Saturday, October 6, 2012

What am I doing here? Isotope Geochemistry.

As most of you know, I'm currently in Lyon, working in Dr. Janne Blichert-Toft's isotope geochemistry lab.  I spent the past several months at Rice separating minerals (garnet and clinopyroxene) from 3 pyroxenite xenoliths from the Sierras.  These two minerals, combined with their respective whole-rocks, will be used to get Lu/Hf and Sm/Nd internal isochrons for the 3 pyroxenites.  The slopes of these isochrons will give the ages at which the pyroxenite samples record their final equilibration P-T conditions.  Garnet is an excellent mineral to use for the Lu/Hf method (and likewise clinopyroxene for Sm/Nd) because of this mineral's strong preference for the parent isotope (Lu for garnet).  Another advantage of Lu/Hf is the higher closure temperatures compared to Sm/Nd.   So, we have very good candidate samples to do this with, plus the results could potentially be very interesting geologically.  What age will these pyroxenites give?  Will it be an age coeval with formation of the Sierran batholith, or an age reflecting a tectonic event much later, such as the Laramide orogeny?  Either way, whatever we get will be interesting.  I just hope that we don't get the age of the volcanic eruption that brought up these xenoliths - but the high closure temperatures of Lu-Hf should prevent that.

Below is a sketch of an isochron.  For better explanation, see intro isotope geology textbooks, like Dickin or Faure.  This is just to illustrate some basic principles.  The dashed line represents "time zero", meaning when the whole rock and its constituent minerals (garnet, clinopyroxene) became homogenized and closed to diffusive exchange.  At this time, they will all have the same 176Hf/177Hf ratio, but different 176Lu/177Hf ratios due to the different affinities for Lu of the different minerals (garnet has a high Lu/Hf ratio because garnet prefers Lu over Hf).  Over time, these different initial Lu/Hf ratios will decay, rotating the dashed line to the solid black line, the isochron.  It is the slope of the solid black line that yields the age (for the math behind this, you can go to any intro isotope geology textbook or even Wikipedia).  The black points are what we measure in the lab, after going through complex chemistry procedures that separate the individual elements of interest. 

But, I'm nowhere yet close to getting the isochrons yet.  Before you can even get there, a lot of chemistry must be done, which is what I am doing now. Below is a flow-chart of the whole process (see Blichert-Toft, 2001, Geostandards Newsletter and Blichert-Toft et al., 1997, CMP for full details):

I'm only barely done with the very first step, sample attack and dissolution.  Of course, this is the most important step because it dictates everything that will go on next.  We first weigh the samples and add mixed spikes to them (a Sm-Nd spike and a Lu-Hf spike).  Usually, you would add a separate spike for each element, but a mixed-spike makes things easier because you could make a mistake weighing your separate spikes, thus irrevocably screwing up your parent/daughter ratios.  What is a spike and why do we need to add it?  First, the spike is a solution of an artificially enriched isotope of the element(s) of interest, which you know exactly the concentration of (this requires carefully making and calibrating of the spike).  Why do we need it?  When you make isotopic measurements in which the parent nuclide is of interest, as in my case (or any case when you want to get an age from an isochron), you need to measure the parent/daughter ratio very precisely.  However, when you go and measure your samples on the mass spectrometer, you usually do not know exactly how much of your total sample actually makes it all the way through the complex machinery of the instrument and to the detector.  Several things can occur, such as unequal transmission of the parent and daughter isotopes because they have different ionization potentials.  So, because you added a spike of which you know the exact amount of and composition of, when you go and measure your spiked samples on the mass spectrometer, you know that the ratio of spike sampled to spike detected will give you the detection efficiency.  The ratio of spike to natural isotope gives the isotope's concentration. 

Before any chemical separation of the elements can proceed, the sample and spike must both be completely homogenized.  This can take weeks especially for refractory minerals like garnet or zircon.  To ensure complete homogenization, we use high-pressure, steel-jacketed Teflon bombs.  HF (the only acid readily capable of dissolving silicates) is added to the bombs, and then they are heated at 156 C for a week.  A spring inside the bomb ensures the samples are kept under high pressure.  Then after a week they are taken out, and we add perchloric acid (HClO4) to them. HClO4 is added to help expel any remaining HF (HClO4 has a higher boiling point than HF) from the digestion, thus decreasing the probability of insoluble fluorides from forming.  At this point, the sample should be completely attacked.  If it is, we can add 6N HCl to the bombs, close them up, and put them back in the oven for another several days.  After this last step, the sample and spike should be fully homogenized and the separation chemistry can begin.

Below is a sketch of how the bombs work:

Next time, I'll explain a bit about the column chemistry and element separation.  Toodles. 


No comments: