Phase transitions are quite possibly one of the most fascinating areas of physics: different substances behave differently at various temperature and pressure points -- sometimes in very remarkable ways. Yet it's easy to lose one's sense of wonder, because we see phase transitions around us every day. Water boils, ice melts (diluting the taste of one's otherwise killer cocktail) -- no big deal, right? Wrong. Sure, you can produce a phase transition of sorts for just about any common material, but there's still a lot to learn about the underlying physics.
Case in point: Earlier this summer, an article appeared in Physical Review Letters (July 7, 2006), describing the results of a new computational study by scientists at Sandia National Laboratories indicating that a new conducting form of water -- dubbed "metallic water" -- could occur at a temperature of 4000 degrees Kelvin and a pressure of 100 gigapascals. Can we just say... yowza! Those are some high temperatures and pressures! And yet they are much lower than previous theoretical estimates calling for temperatures of 7000 degrees Kelvin and 250 gigapascals. (Confession: We do not actually know what a gigapascal entails, but we like the sound of it, and anything with "giga" as a prefix isn't exactly small.)
I meant to write a blog post about this whole metallic water business back in August, but as often happens, time got away from me. Also, the news broke while I was vacationing in Buenos Aires, so I was rather late in the game. My notes have been sitting in the blog fodder file ever since. Fortunately, I have a second chance to address the topic: the same researchers were on hand to present an update on their paper at the annual meeting of the APS Division of Plasma Physics, held this past week in Philadelphia. The authors -- Thomas Mattson and Mike Desjarlais -- summarized their earlier findings from their quantum simulations, which were sufficiently different from previous calculations to warrant some significant alterations in the standard theoretical phase diagram for water under such extreme conditions. (A phase diagram, for non-scientifically inclined readers, is just a graphical representation of the effects of pressure and temperature on the phase of a substance. Go here to find out how to build your very own phase diagram from scratch. We make our own fun here at Cocktail Party Physics.)
Residents of Washington, DC, might assume "metallic water" refers to things like high lead levels in the drinking water, when in fact it describes a form of water that is electrically conductive. It wasn't immediately clear to me what's so special about this, since water is pretty conductive in its normal state, but Jen-Luc Piquant reminded me that very high temperatures and pressures can give rise to new and intriguing properties for even a common substance like water -- in this case, ultra-high energy densities. How high? Far beyond those that would occur naturally anywhere on earth. In fact, it would take temperatures and pressures on a par with those believed to exist in the interiors of the gas giant planet Jupiter to produce equivalent energy densities in a lab like Sandia's Z-pinch machine.
We should have known this was coming. Back in 1996, William Nellis, a scientist at Lawrence Livermore National Laboratory, announced the successful achievement of metallic hydrogen, which, among other things, gave us some fascinating details about the compressed interior of Jupiter. (You can find out more than you probably ever wanted to know about metallic hydrogen here, along with some really nice graphics.) Hydrogen as we know it is a gas, but on Jupiter, it's believed to exist as a super-hot liquid metal because of the extreme pressures and temperatures that typify that planet. Eugene Wigner predicted back in 1935 that if you squeezed hydrogen gas hard enough, it would eventually metallicize, but the requisite pressure was so intense that physicists weren't able to achieve it for 60 years. Hydrogen being a primary component of water, it stands to reason that one day physicists would move on to exploring the possibility of metallic water. (Jen-Luc loves the fact that both these phases are known as "degenerate matter." Clearly the term means something completely different in physics.)
The usefulness of these new results isn't limited to the far reaches of our solar system. The whole point of doing the quantum simulations in the first place was to better understand the conditions deep in the heart of the Z machine, which emits incredibly powerful X-rays in very short (nanosecond) bursts -- the equivalent of many times the electricity generation of the entire world. Sandia uses water both as an electrical insulator and as switches. Zap water with that much electricity, and you'll vaporize it, causing the surrounding water pressures to rise as the resulting shock wave spreads outward. The Z machine exploits this effect to achieve its impressive X-ray energies.
The operators send electrical pulses of about 20 million amps through the row of water switches. Initially, the water acts as an insulator, but eventually it's overcome as the incoming charge builds up. At that point, the water transmits the pulse, in the process shortening (or compressing) it from microseconds to nanoseconds. Voila! Extremely high temperatures and pressures ensue, in the form of a really, really big "spark." The Z machine is undergoing a major upgrade, and Sandia scientists want to understand how their water switches work at deeper, "first principle" levels, so they can attain the optimal transmission of electrical pulses in the future. In the same way that one can never be too rich, or too thin, physics experiments can never really be too high in energy -- because higher energies invariably translate into discovering new and interesting physics.
Sandia's on a roll these days when it comes to weird temperature and pressure phenomena. Also on hand in Philadelphia this week was Sandia scientist Marcus Knudson, who uses strong acoustic shock waves to melt diamond -- more precisely, to determine its melting point. Diamond is one of the toughest substances known, standing its ground in defiance of all manner of onslaughts, but apparently it goes weak at the knees when confronted with the dulcet tones of Knudson's sound-speed technique. Okay, the tones are hardly dulcet: Knudson and his colleagues are producing acoustic shock waves with strengths between 6 and 10 Mbars, producing a mixture of molten carbon and solid diamond. I'm casting about for an appropriate analogy that a non-scientist can relate to, and coming up empty. Suffice to say that melting diamond completely requires shock waves stronger than 10 million times earth's atmospheric pressure.
Why does this happen? Just like with the Z machine, the shock wave transfers large amounts of energy to the diamond material when it strikes, increasing not just the pressure, but also the temperature. If the shock wave is powerful enough, the temperature will get so high, the diamond will begin to melt into liquid carbon. The Sandia scientists determined the "melting properties" of diamond by measuring the speed of sound in the shocked material. Basically, the strength of the sound wave correlates pretty neatly with whether the material in its shocked state is in solid or liquid form. They observed a steep drop in sound speed at the 6 and 7 Mbar levels -- a sign the diamond was beginning to melt.
Knudson's work is particularly relevant to continuing research on inertial confinement fusion (ICF), which requires a material like diamond or beryllium for the fuel capsules used to produce fusion. For ICF purposes, a low melting point is desirable. So diamond's high melting point is a bit of a disappointment in that respect: beryllium has similar properties, but a lower melting point, making it the preferred material for ICF applications.
And here's another fascinating temperature-related tidbit gleaned from the pages of an upcoming issue of PRL: slo-mo boiling. Vadim Nikolayev of the Ecole Superieure de Physique et de Chimie Industrielles in Paris thinks he can explain why certain industrial heat exchangers -- such as those commonly used in power plants -- occasionally experience a "boiling crisis." [Hat tip: AIP's Physics News Update] That's what happens when the steam gets so intense, the machines actually, um, melt. (Jen-Luc shudders to think what might happen were the heat exchangers subjected to the Sandia team's acoustic shock waves.) It's a bit like what happens when a water droplet hits a hot frying pan: it evaporates very slowly. Boiling is generally considered a form of accelerated evaporation by scientists who study the phenomenon, characterized by highly efficient energy transfer: energy moves from a heater to a liquid via the formation of vapor bubbles. Things get dangerous when the temperature gets too high: so many bubbles form that the entire heating element gets covered in a vaporous film, preventing the liquid above from absorbing heat. So heat builds up, and the machine melts down. You can see the process in action here.
Scientists know that much, but they still don't understand the "boiling crisis" problem sufficiently well to devise effective counter-measures. Enter Nikolayev and other colleagues from the Commission of Atomic Energy in Grenoble and the University of Bordeaux. They simulated the boiling crisis to test a theory that the meltdown occurs as a result of vapor recoil. Just like the thrust from a rocket blast, a bubble will grow under intense heat and push aside any liquid near the heating element, giving rise to that dangerously insulating layer of vapor. Subsequent experiments confirmed the theory. The cool thing -- literally -- is that those experiments weren't performed at skin-blistering temperatures, but hear the critical temperature of liquid hydrogen: 33 degrees Kelvin. That's because boiling at normal thresholds is so rapid, it's difficult to make precise observations. At 33 degrees Kelvin, boiling happens much more slowly. And since the laws of fluid dynamics are universal, they should be able to extrapolate the same basic principles of behavior all the way up to fluids at 100 degrees Celsius.
Similarly, the lessons learned for large-scale industrial plants can be scaled down to the individual consumer level, perhaps to address the "boiling crisis"that occurs in, say, in exploding laptops. And you should definitely care: the smaller those electronic components get and the more densely they're packed, the higher the rate of heat dissipation will be. Heck, my spiffy new MacBook Pro is a great machine, but I've noticed that it does run a bit hotter than the older Power Book That Died (when the hard drive had a meltdown). I've invested in a "portable insulated laptop computer desk" (a.k.a., the Lapinator) accordingly, which mitigates the problem a little. At least it hasn't exploded or caught on fire. Yet.
There is a point to this typically lengthy and rambling post, if you recall: Phase transitions are cool. (I didn't say the point was especially profound.) They are still, centuries later, giving rise to cutting-edge science -- like telling us more about the composition of Jupiter, or the potential for achieving a practical fusion energy source -- and yet also helping human beings stay cool as they obsessively blog and surf the Web. So thanks, Sandia scientists, Dr. Nikolayev (and cohorts)!