So I’m in a solid state lab, and we use magnets for lots of things. Whenever we have a reaction batch, trying to make more of a ferromagnetic compound, I always separate those crystals from the rest of the stuff with a Gd-based magnet. That way I get big shiny crystals for diffraction, transport measurements, etc.

Stirring?

Mostly kidding. I was curious as to how they stirred their solutions, but they used an orbit shaker instead.

Awesome paper! I wonder if this would be relevant on an industrial scale; imagine being able to skim off the product of your reaction without any further purification!

This paper uncannily resembles a theoretical idea I had a few years ago and wrote it as a paper for P-Chem class; and the professor sorta laughed at it.

Anyway, we use them in lab all the time too. Our cabinets are metal, and so we buy cheap long legged stuffed creatures, stick some magnets in the feet and throw them at the cabinets and see the hilarious and entertaining ways they stick.

Ohh, you mean for science?

I was thinking about posting that paper too. It looks like it could be useful. Even if it isn’t, it’s cool!

Its really interesting info, thanks mitch. keep going with interesting blogs, I really enjoy reading ur blogs, awesome.

George Whitesides’ papers are, most of the time, real forehead-smackers: the best kind of paper. Why didn’t anyone else think of that? The group’s chemical intuition is uncanny.

I have nothing intelligent to contribute except:

This is really cool. (It really is)

This is a neat paper, but the conclusions presented by Whitesides are rather misleading. First of all, despite what the authors claim in the last paragraph of the paper, this method would be a very tedious way of monitoring reactions. Note that they do not monitor the reaction mixture directly–instead they take an aliquot of suspended beads, then wash them 6 times (3x w/2 solvents) and then dry them for 30 minutes under vacuum (see the SI). It is at this point that they suspend them in a fresh cuvette of GdCl3 solution and wait 15 minutes for the beads to settle. Thus, when the authors claim “[the technique] is rapid (measurements require 15 min)” they have (conveniently) forgot to account for all of the washes and the 30-minute drying step. That is a lot of time!

I suspect that these steps are crucial, because it appears that the levitation height is *very* sensitive to what molecules are appended to the beads (that was the first half of this paper). If you carry solvent (from swelling) or physi-sorbed reagents over, you will surely get an anomalous reading.

The authors also claim that the technique is less expensive than NMR, but that is only the case if you need to buy an NMR to run your experiment. Most departments have NMRs, and it is cheaper to buy NMR time than it is to go through copious wash steps and buy a magnetic levitation set-up. I imagine the cost of CDCl3 is comparable to GdCl3, too.

Finally, this technique would only be convenient for known reactions. That is, reactions where you know the density of the starting beads and product beads. The authors note that they have to fiddle with the concentration of GdCl3 (or the applied field) in order to find a convenient “window” for the vertical displacement of the beads. Thus, you may get lucky with the concentration of Gd you select for running a new reaction, but you might not. A common case would be: too much Gd and the beads will stay near the top and not move; too little Gd and the beads will sink completely before they are done reacting.

So, it’s a neat paper, but it is *waaaaay* oversold by the authors (and in a misleading manner, too). I think most people will stick with TLCs and NMRs and be wise for doing so.

Are you sure it is a Gd-based magnet and not a Nd-based magnet?

I am not disagreeing with a lot that you say, I am just pointing out that maybe solvent and usage wise it is comparable to NMR, but NMR’s require tens of thousands of dollars a year in maintenance cost just to keep them running.

I use my magnets to levitate frogs. It’s a convenient storage option when we run out of aquaria.

Yes, NMRs are expensive to maintain, but my point is that they are a sunk cost. The department is going to pay that money anyway–magnetic levitation can’t be used for everything–and NMR time is usually not that expensive (at least, relative to going through all the fuss of using magnetic levitation as an alternative). Also, an NMR spectrum (or even a TLC) will give you so much more information than a levitation height.

I use my magnets to try and lift steel. So far it’s very hard (I want to create light-weight body armor parts).