Low-temperature matrix isolation has been the key to studying catalytic reactions so far. By slowing the reaction to a rate at which the intermediates are essentially suspended for longer between reactant and product, conventional spectroscopic methods can be used to get a close look at their structures. For instance, an organometallic might be frozen in a noble gas or hydrocarbon glass between 10 and 77 K. The main drawback is that any thermal steps in the reaction cannot then be seen because of the low temperatures involved.

Ultraviolet (UV) flash photolysis offered an alternative view that could be seen at room temperature using time-resolved infrared (IR) spectroscopy. The UV flash essentially blasts apart the intermediate on the femto- to milli-second timescale so that it is caught in a freeze frame — like a dancer under a strobe light. The technique is powerful not least because it allows reactions to be studied under more realistic conditions than in matrix isolation but still some aspects of a catalytic mechanism remain opaque to the flash.

The polymer matrix immobiliser

In their latest work appearing in J Am Chem Soc, the Nottingham team reveal how they have discovered that a polymer matrix can be used to immobilise a reaction in limbo. With the intermediates suspended in this manner, Fourier transform IR (FTIR) spectroscopy and coupled gas chromatography–mass spectrometry (GC–MS) can be used to look at even the most complicated catalytic mechanism. They have obtained 'a surprisingly high level of control' over otherwise thermally unstable intermediates. The team has applied the method to two catalytic reactions — the hydrogenation of dimethyl fumarate (DF) and of norbornadiene (NBD) to demonstrate proof of principle.

The team used a polyethylene (PE) matrix to freeze out the intermediates in a similar manner to standard matrix isolation, but at high pressure inside a unique cell. The major difference, however, was that they could then warm the matrix without it vaporising to track how any intermediates trapped inside the PE change as the temperature rises.

The team applied the technique to the hydrogenation of DF using [Fe(CO)₄(η²-DF)] as the catalyst. They first irradiated the catalyst in the PE matrix at 150 K. This led to the formation of an intermediate complex in which the DF is coordinated to the iron, revealed by FTIR spectroscopy. Now, when the matrix is warmed to 260 K under hydrogen gas, H₂ is tacked on to this intermediate, a step which they again could see clearly using FTIR spectroscopy. The final stage involved further heating which hydrogenates the coordinated DF forming dimethyl succinate (DS) as the product as identified by GC–MS at room temperature.

Similarly, UV photolysis of [(NBD)M(CO)₄] (where the metal is either Cr or Mo) in the PE matrix under hydrogen gas yields the hydrogenated product norbornene (NBN) and nortricyclene (NTC), and a trace of norbornane (NBA). The team explains that the product ratios are close to those observed in a more conventional hydrocarbon solution. However, for the molybdenum catalyst the team found they could change the outcome considerably if they removed free NBD. NBA then becomes the main product. Further studies with hydrogen and deuterium gas exchange, using the specially designed pressure cell, then allowed them to see how this change in product profile might arise. Such isotopic labelling experiments, the team points out, are highly revealing but nigh on impossible in standard experiments because there is no means of controlling the various gases without releasing the pressures.

Considering that this reaction has been studied for more than 30 years it is intriguing to find that a change of perspective can generate something new that might lead to a clearer understanding of its complex reaction mechanism. The clue lies in the previously recognised non-classical hydrogen-complexed intermediates in which the hydrogen molecule is coordinated to the metal 'through' its bond rather than with direct M–H atom connectivity. The difference between the outcome of the molybdenum and the chromium-catalysed reaction results from the difference in how each of those metals favour bonding in this non-classical way to the hydrogen molecule and the impact this has on the overall structure of the intermediate formed.

The team believe there are still several questions about these particular reactions that remain to be answered. The origin of the doubly hydrogenated NBA, for instance, remains a minor mystery. The researchers also point out that while they could detect and identify many intermediates with their new technique they never actually observed directly the crucial reaction step — the transfer of hydrogen from the metal to the organic substrate. This may become transparent to further studies in time, but the demonstration of the power of the new technique is beyond doubt. Homogeneous catalysis has never before been so see-through.