Table-Top X-Ray Lasers

ResearchBlogging.orgI mentioned in a previous post that one of the cool talks I saw at DAMOP had to do with generation of coherent X-Ray beams using ultra-fast lasers. What's particualrly cool about this work is that it doesn't require gigantic accelerators or nuclear explosions to produce a laser-like beam of x-rays-- it's all done with lasers that fit on a normal-size optical table in an ordinary lab room.

The specific talk I saw was by Margaret Murnane of JILA, who co-leads their ultra-fast laser group, and dealt with a new technique for producing soft-x-rays (~500 eV photons) with ultrafast lasers. We'll do this as a real quick ResearchBlogging post regarding the PRL on the subject (freely available as publication #8 on their 2009 publications list), in Q&A format:

So, what did they do, again? They used infrared laser pulses whose width was only a few tens of femtoseconds (1 fs = 0.000000000000001 s) to generate x-ray photons by blasting them into a hollow optical fiber filled with noble gas. They put in three millijoules of energy per pulse, and got out x-rays with an energy of 500eV per photon (a bit less than 10-16 joules, so this isn't violating conservation of energy, even though the change in units makes things look weird). These have a wavelength of around 2.5 nm, almost 1000 times smaller than the laser they were using.

So, wait, if they put in infrared lasers, how are they getting x-rays out? The laser pulses they use are so intense that they ionize the atoms. It's easier to think about the effect of the laser as being like a classical electric field, which strips electrons off an atom-- in quantum terms, you're adding up a bazillion photons to get the electrons out. Later on, when the electron comes back to the atom and recombines, you get all that energy back out at once, hence the x-ray photons out.

If this is like a classical field, how does this have laser-like properties? The pulses they use oscillate multiple times, so you can think of the light field as first stripping the electrons off, then forcing them back to the original atoms, where they recombine. This means that the emitted x-rays all have the same phase, and you get a coherent beam out.

The hollow fiber is also important for this, as it "guides" the laser pulse along through 2cm of gas, keeping the pulse intensity high and generating more x-rays than they otherwise would.

This sounds awfully simple. Why didn't anybody think of this sooner? The Kapteyn-Murnane group has been doing this sort of high harmonic generation for several years now, and slowly pushing the output photon energy up. This recent work makes a couple of refinements, and benefits from serendipity.

They don't get into it much in the paper, but in her talk, Murnane said that the key to getting over 500 eV was using a longer wavelength laser. They hadn't done this before because a simple picture of the process suggests that it's harder to get coherent x-rays out using longer starting wavelengths. When they did more sophisticated models of the process, though, they discovered that the interaction between the laser and the ionized gas produced by the pulse leads to some effects that make longer wavelengths more effective.

The key factor is "self-compression," where the ions formed by the pulse lead to a different propagation speed in the medium for different laser frequencies, which causes their pulses to become shorter inside the gas than out-- 19 fs instead of 38 fs. That shift also corresponds to a larger bandwidth for the laser. When you take this self-compression into account, the high-energy photons add coherently to make an intense beam, while without the self-compression, they would interfere destructively and prevent the beam from building up to useful intensity.

That's pretty lucky, isn't it? Yeah, you don't get a lot of those moments in science.

So, what is this good for? The main interest in this sort of thing is for imaging. If you can generate a coherent and fairly intense x-ray beam, you can use it to do diffraction measurements the same way you do with an optical laser. This lets you make images of features that are as small as a few nanometers, much smaller than you can do with anything in the visible light spectrum.

These are also extremely short pulses of X-rays, so they talk about doing things like taking pictures of chemical reactions as they happen, and imaging processes that take place in single cells. They've got a bunch of papers on this sort of stuff on their publications page, including both biological processes and materials science applications.

One of the great features of this is that the beams are produced with table-top apparatus and in a highly directional fashion, so you're not just spraying x-rays all around the lab. The beams involved aren't big enough to do human-scale imaging, though, so this won't be replacing the x-ray machine in your doctor's or dentist's office any time soon.

Arpin, P., Popmintchev, T., Wagner, N., Lytle, A., Cohen, O., Kapteyn, H., & Murnane, M. (2009). Enhanced High Harmonic Generation from Multiply Ionized Argon above 500 eV through Laser Pulse Self-Compression Physical Review Letters, 103 (14) DOI: 10.1103/PhysRevLett.103.143901

More like this

Having played with X-ray curable resins, this might be a neat way to make nanostructures.

Novel application for dispersive propagation in a plasma. This is one of those things that seems obvious -- once someone has spent a lot of time figuring out not only the idea but how to tell it to you :D

Kudos to them.

I saw a talk on this general area (actually titled "Attosecond Physics") at Photonics West back in January. It's a really cute bit of physics. One of the things the speaker showed was an image of atomic electron orbitals -- yes, actual pictures of those S and P orbitals we all learned about, generated by X-ray interferometry; the electron density phase-shifts the X-rays. (That's about all the detail I recall, though I can find the reference if anyone really wants it).

One thing the speaker said several times is that it's really easy to make coherent X-rays this way; the hard part was getting lasers that would put out transform-limited fs pulses -- at which point he would wave his hand in the general direction of the (huge) Photonics West exhibit hall and say, "but now you can just go next door and buy those!" as if it was a source of continuing amazement to him.