Sniffing out early reaction intermediates.

P owerful tools for identifying reaction intermediates and byproducts of chemical reactions are crucial for the development of high-yield chemical syntheses, not only for testing theoretical mechanistic models and calculations, but also to obtain improved understanding of reaction mechanisms during trial experiments. This will enable more efficient steering toward optimum yields for a variety of practical synthetic applications. With more demanding challenges regarding synthetic yields, for example for the production of larger and more sophisticated pharmaceuticals, efficient analytical tools become increasingly more important. NMR is one of the most powerful techniques for studying reaction mechanisms in organic reactions, but mass spectrometric tests on the reaction mixture can often be a good complement to time-consuming NMR experiments. The incorporation of a highresolution mass spectrometry technique and the use of a custom-made desorption electrospray ionization (DESI) source as reported by Zare and others (1, 2) offer a remarkably increased ability to identify ions directly, because they are formed in a liquid-phase chemical reaction. In PNAS, Perry et al. (3) study the ions formed when a solution of aminoindanol 2 in methanol (with or without potassium hydroxide) was sprayed onto a surface, by a DESI source, containing a deposited transfer hydrogenation precatalyst 1 ([RuCl2(p-cymene)]2. The reaction starts when the droplets of spray containing 2 hit 1 on a paper surface. Because the surface is close to the mass spectrometer inlet (0.5 cm), the reaction proceeds only for a few milliseconds as the secondary microdroplets travel through the mass spectrometer inlet before formation of gas-phase ions takes place (Fig. 1). The combined strength of early detection and high-resolution MS obviously provides a remarkable richness of intermediates detected for the trial reactions chosen by the Zare group and indicates that these are surprisingly complex when it comes to details. The use of a high-resolution Orbitrap (4–6) mass spectrometer increased the accuracy compared with the quadrupole ion trap mass spectrometer used previously (1) and led to identification of unique ion species not seen with the quadrupole technique. Besides confirming the mechanism and intermediates of the reaction previously studied (7–13), unique complexes were also discovered, indicating unique reaction pathways leading to methyl formate in the absence of a reduction substrate (3). Another finding was a complex consisting of an oxidized species of the ruthenium complex, which was absent when the methanol was bubbled with argon before the experiment. This result demonstrates the importance of removing adventitious O2 from the reaction components, particularly the solvent, to obtain high yield in a chemical reaction and not “poison the catalyst” when a sensitive and reactive catalyst system is used. Mass spectrometry does not provide any 3D structure information but only the composition expressed as formula weights of reaction species. Consequently, careful consideration of all plausible structures that can arise from a given formula is required for a successful interpretation of the experimental data. We envision that the simple approach demonstrated in the article by Perry et al. (3) can be applied to studies of many other interesting reactions in which active catalytic species of precatalysts are important issues. In particular, the approach should be very useful when applied to reactions for which the reaction mechanisms are not known or for which the active components have not yet been determined. We predict that in the near future this technique will become one of the major routine tools used in the laboratory for studying chemical reactions and intermediates, owing to its simplicity and the potential to apply it under a large variety of conditions, such as varied reaction temperature, and reaction mixing times or harvesting times. In addition, the effects of varying diffusion conditions, dielectric and electrostatic properties, etc., could be systematically studied by varying viscosity, phase composition, electrolyte medium, etc. Whereas most reaction mechanistic studies are made according to a minimalistic approach, not considering side-tracks that are not important for the main reaction model, an analytic technique that allows the species distribution and reaction details to be revealed at high resolution is attractive because it gives a richer picture of the total reaction scheme and thereby a picture of the total free-energy landscape of the reaction system. It is not unusual in kinetic studies that, as concentration dependence is studied in more detail, observations appear in conflict with initially assumed reaction mechanisms, which of course confuses. This is almost always because one or several parallel reaction paths have been ignored, thus emphasizing the need for a complete map of the speciation of the system, including temporal resolution. Fig. 1. DESI setup used to intercept transient intermediates of the reaction of 1 (5 μL of a 10M solution in CH2Cl2 deposited on paper) with 2 (10 −4 M in CH3OH infused at 20 μL·min; N2 flow rate = 1 L·min).