Scientists Finally Solve the Mystery of Why Comets Glow Green (popsci.com) 13
A team of chemists just solved the mystery of why comets' heads -- but not their tails -- glow green, which had puzzled researchers for decades. From a report: Studying an elusive molecule, which only fleetingly exists on Earth, was the key.
Comets are speeding chunks of ice and dust left over from the formation of the solar system, which occasionally venture from the system's cold outer reaches to pass by Earth. Back in the 1930s, Gerhard Herzberg, who later won the Nobel prize for his research on free radicals and other molecules, guessed that the process behind the green comet glow might involve a molecule made from two carbon atoms bonded together, called dicarbon. A new study, published in the journal the Proceedings of the National Academy of Sciences, put Herzberg's theory to the test.
Dicarbon is so reactive that the team behind the study couldn't get their supply of it from a bottle, says Tim Schmidt, a chemist who oversaw the study at the University of New South Wales in Sydney, Australia. In space, it exists inside stars, nebulae, and comets. But when exposed to the oxygen in Earth's atmosphere, dicarbon will quickly react and "burn up," Schmidt says. Schmidt says this is the first time scientists have been able to examine precisely how the molecule breaks apart when exposed to powerful ultraviolet rays. In the lab, the team had to simulate the environment of near-Earth space with vacuum chambers and three different ultraviolet lasers. Because dicarbon reacts so quickly, they had to synthesize it on the spot by whittling away a larger molecule with a laser.
Dicarbon is so reactive that the team behind the study couldn't get their supply of it from a bottle, says Tim Schmidt, a chemist who oversaw the study at the University of New South Wales in Sydney, Australia. In space, it exists inside stars, nebulae, and comets. But when exposed to the oxygen in Earth's atmosphere, dicarbon will quickly react and "burn up," Schmidt says. Schmidt says this is the first time scientists have been able to examine precisely how the molecule breaks apart when exposed to powerful ultraviolet rays. In the lab, the team had to simulate the environment of near-Earth space with vacuum chambers and three different ultraviolet lasers. Because dicarbon reacts so quickly, they had to synthesize it on the spot by whittling away a larger molecule with a laser.
Truncated summary (Score:5, Informative)
The next line of the summary was:
They confirmed comets’ green light comes from the dicarbon molecules, which, when exposed to sunlight in space, can absorb and emit visible light, Schmidt says. Herzberg was right about dicarbon, he says, if not quite correct about the mechanism–but it was the 1930s, Schmidt says, so “he can be forgiven.”
It was solved in the 1930s (Score:5, Informative)
Re: It was solved in the 1930s (Score:2)
Certainly an enormous amount of credit should be given to Herzberg (as a spectroscopist myself I see him as one of the real founding scientists of the field). But a theory is right when it's experimentally proven, so I'm glad they were able to prove it in the lab and kudos to them for doing it.
Folks like John Maier at Basel have spent years studying atmospheric carbon chains and I'm always amazed at the fact that they even exist.
Re: It was solved in the 1930s (Score:2)
Sorry, I said "atmospheric carbon chains", but I meant "astronomical carbon chains"
Re: (Score:3)
Can't we do this in software? (Score:1)
A few weeks back there was an article about a team of biochemists using time on one of the government supercomputers to simulate a sars cov 2 virus in a droplet of water for ten nanoseconds.
It seems to me that calculating the behavior of dicarbon and other exotic molecules that might appear in a comet's tail would require far less computer time.
Putting my pedantic hat on even tighter (ooh yeah that's the stuff) I would say that creating and measuring the behavior of dicarbon in the lab isn't quite the lynch
Re: Can't we do this in software? (Score:5, Informative)
That's a great question that I may be able to answer. The vast majority of protein folding computations are based upon classical mechanics (in the field, called "molecular mechanics"), where the charge distribution (e.g. dipoles, quadrupoles, etc), dielectric constants, bond lengths, strengths, etc are estimated, then classical (non-quantum) calculations are propagated to simulate the folding. The incredibly hard part about something like proteins folding is that the potential energy surface is huge, with a crazy number of minima. So finding a true global minimum, or even lowest energy path, is very hard. Adding in a solvent like water makes it even harder. There are some calculations that are a hybrid of quantum and classical mechanics ("QM/MM"), where something like the solvent is simulated with the classical equations, but quantum mechanics may be used to describe a few pieces of the protein that the scientist thinks may need to be treated as such. The point is that something like protein folding (or large molecule behavior in general) is frequently well-described by classical equations. There are exceptions of course.
Spectroscopic calculations are, by definition, quantum in nature. You must solve quantum equations, which is really only possible for H and H2+, systems with 1 electron. Everything with multiple electrons requires quantum simulations that have inherent approximations that model the electron-electron interaction (starting with Hartree-Fock). As you try to make those approximations more accurate, your computational effort increases rapidly. Making that molecule a radical like C2 makes it yet harder.
Chances are folks have simulated the C2 spectrum for years, but it hadn't been actually measured. I bet the calculations were pretty good, but you only know for sure when you actually measure it.
Re: Can't we do this in software? (Score:1)
Well, I buy that quantum modeling is more complex than classical models, but if we're talking about dicarbon, with a grand total of two nuclei and 12 electrons interacting with either nothing or a single photon, surely those equations better behaved than a purely classical model of a billion atoms?
I mean, heck, even classical models of pointlike beads on ideal elastic wires become "tricky" when there enough beads such that the ratio of min to max eigenvalues start to approach the dynamic range of floating p
Re: Can't we do this in software? (Score:3)
I think it's a matter of the level of accuracy involved. Yeah, you can pretty easily determine that it will absorb in the visible somewhere, and probably generally where it would absorb. So my guess is that people were able to say it would probably give you the green color based upon calculation. Can it get you accurate to 50 nm, 5 nm, 0.5 nm? Depends on how much computational time to put into it. However, these calculations have a history of being unreliable, and the spectroscopy field uses even "simple"
PNAS link (Score:3)
Now to discover dilithium (Score:1)
Certainly then it can be mined from other planetary systems, and subsequent empires based upon.