Laser cooling in a Penning trap

[kelly0303]

Hello! I tried to find some papers about laser cooling in Penning traps, but all I could find were theoretical papers, no experimental ones. Has laser cooling been done in a Penning trap? And if no, why so? I assume that sending the lasers is more complicated than in a Paul trap, but there are cylindrical Penning traps that would give enough space for the lasers to interact with the ions. Thank you!

 

[EigenState137]

Greetings,

Laser cooling of externally produced Mg ions in a Penning trap for sympathetic cooling of highly charged ions

Resolved-Sideband Laser Cooling in a Penning TrapES

 

[Twigg]

Yep, as @EigenState137 showed above, you most certainly can. Usually the cooling laser comes in parallel to the B field, as in the first link.

In experiments where field uniformity is critical, sometimes you see ions shuttled between a cooling trap and a precision trap. If you need horizontal optical access (which messes up field homogeneity), you do it in the cooling trap. Once things are nice and chilly, you shuttle them over to the precision trap to do your fancy experiment with well-behaved fields.

 

[kelly0303]

@EigenState137 @Twigg Thanks a lot, this was really helpful! Do you know if this was also done on molecular ions, too?

 

[EigenState137]

@EigenState137 @Twigg Thanks a lot, this was really helpful! Do you know if this was also done on molecular ions, too?

Greetings,

I do not know about molecular ions. The physical conditions that must be satisfied for magneto-optical trapping to be effective are non-trivial. It would not surprise me if the additional degrees of freedom within molecular ions adversely affect the desired process.

It might be helpful to your readers if you could provide a more expanded description of your interests in these techniques. Laser cooling and ion trapping are after all rather widely used technologies. Just a thought.ES

 

[Twigg]

@EigenState137 I think what @kelly0303 is suggesting is to penning-trap molecular ions, then laser cool them. I don't think you need to MOT the ions for that, but I agree that the additional degrees of freedom are problematic.

I haven't heard of anyone doing molecular ions in a Penning trap. There are molecular ion expeirments in Paul traps (HfF+/ThF+ eEDM searches comes to mind, but it's a bit of a stretch to call that a Paul trap). The same principles of optical cycling apply in a Penning trap.

Are you thinking of a single-molecule experiment? If so, you might consider sympathetic cooling with a co-trapped atomic ion. Heck, maybe you could even do quantum logic spectroscopy, and avoid the minefield that is molecular structure (almost) entirely. There was a proposal for doing just that recently, but designed for protons and anti-protons not molecular ions. Here's the proposal, and here's a summary I put up on PF a little while back.

Edit: actually, now that I think of it, I don't think I've ever heard of folks MOT-ing any ions (not even atomic ones)? I seem to recall there are problems with shimming the electric fields... I can't remember the details though.

 

[Twigg]

Oh, if you do go the all-out brute force method of laser cooling molecular ions in a Penning trap, there was one cool idea I remember. I forgot who came up with this though... Darn. It was one of the big wigs for molecules in AMO. The idea was that you could repump a atom/molecule that has lots of loss channels by shining a frequency comb at it, and using absorptive filters to remove the teeth of the comb that excite the ground state. Essentially, you're driving every transition in the molecule EXCEPT the ones that de-populate the ground state, so as things spontaneously emit downwards, you'll eventually trickle back to the ground state. I don't know if anyone's actually done this, I just remember discussing it with my old group.

 

[kelly0303]

Oh, if you do go the all-out brute force method of laser cooling molecular ions in a Penning trap, there was one cool idea I remember. I forgot who came up with this though... Darn. It was one of the big wigs for molecules in AMO. The idea was that you could repump a atom/molecule that has lots of loss channels by shining a frequency comb at it, and using absorptive filters to remove the teeth of the comb that excite the ground state. Essentially, you're driving every transition in the molecule EXCEPT the ones that de-populate the ground state, so as things spontaneously emit downwards, you'll eventually trickle back to the ground state. I don't know if anyone's actually done this, I just remember discussing it with my old group.

Thanks a lot for this. Yes you are right. I know that people laser cooled molecules (neutral or ions), and I know that people trapped and laser cooled ionic atoms in Penning trap. I was wondering why no one combined the 2? Is it just that you gain nothing compared to a Paul trap? Or is it just too difficult?

By using a logic ion you mean laser cooling the ionic atom which in turn cools down the molecular ion? Don't we need first to bring them down to low temperature by other techniques? Or does this work without laser cooling the molecule at all? Also if I were to do some high precision measurement on the one molecular ion, won't the other ion affect its properties? Or that is relatively straightforward to account for (e.g. gradient electric field due to the atomic ion)?

About MOT, I assume that using MOT with ions doesn't make much sense, as using electrodes is easier than light? Based on my really basic knowledge I would trap ions with E and B fields rather than laser light.

 

[EigenState137]

I haven't heard of anyone doing molecular ions in a Penning trap. There are molecular ion expeirments in Paul traps (HfF+/ThF+ eEDM searches comes to mind, but it's a bit of a stretch to call that a Paul trap). The same principles of optical cycling apply in a Penning trap.

Greetings,

This is why I suggested greater context for the initial question. Penning traps are certainly used in molecular ion experiments such as Fourier-transform ion cyclotron resonance mass spectrometry. I realize that leaves out the LASER cooling.

... Heck, maybe you could even do quantum logic spectroscopy, and avoid the minefield that is molecular structure (almost) entirely.

But its that minefield that is molecular structure that I am interested in ES

 

[Twigg]

I was wondering why no one combined the 2? Is it just that you gain nothing compared to a Paul trap?

You ever tried to keep 15 lasers locked at once while hopping on one leg and praying to the ion gods that your ablation target will last a few more days? Jokes aside, do you have a particular measurement in mind, or just interested in building it cause you can?

I would speculate that there's just no application for laser cooled molecular ions in a penning trap. Parity searches in molecular ions are hard because you need a rotating E-field to Stark shift/polarize the opposite parity states (like the HfF+/ThF+ experiments). I don't think you can make a rotating E-field in a Penning trap (watch someone make a fool of me for saying this). As far as quantum simulation/computing, molecular ions are worse than atomic ions or neutral molecules. Neutral molecules have big dipole interactions over short distances, but molecular ions have long-range coulomb interactions that prevent ions from getting close enough for dipole-dipole stuff to happen. Atomic ions interact strongly via coulomb forces just like molecular ions, but the atomic ions are waaaay easier to repump and just cleaner systems overall (in molecules (esp. heavy ones), some transition "rules" are more like "transition suggestions").

By using a logic ion you mean laser cooling the ionic atom which in turn cools down the molecular ion?

Yep!

Don't we need first to bring them down to low temperature by other techniques? Or does this work without laser cooling the molecule at all?

In the proposal paper I linked, the first round of cooling is done by resistive cooling in the penning trap. So yes, there is no need to directly excite the molecule at all!

Also if I were to do some high precision measurement on the one molecular ion, won't the other ion affect its properties?

You're right, it will. In that proposal paper I linked, they do the sympathetic cooling + quantum logic in a different trap configuration than they do the spin precession. In short, they bring the two much closer to do logic, than they pull them apart before doing the precision measurement.

About MOT, I assume that using MOT with ions doesn't make much sense, as using electrodes is easier than light? Based on my really basic knowledge I would trap ions with E and B fields rather than laser light.

Yep, you're right. It's much easier to use DC and/or RF E and B fields to control ions than it is to MOT them. The only people I know who tried to MOT ions attempted it as a flex.

 

[Twigg]

This is why I suggested greater context for the initial question. Penning traps are certainly used in molecular ion experiments such as Fourier-transform ion cyclotron resonance mass spectrometry. I realize that leaves out the LASER cooling.

I never knew the fancy name for that (ICR)! Thanks! And yeah, i meant to say ultracold molecular ions in penning traps.

But its that minefield that is molecular structure that I am interested in

Too much trauma for me! The wigner matrices haunt my dreams. That and all the different lasers it takes to cycle a molecule! Too many hours spent wiping laser dye off the floor

 

[EigenState137]

Too many hours spent wiping laser dye off the floor

Better off the floor than off yourself!

 

[kelly0303]

You ever tried to keep 15 lasers locked at once while hopping on one leg and praying to the ion gods that your ablation target will last a few more days? Jokes aside, do you have a particular measurement in mind, or just interested in building it cause you can?

I would speculate that there's just no application for laser cooled molecular ions in a penning trap. Parity searches in molecular ions are hard because you need a rotating E-field to Stark shift/polarize the opposite parity states (like the HfF+/ThF+ experiments). I don't think you can make a rotating E-field in a Penning trap (watch someone make a fool of me for saying this). As far as quantum simulation/computing, molecular ions are worse than atomic ions or neutral molecules. Neutral molecules have big dipole interactions over short distances, but molecular ions have long-range coulomb interactions that prevent ions from getting close enough for dipole-dipole stuff to happen. Atomic ions interact strongly via coulomb forces just like molecular ions, but the atomic ions are waaaay easier to repump and just cleaner systems overall (in molecules (esp. heavy ones), some transition "rules" are more like "transition suggestions").Yep!In the proposal paper I linked, the first round of cooling is done by resistive cooling in the penning trap. So yes, there is no need to directly excite the molecule at all!You're right, it will. In that proposal paper I linked, they do the sympathetic cooling + quantum logic in a different trap configuration than they do the spin precession. In short, they bring the two much closer to do logic, than they pull them apart before doing the precision measurement.Yep, you're right. It's much easier to use DC and/or RF E and B fields to control ions than it is to MOT them. The only people I know who tried to MOT ions attempted it as a flex.

Thanks a lot! One more question about resistive cooling (I will also read that paper too soon). In principle you can't cool down the molecule below the T of the external circuit. So do they cool it down to that temperature, decouple the external circuit, then cool it down further with a logic ion?

 

[Twigg]

Yep! The resistor is just like a heat bath, so if you leave it connected and start laser cooling, you won't get anywhere.

 

[kelly0303]

Yep! The resistor is just like a heat bath, so if you leave it connected and start laser cooling, you won't get anywhere.

I assume that disconnecting the RLC circuit would induce an electric field. Is that something that is well understood in general? Or are there some tricks to not have to deal with it?

 

[Twigg]

I've never worked a Penning trap, so I couldn't say for sure. If it was a Paul trap, you can always correct the lingering field out by imaging the ions, recording their position, and applying a "shim" voltage to the endcaps. There is probably a similar procedure for Penning traps (but without the imaging), but you'd have to ask someone who's done it.

 

[Twigg]

You know that feeling when you read a paper, and then realize something you said earlier was utter BS? That just happened to me

As far as quantum simulation/computing, molecular ions are worse than atomic ions or neutral molecules. Neutral molecules have big dipole interactions over short distances, but molecular ions have long-range coulomb interactions that prevent ions from getting close enough for dipole-dipole stuff to happen. Atomic ions interact strongly via coulomb forces just like molecular ions, but the atomic ions are waaaay easier to repump and just cleaner systems overall (in molecules (esp. heavy ones), some transition "rules" are more like "transition suggestions").

https://arxiv.org/abs/1909.02668
https://arxiv.org/pdf/2008.09201
I'm a goof. There's been theoretical work showing you can use phonons on the electrodes of surface traps to couple Paul-trapped molecular ions.

Also, there was a result from John Preskill's group that (theoretically) implemented error correcting codes in the molecular Hamiltonian:
https://arxiv.org/pdf/1911.00099.pdf

I'm going to show myself out the back now lol

As far as doing this with a Penning trap, I'm sure you could induce some coupling between polar molecule qubits in separate penning traps by connecting their endcap electrodes. I'm not sure what that'd look like or how strong the entanglement would end up being.

Also, I remembered where I got one of the things I said but couldn't remember:

he idea was that you could repump a atom/molecule that has lots of loss channels by shining a frequency comb at it, and using absorptive filters to remove the teeth of the comb that excite the ground state. Essentially, you're driving every transition in the molecule EXCEPT the ones that de-populate the ground state, so as things spontaneously emit downwards, you'll eventually trickle back to the ground state. I don't know if anyone's actually done this, I just remember discussing it with my old group.

It was Brian Odom's group, I think!
researchgate.net/publication/51667448_Optical_pulse-shaping_for_internal_cooling_of_molecules
https://arxiv.org/pdf/2007.15713.pdf