
Turning Setbacks into Progress: Unlocking Design Principles for Enhanced Dynamic Nuclear Polarization Through Electron Spin Clusters
Have you ever tried to get an exciting result, whether it be a good test score, beating a tough game boss, or nailing an interview, only to find out that you didn’t get that cool outcome you initially expected? Even when that happens, there is something quite powerful about how we make the most out of unideal circumstances and make a solid impact anyway.
That is the story behind our recently published paper “Dynamic Nuclear Polarization Using Electron Spin Cluster,” a story that initially aimed to enhance NMR signal in an unconventional way, failed to do so, but instead became about surefire design principles on how to make this work in the future, backed up by quantum mechanical simulations.
Backing up a bit, what is our research about anyway? Well, you know how when you go to the doctor and get an MRI scan, they can peer into your body and figure out what’s going on with your health? The reason why that MRI works is because of a phenomenon called Nuclear Magnetic Resonance (NMR), which uses radio waves to detect the structure and dynamics of nuclei via a nuclear property called “spin.” If the radio waves sync up with the nuclear spin’s resonance frequency, then you get detection. NMR is super versatile, able to study the nuclear spins of a wide array of things, from people to cells to materials to even swamp water.
However, NMR suffers from a fatal flaw–the poor sensitivity of nuclear spins. We’re trying to improve NMR’s sensitivity by activating the high sensitivity (or spin polarization) of electron spins via microwaves and transferring that over to nuclear spins in a process called Dynamic Nuclear Polarization (DNP), which often leads to orders of magnitude of improvement!
The way that we use electrons to improve MRI is by introducing organic radicals into whatever system we are enhancing the signal of. These organic radicals may be optimized for different DNP mechanisms of transferring electron spin polarization over to nuclear spins, and characterizing these DNP mechanisms is critical to improving DNP.
We knew from some of our earlier work that radical clustering improves DNP signal enhancement by quite a lot and figured that this is because of the Thermal Mixing DNP mechanism via strong electron-electron spin coupling, which may enable more efficient transfer of their spin polarization to nuclear spins. Through a collaboration with the Bagryanskaya research group, who synthesized organic radicals for us, that contain four electron spins each, called “TetraTrityl,” hoping to study how a smaller controlled cluster gives rise to Thermal Mixing.
As this was my first research project in the group I had a naive outlook on our initial results, thinking everything was exciting. When I ran the first sample with TetraTrityl and saw its 1H DNP enhancement profile, I was initially excited, as the profile shape showed exactly the dispersive feature of Thermal Mixing. However, my more knowledgeable mentors and advisor realized something was wrong–while the profile shape was good, the DNP enhancement was poor. However, this could be okay if the profile was coming from individual TetraTrityl rather than a TetraTrityl cluster, as the former would still give insight into how Thermal Mixing DNP arises. So, we went to EPR to study the electron behavior and interestingly saw multiple peaks. This seemed exciting, but upon further study we realized all we had found what seemed to be an uninteresting result—TetraTrityl does not give Thermal Mixing DNP via 4-spin clusters but instead due to stochastic clustering, just like what was already known. Now what?
Well, looking carefully at the expected electron-electron spin distances in TetraTrityl based on molecular dynamics simulations, we found that those distances yield coupling strengths that are weaker than expected for matching with the 1H nuclear resonance frequency. Perhaps the issue is that the TetraTrityl itself doesn’t have short enough electron-electron spin distances to tune to the nuclear resonance frequency, like a radio that doesn’t have enough bandwidth to tune into your favorite station.
From these results, I was able to perform quantum mechanical simulations to show that TetraTrityl cannot give Thermal Mixing DNP even in simulations and that bringing in another electron very close to TetraTrityl can replicate our experimental DNP profile. From here the study took off and we ended up with clear design principles for what we now called ESC-DNP or electron spin cluster DNP as a subset of Thermal Mixing DNP that focuses on more well-defined electron spin clusters. These design principles should hopefully make the next radical synthesis group able to yield radicals that give us that high DNP enhancement that we initially aimed for.
That is the story behind our paper. Yes, we did not get the results we initially expected. But we turned lemons into lemonade and figured out what is necessary to get those results, based on what we found. We hope this is a useful lesson not just for sticking to the scientific journey even through the deepest of challenges, but also a useful life lesson that even when things look grim, there is always some way to make the most of it and push forward to the light.