read more journal articles

One way to read lots of papers is to volunteer to be a “crash reviewer” for a conference. Had to review a dozen or so speech papers in a weekend. I was surprised how easily I could get the gist of them, even the ones in bad Taiwanese english with all the “the”s in the wrong places. Doesn’t quite satisfy this goal though – I need a habit of regular reading…

RSS feeds help A LOT for this goal, as journals are all online, now. I recommend doing both RSS and having the titles/abstracts e-mailed to you when the issues come out. It makes the whole thing much more managable.

This is probably one of the most useful things you can do as a grad student: read, read, and read. Don’t worry if you can’t understand everything; I think when my amazingly smart boss told me that he understands from 0-50% of any article, that’s when I realized that it’s best just to go through and get the broad strokes. If you really need to know things in detail, you’ll learn it then.

Consequently, I feel like I now have a ton of direction for my last year of grad school, I feel I know who is doing what right now, and all the rest. Quite worth doing!

Sprained my ankle in a really stupid way trying to snowboard on the slopes of Stowe, VT on Saturday and so spent the following afternoon today in the ski lodge pissed off, drinking and reading research articles. I’ll write in the names and references later ‘cuz I don’t feel like hobbling to my bag to dig ‘em up.

I’ve definitely noticed that there’s a strange curve to this. The first beer increases your comprehension (at least mine), but all subsequent ones really make it suffer, and this downturn is very sharp and sudden. Interestingly, I’ve heard it said that a small amount of alcohol (or even being totally blitzed) will allow you to speak a language you are insecure with a little bit better.

Tried to read an article on String Theory published this week in Physical Review Letters (PRL 96, 031301 (2006)) that describes the birth of the Universe and the fact that it’s likely two Universes were created simultaneously. It had this nifty image to describe it:

I took a class on String Theory, but it was at a (relatively speaking) basic level and I couldn’t get much out of this paper beyond the fact that this interpretation avoids some mathematical singularities. String theorists inhabit a different planet, methinks (perhaps in this other Universe they’ve postulated)...

Have read a couple more (a PRL and a PRE). Some of the things that get published are so suspect…

Have been really bad about writing the summaries, though. I definitely get a lot more out of them when I do that.

In the past two days, I’ve read six (I think). 4 PRL’s, 1 JCP, and 1 PRE. Understood them to various degrees. This is really one of those things that gets easier the more you do it (it’s that whole learning curve thing).

Science 302, 849 (2003)

Basic points:

1. Fragility is generally measured above the transition temperature T_g and typically correlates with the strength of the boson peak, the stretching parameter of nonexponential relaxation, and the temperature of the shear modulus.
2. Looking at the non-ergodicity factor f_, one can identify a factor that controls how fast it decreases as _T goes up.
3. A lot of effort is spent linking f to the dynamic structure factor, which is measured by inelastic X-ray scattering.
4. The harmonic approximation is taken for vibrational dynamics and this leads to a linear relationship between the inverse of f and the temperature. The slope of this line, \alpha, is directly proportional to the fragility. So, this is a link between glass properties and liquid state properties.
5. More generally, it means that in the potential landscape picture of glasses, the structure of the local minima that the system falls into upon vitrification is related to other, more global properties of the potential energy surface.

References to look up:

1. PRL84, 5788 (2000)
2. PRE66, 031205 (2002)
both for the link between the harmonic approximation and the structure factor.

Ideas:

1. We can measure the non-ergodicity parameter, but quite poorly in contrast to IXS techniques. I thought it was shown to display a sqrt behaviour, as predicted by MCT… Might be nice to see if some of these old data (and new ones yet to be taken) can correlate with these IXS results.

PRE 69, 062501 (2004)

Things of interest:

1. for relaxation dynamics, vitrification
van der Waals liquids: T and V equivalent in importance
polymers: T slightly dominant
strongly H-bonded: T clearly dominant
2. this T¹V^\gamma model fits 9 decades of frequency in relaxation dynamics!
3. The ratio of Enthalpies: H_V/H_P (which must be a measure of how much energy goes into changing the local configuration and isn’t stored for physical work in the two different cases, I think) goes to 1 for T dominance, 0 for V dominance.

This article wasn’t much different from the last one. But, there are a couple of references to chase when time permits:

1. J. Phys. Chem. 91, 4169 (1987) for enthalpy ratio must behave as it does (though it kinda makes sense from the way it’s defined through derivatives).
2. Contemp. Phys. 12, 339 (1971) for why interaction potentials go as r^{-3\gamma}.

Ideas:

1. It would be nice to see if this analysis could be linked to an analysis of elastic moduli, which are also variables we can measure. This would also be interesting as the shock experiment is able to change the volume in a time that may be shorter than some of the relaxational modes (at least for the alpha process)—and we recover an elastic modulus from that experiment. By inhibiting this kind of relaxation, would the glass be fundamentally different somehow? Would the fragility be inherantly different, and how? Would it behave “stronger” because V doesn’t have as much of an impact as it would normally?

Next article: “Correlation of fragility of supercooled liquids with elastic properties of glasses.” It’s kinda long, so it may take two days…

Things of interest:

1. Generally speaking, strong liquids primarily vitrify by jamming, weak liquids are more thermally activated. Hydrogen bonded liquids are likely to deviate from this behaviour.
2. fragility is a measure of the loss of local structure with increasing T. Strong liquids maintain short range order longer to higher T. Opposite for fragile ones. This is why strong are preferable for glasswork.
3. Can quantify relative influence of T and spec. vol. using a scaling function of form log(\tau)=F(Tv^\gamma). \tau is a relaxation time (here dielectric relaxation). \gamma is material specific and can be interpreted as measure of steepness of intermolecular potential. \gamma->0 fragile liquids, \gamma large for strong.

GET REF 21: PRE 69, 062501 (2004)

4. Put it together: dominant short range repulsive potential implies stronger dynamics, and a more harmonic potential.

ideas:
1. Would be a great way to analyze \tau from DAC ISTS measurements, also see if dielectric relaxation measurements (probing rotation degress of freedom) differ from our mechanical relaxation measurements.
2. Fragility should also correlate with thermal diffusivity as it apparently correlates with the anharmonicity of the potential. We should be able to test this assertion, too.

Next: “Thermodynamical Scaling of the Glass Transition Dynamics”

Am going to start my article regimen tomorrow. First victim: “Why Liquids are Fragile.” I’ll then write some notes from it on this website and give a brief summary!