See shear waves in a supercooled liquid

So, I finally got them thar shear waves. It happened a week ago, on the 20th. All it took was for me to modify the laser I had been using for a while so that I could get more power but longer pulses (I bypassed the compressor) and voila, shear waves. They were weak, but they were there, and my boss confirmed that that’s what I saw: shear waves in Triphenylphosphite.

In the future, I’ll edit this victory lap post to show the shear waves. (Edit). OK. Put it in, and there they are at 126MHz.

I’ll also edit it to show me holding the jar of Nutella I bought a few days later after the shock wore off. Now that I can buy Nutella, I actually don’t even care for it that much, anymore!

Anyway, now comes the fun part: doing science with the shear waves!

But first I’m going to pop off to Korea and Hong Kong to meet my girlfriend’s family! While I have to analyze data I’ve been taking for a different project (I’m giving a talk on this data two days after I return!), it will be nice to know that, for me, the shear waves are there for the seeing.

You learn something new every day.

I was just talking to one of my colleagues, a postdoc who has been with us since January. He pointed out to me that the detection equipment we use is most sensitive to a wavelength of light put out by a different kind of laser than the one I have been using for, oh, the past FOUR AND A HALF YEARS. I looked, and indeed, it is six times more sensitive! D’OH!

But, I’ve been in this lab for five years now. So, I did use these other lasers back then… and I hated them, partly because I was too uneducated to know that they are actually better than the one I was using, but also because, being at the edge of the visible spectrum, they are hard to see, the output has a funny shape, etc.

So, here we go. In the past couple of weeks, I’ve determined that both of the lasers I use are not ideal for this experiment. And now I know otherwise. I now have a zillion different ways in which I can make the measurement I’m trying to do a lot more robust. I’d say that’s good news.

I spoke with my boss about it, and I also spoke with my colleague who is the main guy on the other laser, and it’s all a go!

This laser may REALLY be the one I needed. The pulse length is just right, and there is more energy than I ever got out of the last one. Too bad it just means I have to rebuild everything from scratch (including ordering parts).

I figured it out, ‘cause it broke on me while I was in Pullman, Asilomar, San Francisco, San Diego, and Seattle.

It’s because it shouldn’t be used for this experiment! The laser has been trying to tell me this by breaking all the time!

I need a good ol’ Titanium Sapphire system, not my Ytterbium based system… and we’ve got five or six of those kicking around the lab!

All I need is an electric field. With the large bandwidth of Ti:Saph, it would be easy to stretch the pulse out long enough to avoid unwanted nonlinearities and still remain in the impulsive limit to a shear wave. Plus, since the pump is at 800nm this has the added benefits of not being absorbed into either vibrational overtones or even via two photon absorption at 400nm. I would have to get THREE photon absorption to have the same problems I used to have with my laser’s green. Plus, I can access smaller grating spacings because the angles don’t need to be as big to get them.

Finally, these other systems seem to put out more energy than mine does!

No wonder! Shear waves and Nutella, here I come!

In my insomnia, I just realized something.

I am building a stretcher for IR for this experiment. The IR gets absorbed, needs reflective optics, and can’t get the large angles needed for the extremely high wavevectors I need to get this experiment to work.

When I first tried this experiment in spring of ‘05 (before the laser died), I tried doing this all with visible light, but the two photon absorbtion was too great and caused all sorts of non-linear things to happen that eventually would carbonize the liquids. Bad news all around.

I think it’s time to revisit this.

The nice thing is that the laser puts out a pump wavelength that is so close to the probe, I can use transmissive optics, which makes alignment trivial and the ability to go to large angles for big wavevectors correspondingly trivial. Large angles are essential for this to work because I need the large electric field gradients to do the stimulated Brillouin scattering.

If I’m smart about designing it, putting cylindrical and sperical lenses in the right place, then the diffraction efficiencies are good through a common 532 phase mask, and all the optics are common path, making vibrational noise A LOT less of a problem.

Plus, I can get a holographic diffraction grating from Newport for like $250 and only lose 40% of my light… which I can compensate for by focusing tighter. I can then stretch it out long enough to not do two photon absorption, but still be short enough to be impulsive w.r.t. the shear waves. It should work! So, this is a very viable alternate method.

I’m gonna order that grating next day air tomorrow and get this thing built ASAP.

Thing is that I still need to autocorrelate the pulse to see what I’ve done… but, I think that any old chunk of BBO will be able to double the 515nm light… I’ll just be at an inconvenient angle. Good thing I used silver mirrors (and not dielectric ones) for the autocorrelator.

Tried this yesterday. For once, the sample didn’t crystallize! It must have been because of the vacuum distillery that a friend of mine and I built. So, at least sample issues are largely taken care of.

Plus, the temperature stability was excellent, so at least I did a good job of learning how to use all this crap.

I didn’t see shear waves. For a second, I thought I did. But, it turned out to be some electrical noise at 60MHz. D’oh.

However, the longitudinal waves looked beautiful all across the glass transition, which was a definite positive step. If I can see that at high pressure, at least that will soften the blow if there are no shear waves forthcoming…

Given the experimental conditions last night, the only problem was that the sample didn’t have large shear waves. Everything else was working OK.

So, I had to give up the cryostat. But, we’re getting a couple of new ones! That means I’ll get back to it when I return from the West Coast on the 15th of August.

before I give it a rest until I get back from the West Coast in the middle of August (am leaving for work/conference/vacation in late July).

Liquid of choice: dibutyl phthalate. I may be setting myself up for failure here, because I can’t be too sure that this thing’s shear modes can be coherently driven or not. I know nothing of its polarizability, no less if there is a polarizability anisotropy (which is required for driving shear waves). The molecule doesn’t look very isotropic, so that’s at least a good start.

Oh, well. It’s what I put in the vacuum distillation apparatus that Josh and I built, so it’s what I’ve got to work with. At the very least, I could measure some longitudinal acoustics in it, as well as relaxational behaviour… having that data will at least be useful.

I’ve been getting really nice longitudinal wave data out of this silicone oil I’ve been doing experiments on (it also forms a glass when cooled) because I switched to a more sensitive detector and this is making a big difference. It’s highly likely that this data will go towards a publication.

The signals are large, and the laser powers are turned waaaay down and things still work great. My feeling is that I should be able to get some success out of this for the shear waves. Now if I can just get the parts I need to modify the cryostat, as well as build the stretcher/autocorrelator, I’ll definitely be all set!

I finally got all the optics for an octuple pass. My idea actually works! It doesn’t work as well as it could because the optics aren’t perfect… but if they were perfect, it would work quite well (or at least well enough). I still need to deal with a couple of issues of beam divergence and learn how to deal with fiber optics better, as I am using them as a means to erase the effects of the beam quality degrading over the octuple pass.

Nevertheless, this allows me to generate 32 nanoseconds of optical delay for the pulse-probe, which is A LOT. If I spend a day or two aligning carefully, I should be able to test-drive it with isopropanol in short order!

are on the horizon for next week while I work on building the stretcher. This will give experience with the cryostat and supercooling the liquids. Plus, it will help gather forward momentum and be relevant to the shear measurements, as well.