Hyperpolarization Update #5 (the creation of frankencoil)

A quick recap: I’m troubleshooting a setup to do hyperpolarization (a technique that allows us to take much faster measurements of samples using NMR). Our technique for performing hyperpolarization involves transferring nuclear spins from parahydrogen to our sample. Parahydrogen is a spin state of hydrogen that you make at low temperatures. Regular hydrogen will show up in an NMR measurement, but parahydrogen will not. The experiments that I am performing involve comparing the amount of signal we get from samples of regular hydrogen to the amount of signal we get from samples that we have attempted to enrich with parahydrogen. In order to measure the hydrogen I am pressurizing a solvent (acetone) with the gas to dissolve it in the solvent. If everything is working correctly, we should see a ⅓ reduction in signal (the signal doesn’t entirely disappear because we are only converting some of the regular hydrogen to parahydrogen).

I spent this week testing the new FeO(OH) catalyst. Our first catalyst (activated charcoal) was packed into an intricate double spiral of copper tubing, to allow for a very long path length so hydrogen would get cooled properly. For the new catalyst, I packed a short length of copper tubing with the catalyst and bent it into a “u” shape. This new setup has a much shorter path length than the original setup, but I didn’t think it would be a problem because I had read papers with similar setups and hydrogen (as a gas) has a very small heat capacity.

After extensive testing, I determined that the new setup didn’t work at all. This was very confusing because I had read a lot about how well the FeO(OH) catalyst worked. After some thinking, I ended up deciding to attach the new catalyst tubing after the old catalyst coil. The old catalyst coil was completely ineffective, so it wouldn’t catalyze anything, but it would provide a long path length for the hydrogen gas to cool before it passes over the FeO(OH) catalyst.

Frankencoil, as I dubbed the resultant monstrosity (see picture), passed extensive tests with flying colors! It consistently produces exactly the reduction in signal that I am looking for (see figure). After eight weeks, I’ve finally got some parahydrogen-enriched gas to work with!

This tangle of copper tubing haphazardly patched together is Frankencoil.

This tangle of copper tubing haphazardly patched together is Frankencoil.

 

Some of Frankencoil's test results. The key thing here is that all the red lines fall at or below the level of the green lines, which mark roughly the reduction in signal I am looking for. The red lines represent gas that I have attempted to enrich, the black lines are regular hydrogen gas, and the green lines are the black lines reduced by a factor of 1/3 for reference.

Some of Frankencoil’s test results. The key thing here is that all the red lines fall at or below the level of the green lines, which mark roughly the reduction in signal I am looking for. The red lines represent gas that I have attempted to enrich, the black lines are regular hydrogen gas, and the green lines are the black lines reduced by a factor of 1/3 for reference.

The next step is to actually attempt hyperpolarization. My eventual goal is to get this to work with single-sided NMR magnets, but I’m going to try to get it to work on traditional magnets first, since this has been done before. Also, traditional magnets make it easy to separate the enhanced signal from all the other signals my sample produces.

Comments

  1. jlpalumbo says:

    Hello Ruth!

    Fantastic work! Frankencoil is an amazing twisted network, and I am glad to hear it worked. The results from Frankencoil seem remarkably consistent, and, although I am not a chemist, your explanations make the data remarkably clear. As I understand it, the FeO(OH) charcoal catalyst failed in the standard double spiral shaped but also failed in the u shape. Do you know why this would fail while your Frankencoil succeeded? Was it the long path length, and if so how did that affect the chemical reaction? Overall this experiment looks fascinating, and I have really enjoyed watching it progress this summer.

    Thanks and once again great work,
    Jonathan Palumbo

  2. Hi Jonathan!
    Thanks for your comment! I think the longer path length cools down hydrogen gas to a lower temperature than a shorter path length. The longer the coil, the more time the hydrogen gas is exposed to the liquid nitrogen bath that the coil is submerged in, and the cooler it gets. This is important because the catalyst only converts hydrogen to its equilibrium state. At 77 K (the temperature of the liquid nitrogen bath) the equilibrium state is 50% enriched parahydrogen. However, if the gas is not given enough time to cool down and is instead passing over the catalyst at, say, 150 K (which is closer to room temperature) it will be converted to an equilibrium state that has much less parahydrogen enrichment. So in order for the coil to work, it needs sufficiently cold hydrogen gas. Since making the coil longer without substantially changing the catalyst fixed my problem, I think it was the path length.
    Thank you!
    Ruth Ann

Speak Your Mind

*