Chemistry in retrospect: Personal reflections on a polarizing story

eisenbergMany scientists will at some point come up against the question regarding what good their work might do for the world. For those working in applied and interdisciplinary areas, this is often easier to rationalize to the public than for those working on fundamental, basic research. In this week’s “Chemistry in retrospect,” Rich Eisenberg tells us the story behind PHIP as a case study for why answering fundamental questions in chemistry is so important to the development of the field.

The word “hype” has been used — and abused — when chemists are asked to explain what they do and why.  This extends beyond simple social conversations with friends and folks newly met who will invariably ask with a slightly glazed stare after we explain what we do, “What good will that be?” Even our introductions to papers and proposals promise social and economic good before the science is discussed. And yes, sometimes we do overstate the ultimate social, economic and technological benefits of our science when we are just doing basic research. The key word here is “basic” and it really addresses the key question that drives experimental science, “Why does this happen?”

I want to talk about a basic research story that commenced in my lab more than 30 years ago into which we were drawn by the question, “why?” Thanks to the current efforts of a former postdoc and collaborator, the answer will likely have significant practical applications in magnetic resonance imaging (MRI).  Despite its widespread use, MRI lacks sensitivity that requires use of large-magnet instruments and specialized contrast agents to achieve current results which are often less than optimal.  But let’s go back to the beginning and see how that simple question “why” drove us to where we are now.

In 1985, my group was examining hydrogenation reactions of alkynes and olefins promoted by rhodium complexes using NMR spectroscopy and observed an extraordinary enhancement in the intensities of the resonances of newly-added H atoms in the products.  We initially considered the cause of the enhancement in terms of radical pair reactions and resultant chemically induced nuclear polarization (CIDNP) but tests of this hypothesis, which included the arduous task of synthesizing and studying isotopically labeled substrates for hydrogenation, proved negative.

It was the innovative work of Dan Weitekamp, then a new Assistant Professor at Caltech, that provided the answer.  From the earliest days of nuclear spin and quantum mechanics in the 1920’s, it was known that molecular hydrogen exists in two forms, orthohydrogen and parahydrogen in a ratio of ~3:1 at room temperature, since orthohydrogen has three nuclear spin states and parahydrogen has only one.  Parahydrogen is slightly more stable than orthohydrogen and, as these molecules approach lower temperatures, this slight difference becomes apparent.  At 77 K, the ortho:para ratio is nearly 1:1, and when cooled to liquefaction, H2 is nearly 100% para. The spontaneous interconversion of the two forms is essentially forbidden from a quantum mechanical perspective, and is only promoted by interactions with third party molecules or materials.  Weitekamp’s hypothesis was that if a hydrogenation were done using H2 enriched in the para spin state, one could observe enhanced resonances if — and only if — the two hydrogen atoms that were added to a substrate originated from the same H2 molecule.

We were informed of Weitekamp’s proposal, but were uncertain how such an enrichment of parahydrogen relative to orthohydrogen could happen, but it soon became apparent. Because of local NMR sign-up rules, we needed to conduct our experiments at night, and to ensure that we had a number of samples to examine, we would prepare them many hours beforehand and store them in liquid nitrogen during which nothing would happen —    except for the slow conversion of orthohydrogen to parahydrogen promoted by the compounds in the NMR tube.  Only after the sample was rapidly melted and shaken to mix in the para-enriched hydrogen did we observe the extraordinary NMR enhancements we were seeing. So began our studies of parahydrogen induced polarization (PHIP).  After finding — or more correctly, stumbling into — PHIP, we examined H2 addition to metal complexes and to organic substrates catalyzed by metal complexes using this phenomenon in order to obtain positive evidence of one aspect of the mechanism of reaction.  If PHIP was observed, then the two H atoms that were added to a metal complex or an organic substrate came from the same H2 molecule.  There was, however, another aspect of PHIP that is key to the present story.

When a 1H NMR resonance is observed experimentally, it is because there is a difference between the number of protons that have one spin state (a) versus the other possible spin state (b).  This population difference is, in fact, the basis of NMR, and the simple two-level approach holds true for other  S = ½ nuclei such as 13C, 19F, 15N and 31P as well as for 1H.  In a 400 MHz spectrometer, the difference in populations of the two 1H spins for a given resonance is only 1 in approximately 32,001 — the a state has 16,001 spin and the higher energy b state has 16,000.  If the difference between the populations of these two spin states could be increased, then resonances could be enhanced in magnitude and/or detected from smaller or more dilute samples. This was an objective of much-heralded research on pulse sequences in NMR spectroscopy, including how these methods could be used to affect populations of other nuclei in the molecule such as 13C, 31P and 15N.  Since PHIP also resulted in a change of the relative amounts of a and b spins for the added H atoms, we examined and found that signal enhancement was transferred to other nuclei in the product compounds.

It was at this time that Simon Duckett joined my research group as a postdoc. Duckett is now a Professor at the University of York and one of the leading practitioners of PHIP methodology today.  In my lab, Duckett combined PHIP with a signal enhancement  pulse sequence to observe an enhanced resonance for the 13C nucleus in a natural-abundance sample  (1% 13C) of an iridium carbonyl compound, corresponding to a signal-to-noise increase of 25,000. In the last two decades, Duckett’s studies on PHIP for enhancing resonances of 1H, 13C and 15N in organic compounds that were in solution with, but not direct acceptors of, parahydrogen, have proved successful. In these studies, parahydrogen was adding to a metal complex but the resultant hydrides were not adding to the organic compound. The process was referred to as SABRE (Signal Amplification By Reversible Exchange).  In an article in Accounts of Chemical Research, 2012, 45, 1247-1257 by Duckett and Mewis, the following specific points (only slightly modified for context) were made:

Hydride-derived PHIP was transferred to 15N nuclei by an INEPT based procedure, after which ligand exchange yields a free pyridine 15N signal 120-times larger than normal even though it does not contain any parahydrogen derived protons. A more dramatic effect has been observed when systems with weaker association between the metal and substrate are studied in low magnetic field. In this case, transfer of nuclear spin polarization from parahydrogen to free pyridine molecules is observed without the need for radio frequency irradiation. Blümich and co-workers have recently shown using SABRE that nanomole amounts of pyridine can be detected; a further report has demonstrated its use with peptides. These approaches therefore polarize an analyte without the need for chemical incorporation of parahydrogen-derived nuclei into it; all that is necessary is to bind reversibly to the metal center……Here, no chemical modifications were required to a substrate even though a 10,000-fold increase in MR (magnetic resonance) signal has been demonstrated. “

This research by Duckett and other investigators using PHIP and SABRE for magnetic resonance imaging is impressive. If done at low magnetic fields, as has been demonstrated in certain ways at this time, it could be truly transformative.  One may be able to do sensitive MRI without the use of a large magnet, and possibly away from any laboratory using the Earth’s magnetic field. So is this hype?  The future will tell us.  But the beauty of the story to me is that it was fundamental research and seeking answers to “Why does [fill in the blank] happen?” that drove me, and now others, even while leading to a compelling new approach for magnetic resonance imaging.

Richard Eisenberg is the Tracy Harris Professor Emeritus in the Department of Chemistry at the University of Rochester.