Tuesday, September 25, 2007

The earth, the atmosphere, and 40Ar retention

When I started getting into noble gas thermochronology (i.e. 40Ar/39Ar and (U-Th)/He) I realized that there were different types of geoscience literature involving noble gas research. First, there are the studies I was most interested in, involving either the behavior of radiogenic noble gases in common crustal minerals or their application to understanding tectonic problems. The other noble gas studies, which I tended to ignore, used them as geochemical tracers of a whole boatload of earth processes, including whole-earth degassing and formation of the atmosphere. In the past year through collaboration between my research group and some excellent geochemists, and in light of a recent and excellent Nature article announcing some really surprising findings, I have gained a new appreciation for the role of noble gases in geoscience.

The article, which came out in the September 20th issue of Nature, is called 40Ar retention in the terrestrial planets. It presents the results of a whole series of experiments examining the behavior of 40Ar in forsterite ( Mg2SiO4) and enstatite (MgSiO3), the two minerals that make up most of the mantle. The common perception of noble gases is that they are relatively incompatible in minerals. Not only do they diffuse out quickly, but during partial melting events, the noble gases are strongly partitioned into the melt. The melt then ascends (say at a mid-ocean ridge), the gas exsolves, and escapes into the atmosphere. For 40Ar, we should be able to calculate the percent of the planet that has degassed if we know the K content of the planet (40K being the radioactive parent of 40Ar, and therefore the source of most of the Ar) and the total amount of 40Ar in the atmosphere. Some studies have concluded that the earth may be only ~50% degassed. This is confusing if you accept that noble gas diffusion is relatively fast and noble gases are strongly partitioned into melt during partial melting events; why hasn't the whole mantle degassed by now?

The data presented in this paper are discussed in terms both diffusivity (the speed at which 40Ar atoms move through the crystal lattice) and solubility (the total amount of 40Ar that could be stuffed into a crystal lattice given unlimited time, temperature, and 40Ar). The authors took highly polished slabs of minerals (both natural and synthetic crystals) and put them in 40Ar rich atmospheres at different pressures and temperatures. After set amounts of time, they removed the samples, and looked at concentration profiles of 40Ar in the crystal using Rutherford Backscattering Imaging. From this they are able to construct concentration versus depth plots for all of the various temperature and 40Ar pressure scenarios, which are then fit these profiles with equations relating to diffusive uptake, which then allow for the calculation of some of the fundamental parameters of diffusion (diffusivity and solubility). These measurements are different from the bulk loss profiles I am used to, where we infer the concentration profile based on step heating experiments. These instead are direct measurements of the distribution of 40Ar in the solid. The paper discusses many of the potential problems of the experiments and measurements, but I won't go into that here. Their punchline is that 40Ar solubility is actually fairly high in both forsterite and enstatite, and that 40Ar diffusivity is actually fairly low. In fact, during partial melting events, 40Ar can almost be thought of as a compatible element, that is, it is not strongly partitioned into the melt at all; both forsterite and enstatite can hold onto significant amounts of their 40Ar during partial melting! As I mentioned earlier, this is in contrast to previous thoughts and experiments on the topic, but does at least fit with the suggestion that the earth is not fully degassed.

Like most Nature papers it is not terribly long (they have a maximum of 4 pages to work with), but well worth the read. The implications could be tremendous. Even from a thermochronology perspective, it makes me wonder about the validity of our diffusion experiments that try to infer the concentration profiles of gases in minerals indirectly. Hmmmm.

Because it just got printed I decided not to include any of the figures in this post, but anyone interested should check out the original paper (Nature has the advantage that even most public libraries carry it). Allegedly there is a much more detailed version in the works, I'll keep you posted.


Brian R said...

I was hoping you'd post about that paper. Can you think of a few hypothetical and significant implications? If they are right, what does it mean?

Thermochronic said...

Well, from a whole earth perspective. If they are right, then there must be another mechanism for degassing the planet besides partial melting. One they propose is the potential importance of hydrothermal alteration at the ridges as a means to extract the Ar. What I thought of then was how important liquid water at the surface is for degassing the planet, and the potential consequences for other planets with similar bulk chemistries but very different atmospheres.

From a thermochronologist's perspective, I think the idea that Ar is actually fairly soluble in some minerals could go a long way to explaining some of the very weird data people often get from metamorphic rocks. We always test for the presence of "excess argon" in our samples, and have to discard data which has obviously been contaminated by non radiogenic gas. But the mechanics and chemistry of how that gas is incorporated into the minerals is unknown, and therefore there is nothing we could even do with it. I think this could go a long way to explain the behavior of minerals in the crust. Although it should be noted that their experiments were limited to phases most thermochronologists don't care about (olivine and pyroxene).

There is a comment to the paper that appears in the same Nature issue that raises some interesting points and discusses some of the potential ramifications.

Personally I've never thought of the solubility of noble gases in minerals before, that has been interesting from my perspective. They were working with much higher pressures of gas than the average mineral in nature, but the interplay of solubility with diffusivity has made me rethink noble gas diffusion in general.

Kim said...

But the mechanics and chemistry of how that gas is incorporated into the minerals is unknown, and therefore there is nothing we could even do with it.

Except bang our heads on desks, or occasionally (in my case) on a robotic arm used for analysis of paleomag samples.

I've had some pretty miserable experiences with excess argon in muscovite/phengite (in the earliest days of the Stanford argon lab, before Green Earth Sciences Building was built). I hypothesized that the micas picked up excess argon while they were being deformed and rocks at deeper levels were undergoing granulite-facies metamorphism... but honestly, I never really knew. (And after my experiences in the lab, I didn't want to know any more; I ran fast and far away from both the field area and the technique.)

But it would be very interesting to be able to use evidence for excess argon to say something else about the environment - the partial pressure of Ar in fluids, for instance.

(I think at least one of your collaborators knows about my old rocks... if they're interested in messed-up micas, I can tell them where to find them. Though the Alps or the Aegean would be more interesting to visit, and just as messed up.)

Thermochronic said...

Unfortunately with the move to Green Earth Sciences the "excess" problem did not get resolved! I have a similar experience working on micas (multiple samples) from the amphibolite rocks of the Funeral Mountains, which is one of the reasons I am so interested in this. The deformation enhanced Ar retention was something he didn't touch on, although if solubility is high, then the amount of Ar that a crystal could pick up from the environment would be in some part a function of it's surface area....interesting...

One other thing I discussed with the authors was how they found almost no evidence for the adsorption of Ar to the outside of crystals. I had always assumed that the first gas you get from a sample, at low temperatures, which is almost always atmospheric, was gas that had been adsorbed to the outside of the crystal.

Again, their results were only for forsterite and enstatite, but it is interesting.

Chuck said...

Can you email me a pdf? I can't access Nature.

But one of the experiments I wanted to do during our field off-season as an "amateur" was to determine Ar partition coefficients directly (If it was easy it would have been done decades ago, but I think it is tractable), so I'm pretty keen to see this paper for scoopitization reasons.

On a theoretical level, though, isn't the solubility of Ar in Olivine and Opx irrelevant? After all, K in the mantle partitions into Cpx. And the host of Ar generated in the lower mantle passing up through the trnasition zone should depend on the relative partitioning between Ol, Cpx, and majorite (which in turn transforms into Opx, I think).

And on the gossip side of things, is it just me, or has Bruce been on a bit of a roll the last couple years wrt high-impact papers?

Thermochronic said...

Emailing pdf's is a highly unlawful and horrible thing to do, violating copyrights and taking money away form needy organizations like Nature, Thermochronic does not participate in illegal activities.

As far as the opx and cpx issue, one fo the reasons it was never measured in cpx is that the method (rutherford backscattering) does not work with phases that have Ca or K in them (it interferes with the Ar peaks too much). I think the assumption is that the pyroxenes act similar. And, truth be told the deviation from the expected value that they measure is enormous. I don't know enough about mantle mineralogy to say much more. If perhaps some less legal conscious person sends you the pdf I'd love your comments. There is also a Chemical Geology version in the works that should be out soon.

Bruce et al seem to have been on a role for a long time. I actually saw him give another talk the other day that blew my mind, Ti-in-zircon thermometry and it's relatives, Ti-in-quartz, Zr-in-rutile, and Zr-in-sphene. Amazing stuff I hope to post on soon. When I first met his I honestly expected him to be really old, just because he has done so much great work. Turns out he is just super smart and hard working.