Wouldn't it be cool if this happens on a planetary scale somewhere. - Andrew Lang

How so, Andy? - Steve Koch

A search for "deuterium stratification" returns interesting results (though not quite like the one Steve shows here :) http://www.igsoc.org/journal... | http://www.sciencemag.org/content... - mkz

Which makes me wonder: If places like Lake Vostok (http://en.wikipedia.org/wiki...) turn out to have a higher concentration of heavy water due to stratification, should we expect to find microbes that have a higher tolerance for it there? - mkz

@Steve I'm thinking about a 'water' planet with a solid core, a layer of heavy water with heavy water 'islands', a layer of light water, then an atmosphere. I think it would make a great setting for a SF story. - Andrew Lang

I got curious and read a bit about the ice cores from Vostok. Turns out the deuterium content does not increase with depth. Come to think of it, the lack of heavy water/light water stratification should have been expected---after all, the ice did not form by the freezing of stagnant water; it probably came down as snow, and got packed over time. The heavier isotopes can go down in liquid, stagnant water and form a deep layer of heavy water, but such flow/diffusion is much slower in ice (ice is solid, after all :). Apparently the concentration of D is used as a proxy of the local temperature at the time of precipitation. Here is a paper that shows the density profile of deuterium in the ice cores from Vostok (430,000 years of Earth history = 2300 citations :): http://www.daycreek.com/dc... You can even download the data: ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/vostok/deutnat.txt - mkz

@mkz, at a mass difference of only 2 g/mol, the energy difference due to gravity of an H2O versus D2O molecule is pretty tiny. Something like equivalent to thermal energy for a height of 1 kilometer. Considering all the mixing, it's not surprising no stratification in oceans. I didn't read your link carefully to understand the stratification in glaciers. It's due to a difference in the rates of evaporation for the different isotopes? @Andy that is sounding interesting :) @all: I am fascinated by the possibility of microbes (perhaps in the 4th domain?) that require very high concentration of D. Which is why bummer that there isn't stratification in the oceans, although I suppose high hydrostatic pressure would be much bigger deal anyway. - Steve Koch

@Steve, how would such microbes evolve? First, there'd have to be an environment particularly rich in D for them to succeed in by adapting to higher and higher D usage... if there is such an environment, that's where to look (but unfortunately I can't think of any such, unless somewhere in a nuclear reactor). How are your seed experiments coming along? - Bill Hooker

@Bill -- well couldn't we just create them in the garage genetic screens? But since you mention heavy water reactors, that rung a very faint bell in my head. I feel like someone either talked about microbes in heavy water reactors, or joked about it with me. Certainly it seems like things can grow everywhere, and the parts of heavy water reactors can't be that tricky of an environment. So the archaea and what not that are there may have evolved a preference or need for D versus H. Even though it's only been half a century, that could happen. I heard a talk last year at my DTRA program review about microorganisms that have evolved in less than 100 years to "eat" products and byproducts of explosives manufacturing. I.e., compounds that didn't exist more than about a 100 years ago and there are microorganisms that have invented the chemistry / enzymes necessary to eat TNT or whatever. It was fascinating. - Steve Koch

Andy has a really nifty camera set up for the tobacco seed experiments. I believe he's making videos of the previous attempt, but I remain skeptical that we'll be able to detect a difference with statistical significance. There were certainly some new technical challenges, such as the seeds tending to go through sinking / floating cycles, which made no sense to me. At the very least, the videos should be entertaining, and reproducing Lewis' 1934 results are successful. Andy and I think that measuring "time to germination" may be the most sensitive test for sensitivity to D-depletion. What "we" should do next time: (A) pre-soak the seeds in the solution at 4C for a day or so, (B) put a screen (sieve) or something on top to prevent floating, and (C) orient the seeds with tweezers so the germination thingy can be seen. Hopefully The Google will help transfer this technology to school science fairs next calendar year. I'm hoping my son's kindergarten class has a science fair :) - Steve Koch

Steve, I did a quick scan of the glacier paper, and it seemed that evaporation and surface temperature changes were indeed important. (The density profiles they present for D are not monotonic with respect to depth.) The Lake Vanda paper says deuterium is stratified in the lake, but has a *lower* density in the deeper sections :) I had guessed the mixing in the oceans would overwhelm any tendency to stratify, but (thermal energy)~(1km) sounds even more discouraging than what I had thought. The deeper parts of Black Sea are known not to mix with the upper layers (and to be anoxic), so I thought it may be a better candidate than the ocean, but a relatively recent paper on the D/T/O18 concentrations there doesn't look encouraging, either... Anyway, looking forward to your videos :) - mkz

I like the science fair idea, but aren't the D-depleted/enriched reagents pretty expensive? - Bill Hooker

D-enriched is actually pretty cheap, I'm guessing because of use in nuclear reactors. You can get 10 ml of 99.9% D water for like ten bucks which would be plenty for tobacco seeds. And you wouldn't need it that pure. D-depleted is more expensive, though, I think. Not sure if following link works for D-enriched sigmaaldrich.com/catalog/ProductDetail.do?D7=0&N5=SEARCH_CONCAT_PNO%7CBRAND_KEY&N4=151882%7CALDRICH&N25=0&QS=ON&F=SPEC - Steve Koch