecules into the gas phase [27]. The non-volatile molecules are deposited on a sample bar made of a material that is believed to assist the desorption process. Various matrices, such as activated carbon, fritted glass [28, 29], polyethylene [30] or graphite [31–35], have been used, although matrix-free desorption (bare molecules) has been performed as well [36]. Besides choice of sample bar material, variations in sample preparation (deposited as a thin layer, doped, pre-mixed), sample bar shape (rods, pressed discs, flat stages) and the choice of desorption laser (pulse length, wavelength, fluence) result in slight modifications in the desorption mechanism. For the exper-iments discussed in Chaps. [37–39] of this book, molecules are desorbed from a sample bar of ultrafine grain graphite with a low intensity pulsed Nd:YAG laser (fundamental or frequency doubled) with output energies of about 1–2 mJ.
Various mechanisms have been proposed to explain the process of laser desorp-tion [40–44]. Here, we focus on the mechanism where the matrix plays a passive role as energy transmitter, i.e. laser-induced thermal desorption (LITD) by indirect heating of the graphite substrate to bring neutral non-volatile molecules into the gas phase. When the desorption laser hits the surface of the graphite sample bar, instant, fast and extreme heating takes place. The main difference between laser desorption and thermal heating lies in the heating rate of the substrate. Rates for resistive heating are of the order of 100–102 Ks1, while for laser desorption they are of the order of 1010–1012 Ks1 (about 1,000 K in a 1-ns laser pulse). Two processes now compete: the molecules either desorb from the surface and enter the gas phase intact or they react and fragment [45]. Although the reaction/fragmentation energy barrier might be lower than the barrier for intact desorption, it was shown that high heating rates allow the energetically unfavourable desorption process to occur on account of entropic reasons [44]. It should be noted that higher heating rates are not always better. After an optimum is reached, higher laser fluences result in a high amount of desorbed sample molecules, which may disturb the supersonic expansion, the cooling conditions and even complete molecular clusters may be ablated from the sample substrate.
The major disadvantage of laser desorption lies in shot-to-shot fluctuations in the amount of sample molecules injected into the supersonic expansion. A small fluctuation in desorption laser power can cause larger fluctuations in the amount of desorbed molecules [27]. Variations can be minimized by maintaining a constant laser fluence and by creating a homogeneous sample. The stability and strength of the signal (the amount of evaporated sample molecules) can be improved by mixing the sample bar with carbon black powder [46], forming a more homogeneous sample.
Figure 1 presents a schematic overview of a typical molecular beam time-of-flight mass spectrometer equipped with a laser desorption source. In the studies presented in this book, the sample bar is made from graphite. Accurate positioning of the sample bar with respect to the nozzle is required for optimal performance. It is typically mounted on a double translation stage (Fig. 1). The vertical travel (x-direction) with a typical accuracy better than 0.01 mm allows for optimal cooling with minimal distortion of the molecular beam expansion. The sample bar is typically positioned about 0.1 mm below the aperture of the pulsed molecular beam valve. Travel in the horizontal direction (y-axis) of 50 mm (length of the sample bar) with a position accuracy of about 0.1 mm ensures desorption of fresh sample at every laser shot. Both positioning options can be controlled under operating conditions. Finally, the distance along the molecular beam (z-axis) between the sample bar and nozzle can be adjusted when necessary, but should be kept as close as possible for optimal cooling. The shape of the sample bar (knife shaped) as well as its distance to the nozzle and the molecular beam axis is chosen to influence the molecular beam expansion as little as possible while obtaining as much signal as possible.
Desorption of molecules ranging in size from single amino acids to systems with molecular weights up to 2,000 Da has been demonstrated, where a monolayer or less is desorbed by the heat generated at the surface. However, some species are difficult to bring into the gas phase without inducing fragmentation. For example, arginine-containing peptides tend to undergo substantial fragmentation [47, 48]. In addition, nucleosides show varying behaviour; guanosine does not fragment at all, but adenosines may show significant fragmentation [49, 50].
In most experiments, argon at backing pressures ranging from 1 to 8 bar is used as a seed gas, although other noble gases have been used as well. For example, for molecules with a molecular weight close to 2,000 Da, xenon has been found to improve the supersonic cooling of the sample [51].