So! I mentioned isotopes above. They're the tool I'm going to use.
To take several steps back... atoms are made up of three types of particle: protons and neutrons (in the nucleus) and electrons (whizzing around outside, ish). What defines whether an atom is hydrogen or helium or oxygen or carbon is how many protons it has - something with 6 protons is always carbon. However, the number of neutrons can also vary - so you can get carbon with 6 neutrons (carbon-12, the vast majority), carbon with 7 neutrons (carbon-13, which is also stable but present in much smaller quantities) and carbon with 8 neutrons (carbon-14, the radioactive carbon isotope used in radiocarbon dating).
If you look at all the carbon on the planet, there's an overall ratio of carbon-12 to carbon-13 (I'm ignoring the 14 for these purposes but the same argument holds with a few tweaks to take into account its radioactivity). However, when you look at individual reservoirs of carbon - for example, a tree or coal or limestone - you tend to find that the ratio of carbon-12 to carbon-13 is all over the place from your expected value, and not consistent between reservoirs (so even if you've got your best guess at what the ratio should be right, you're still seeing significant variations from that ratio).
So: something has (or somethings have!) to be happening to shift the number, and as it turns out they are, and this is called isotope fractionation. Specifically, the examples I'm giving - and the species (i.e. elements/oxides/isotopes) I'll be working with - are all stable isotope fractionation (i.e. no radioactive decay is involved). If you'd like to know about how this all works with radioactive species playing along then by all means ask, but I'm not going to include it in this comment for the sake of keeping it at least a bit short ;)
So! The particular element I'm going to be looking at in the first instance is thallium, which is a heavy metal element (not hugely abundant!). There's barely any thallium in the mantle, and we know that its isotopic ratio is significantly affected during marine sediment formation - different kinds of marine sediment have different thallium ratios. However, the isotope ratio shouldn't be affected by processes like melting.
When sediment is subducted it is expected to melt pretty efficiently (that is, most of it will melt). This means that you're quite right, we don't actually see bits of mud and shells ;) BUT via the magic of going "huh, this lava has a lot of thallium in it... in a very precise isotopic ratio", we're able to say "the fact that this lava with this composition was produced means that we must have had a source material that meled to create it, and that source material must have been at least X% of this particular type of sediment."
Obviously using just one set of elemental isotopes leaves a lot of room for interpretation, and that's part of the problem with existing work using radiogenics (radioactive isotope systems) - it's great as far as it goes, but there are some cases it simply can't distinguish between. However, the ability to detect small differences in isotope ratios of heavy elements at low concentration is relatively recent, so the hope is that by using these new techniques and newly-available stable systems, we'll be able to build up a better picture of what must have happened.
... which possibly got massively technical at the end? I'm sorry, I can't tell, I'm underslept (but enjoying myself immensely! This is good practice). If you'd like more on any of those bits, please please ask (because chances are someone else would too :D).
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
regna
(no subject)
Date: 2013-11-06 10:42 pm (UTC)So! I mentioned isotopes above. They're the tool I'm going to use.
To take several steps back... atoms are made up of three types of particle: protons and neutrons (in the nucleus) and electrons (whizzing around outside, ish). What defines whether an atom is hydrogen or helium or oxygen or carbon is how many protons it has - something with 6 protons is always carbon. However, the number of neutrons can also vary - so you can get carbon with 6 neutrons (carbon-12, the vast majority), carbon with 7 neutrons (carbon-13, which is also stable but present in much smaller quantities) and carbon with 8 neutrons (carbon-14, the radioactive carbon isotope used in radiocarbon dating).
If you look at all the carbon on the planet, there's an overall ratio of carbon-12 to carbon-13 (I'm ignoring the 14 for these purposes but the same argument holds with a few tweaks to take into account its radioactivity). However, when you look at individual reservoirs of carbon - for example, a tree or coal or limestone - you tend to find that the ratio of carbon-12 to carbon-13 is all over the place from your expected value, and not consistent between reservoirs (so even if you've got your best guess at what the ratio should be right, you're still seeing significant variations from that ratio).
So: something has (or somethings have!) to be happening to shift the number, and as it turns out they are, and this is called isotope fractionation. Specifically, the examples I'm giving - and the species (i.e. elements/oxides/isotopes) I'll be working with - are all stable isotope fractionation (i.e. no radioactive decay is involved). If you'd like to know about how this all works with radioactive species playing along then by all means ask, but I'm not going to include it in this comment for the sake of keeping it at least a bit short ;)
So! The particular element I'm going to be looking at in the first instance is thallium, which is a heavy metal element (not hugely abundant!). There's barely any thallium in the mantle, and we know that its isotopic ratio is significantly affected during marine sediment formation - different kinds of marine sediment have different thallium ratios. However, the isotope ratio shouldn't be affected by processes like melting.
When sediment is subducted it is expected to melt pretty efficiently (that is, most of it will melt). This means that you're quite right, we don't actually see bits of mud and shells ;) BUT via the magic of going "huh, this lava has a lot of thallium in it... in a very precise isotopic ratio", we're able to say "the fact that this lava with this composition was produced means that we must have had a source material that meled to create it, and that source material must have been at least X% of this particular type of sediment."
Obviously using just one set of elemental isotopes leaves a lot of room for interpretation, and that's part of the problem with existing work using radiogenics (radioactive isotope systems) - it's great as far as it goes, but there are some cases it simply can't distinguish between. However, the ability to detect small differences in isotope ratios of heavy elements at low concentration is relatively recent, so the hope is that by using these new techniques and newly-available stable systems, we'll be able to build up a better picture of what must have happened.
... which possibly got massively technical at the end? I'm sorry, I can't tell, I'm underslept (but enjoying myself immensely! This is good practice). If you'd like more on any of those bits, please please ask (because chances are someone else would too :D).