Monday, January 22, 2007

(U-Th)/He Thermochronology

UPDATE : If you found this web page while looking for information on labs where you could do (U-Th)/He analyses, please check out the web page for the new He lab at the University of Colorado - Boulder.

In response to a comment I received on an earlier post, I thought it might be a good idea to give a brief overview of (U-Th)/He thermochronology. Chances are I will have pleanty of posts in the future that will touch on the subject, so this could be useful.

Like all radiometric geochronometers, (U-Th)/He chronology is based on the radioactive decay of a naturally occurring parent nuclide (in this case, U and/or Th) to a stable daughter product (in this case He). We typically think of U and Th decaying to Pb, which it does, and this is the basis for another geochronometer. U and Th each decay by what is called "chain decay", meaning that there are many steps and many intermediate nuclides in the decay process, and decay from one of these to the next is accomplished via wither alpha or beta decay. 238U for instance, has 8 alpha decay steps in its decay chain. During alpha decay, an alpha particle (2 neutrons and 2 protons) is emitted from the nucleus. Alpha particles are He nucleii, as soon as they leave the nucleus they pick up some electrons and shabang, a He atom. 235U undergoes 7 alpha decays, and 232Th produces 6. Below is a schematic diagram of the decay chain for 238U, the more abundant U isotope. You can see that there are multiple paths the decaying atom can travel, and that it must experience some number of beta decays, along with exactly 8 alpha decays. Beta decays are when a neutron is converted into a proton and an electron, and that electron expelled from the nucleus (simplified, but you get the point), so during a beta decay, the number of neutrons in the nucleus goes down by 1, and the number of protons increase by 1 (the number of protons is what defines something as a single element, changing the number of neutrons just makes something an isotope of the same element).
235U and 232Th have similar decay chains. So, if you measure the amount of the daughter product (He), and the amount of the parents (238U, 235U, and 232Th), and know the rate at which decay occurs (which we know exceedingly well), then you can calculate the amount of time that the He has been accumulating.

But helium, really, helium in a rock? It seems like a stretch, especially considering our everyday experience with He. It makes up only about 1 ppm (part per million, or .0001%) of the air we breath, it is so light that it can escape our atmosphere. It escapes from balloons, canisters, everything. The idea then was that it must escape from rocks and minerals right away. This was what Strutt and most of the early workers I discussed in an earlier post believed. When they dissolved the minerals in acid of course, the He was released immediately. It turns out that certain minerals actually can retain He relatively well. Apatite and zircon, two common accessory minerals that typically contain 10's to 1000's of ppm of U and Th, can hold onto He for geologically meaningful lengths of time (millions of years), as long as the temperatures don't get too high. Below is an image of some apatite crystals from a granite in western Utah that I worked on, notice the scale bar. It is fairly easy to "date" an individual crystal, typically we measure 4 or 5 individual grains to make sure the ages are reproducible.

He diffuses rapidly out of apatite at temperatures greater than about 70°C (again, talking geologic time scales). Radioactive decay of course doesn't care about temperature (well, not anything close to a temperature found on or in earth), so it is occurring all of the time. At higher temperatures, the He that is being produced is diffusing out of the crystal and leaving the system, into the atmosphere. If you are not retaining any of your daughter atoms, you have a zero age. Once the rock cools, you start to retain the He, so the age you measure with (U-th)/He is not necessarily a formation age, but rather a cooling age, measuring the time since the rock cooled to the temperature at which it could retain He. There area actually many "thermochronometers" like this, systems that geologists use to determine the time-temperature history of a rock or geologic terrain.

Measuring the different nuclides is a little tricky. As far as I know, there is no one machine that can accurately and preceisely measure U, Th, and He. U and Th are often measured together (like using this amazing machine), but He is a different matter.

For a variety of reasons, we tend to measure He from single grains, I work primarily on apatite (Ca-phosphate, the same mineral that makes up our bones and teeth, it is also a common accessory mineral in granites and related rocks). The first step is to get the He that is trapped in the crystal out, so we can measure it. A common way to do this is to heat the crystal with a laser

These are pictures I took of the laser I used in grad school, an Ar-ion laser. Many different kinds of lasers can be used, the Ar-ion has the advantage that the light is in the visible spectrum, so you can take really amazing pictures of it. The He that is generated during the radioactive decay an that we de-gas from the crystal with a laser is 4He. To measure the total amount of 4He we spike it with a known amount of 3He, and measure the isotopic ratios on a Quadrupole Mass Spectrometer. There are a few intermediate steps, mainly designed to clean up the gas and remove any contaminants, that is where the cryogenic charcoal trap I showed in a previous post comes into play.

The lab that measures He cannot measure U and/or Th. So we then take the crystals, dissolve them in nitric acid (for apatite, nastier stuff for zircon), spike it with known amounts of 233U and 229Th (or some other appropriate isotope),

and measure the U and Th on a magnetic sector ICP-MS.

Once we know the amount of U, Th, and He, we can then calculate an age. All of the steps up until now are pretty straightforward, now of course comes the hard part, interpreting that age. Sometimes that is fairly obvious, sometimes it isn't.

There are many other aspects of this technique that I find interesting, and that are very important for anyone wishing to use it. For a technical discussion I'd direct you all to an excellent review paper written in 2002:

Farley, Kenneth A. (2002) (U-Th)/He dating; techniques, calibrations, and applications. Reviews in Mineralogy and Geochemistry, v. 47, pp. 819-843.

So, I hope this clears up a bit about the technique. I am sure posts in the future will explore the topic a little more, for example the issue of alpha-ejection, He diffusion, etc. If you are interested in geochronology, one additional place I'd recomment you check out is the Earthtime project. Cheers.


BrianR said...

This is good...I haven't read your summary in a lot of detail yet, but when I do i'll be able to communicate better w/ those in this field.


C W Magee said...

You forgot the best part.

Sometimes, following He measurements, certain little dipshits dissolve not only the degassed grain, but also the entire platinum capsule.

When this happens, they then end up measuring not only 235U, 238U and 232Th, but also 40Ar195Pt, 40Ar198Pt, and 40Ar192Pt when their sample is ionized by the IC plasma. Not only does this fuck up their analysis, but it means the poor techo then has to replace the ENTIRE SAMPLE DELIVERY SYSTEM when the PGE people want to do ppt level platimun measurements.

On the bright side, at least I'm not bitter...

Thermochronic said...

Yes, fortunately when working with apatite this hasn't been an issue for us. We monitor this (we don't own the ICP-MS so I'd hate to screw it up), and the levels have never been above background. Nitric acid (at the times and temps we dissolve in) just doesn't seem to do mu to the Pt. Of course this is a different story when dissolving zircon, and you must do a nasty HF acid vapor dissolution. This would easily dissolve the Pt foil. For zircon, most people now use Nb foil, you still put a boatload of material into the machine (which takes a long time to deal with), but you don't have the interfering PtAr peaks.

C W Magee said...

I thought Platinum was fairly tolerant of HF.

BTW, have you ever looked at He-rich samples that has suffered U loss?

Uncommon, but fun.

Thermochronic said...

By itself Pt does OK, but the acid-vapor (boiling pressurized HF) seems to do the trick. Truth is I haven't worked with it myself, I stuck with apatite for my PhD, primarily because of the wet chemistry side.

What kind of He rich materials do you mean? For most things the U loss would only be detectable if it happened at low temps. I'm intrigued!

kshatriya said...

Thanks alot, this was really a useful discussion for me. I was just searching for some basic concept of Thermochronology and got it right here.
Thanks again

Alexandria Darcy said...

This was extremely helpful! I'm in a tectonic geomorphology grad student class, and haven't had much exposure to thermochron until now. I needed to know a little more about the basics to read an assigned paper this week, and your blog post did the trick! Thank you!

Thermochronic said...

Much appreciated! I am glad it is useful, He thermochronology is being used more and more in a variety of fields.

FYI, I have restarted blogging at, nothing new thermochronological yet, but soon!