RedOx conditions of ultramafic and gabbroid rocks formation in Yoko-Dovyren intrusive complex

Few days ago I finally received the final preprint of a co-authored paper from master studies times ("RedOx conditions of ultramafic and gabbroid rocks formation in Yoko-Dovyren intrusive complex (based on intrinsic oxygen fugacity measurements in olivines)"), Paper preparation took several years, but finally I'm happy that it got published.

This post described in brief the content of this paper. Since anybody can go inside the paper, this is more informal and backdoor intro. I won't give citations to materials mentioned in this post, however, everybody interested is free to ask :) 

This text in Russian.

The object of studies: mt. Dovyren

Geological problem. Magma crystallization results in fractionation of silicate minerals. Chemical compounds not included in their composition accumulates in residual melt. The same as if we take a heap of coins and pick out cents, number of euro coins increases. But there is no way to accumulate and store those components eternally, as well as keep the conditions in magmatic chamber unchanged, so that once these components will precipitate and separate from the melt. So that finally a metal initially dispersed a seldom becomes concentrated enough to form an ore deposit. One ore deposit won't satisfy needs in a specific component for all the humankind, so there should be a way to look for more deposits.

How to solve the issue. There not so many easy ways to find a vein of minerals in boreal forests. The easiest way is heavy concentrate washing. If you found some interesting grains in sands, their should be a source nearby, so you can track the river channel to find some suspicious rocks. Then one can make some chemical analyses to find the ore and probably be lucky.

The problem that it is not always obvious where you should start make washings or proceed chemical analyses. Science is a way to say how similar are specific objects, so is it possible to find some deposits of economical interest or not. If one "suddenly" finds some thermodynamic conditions worked out in one situation, they should also happen in a similar one.

Geological object. Our research is based on Yoko-Dovyren intrusive complex located to the North from Baikal lake. This is a block of ultramafic rocks (average SiO2 < 45%) formed ~700 millions of years from magma of similar composition. It is in general similar to other ultramafic intrusive bodies and contains some copper and PGE ores.

Goal of survey. One of the key parameters which is hard to determine precisely is oxygen fugacity (fO2, and log fO2 is used usually). Fugacity is a kind of effective pressure, and in that sense it works in the same way as activity of components in solutions. 

Crystallization of sulphide ore minerals mentioned above starts when oxygen fugacity (and other parameters) reach certain values. If we can use magmatic minerals to find out these values, we will be able to predict where (within the magmatic body) ores should accumulate.

Oxygen fugacity is not just a continuous parameter like pressure or temperature. It changes to alter the oxidation state of elements (iron predominantly) in magmatic melt. Those chemical reactions involving specific minerals are buffer equilibriums. The most known are IW (iron-wustite, Fe-FeO) - and all the Earth's magmas are above it besides meteorites can be below it, WM (wustite-magnetite FeO-Fe3O4), which is high enough (conditions are more oxidized) to form Fe3+, and QFM (quartz-fayalite-magnetite), when oxygen fugactiy reach values high enough to disassemble even silicate minerals. These equilibriums buffer the system, because there is a lot of iron (comparable with the amount of free oxygen) to keep the amount of free oxygen by Fe2+-Fe3+ transitions.

What we've done

My ex scientific advisor in GEOKhI RAS wanted to test a device made in another "friendly" lab. 

The device is a locally assembled tool according to schemes and ideas used in several labs in the end of 1970-s and 1980-s. Method is based on comparison of oxygen electrochemical potential measured in mineral grains with air. 

I personally see a lot of issues with these measurements, and I do not know systematic investigation of their influence. Method doesn't have standartization; there are no systematic surveys with experimentally cooked rocks (under controlled conditions) to prove the relationship between measured parameters and real world, and there is no theoretically proved background to apply this method to certain minerals to estimate the conditions. Olivine which we used in our survey is famous for high diffusivity, so that severals tens of days is enough to reequilibrate it under 1100-1200C (this is below melting point). Oxygen can be reequilibrated within some days as well.

The first sample I brought for the investigation was returned in a couple of days with remark "your olivine is ultra-oxidized". I came with my camp gas burner to show that it can happen only if there was no hermetic conditions. The leak was repaired, but it was a good point.

I think that my general feeling is clear. However, it's always interesting to try. Also, if your boss says the hamster is a bird, then this is a cute little birdie.

Sample preparation

Lab to crash the rocks. One can use mechanical crashers on the left side, but normal people do not use them since they spoil all the sample with previous uncleaned powders.

Rocks to crash

Manual crasher, hammer, air compressor

working place

Set of sieves

Magnetic separator. If there was few olivine, one can use it to enrich the crashed material. Magnetic grains settle slower.

It results in white non-magnetic grains settled first and magnetic grains settled next.

The only normal binocular in our campus. Razor is used to make gaks of powders, and then one use needle to split it into two gaks with grains of interest and the rest. Notice a sweet old school PC with 3.5" and 5.25" floppies.

Grains for experiments. To make a control measurement, we've selected randomly more than 80 grains to check whether it is an olivine. 100% grains were the proper mineral.

Results

Oxygen fugacity can be measured under different temperatures, and result will differ in orders of magnitude. Experiments were carried step-by-step under a range of temperatures, and values were reported each several tens of degrees. So we obtain a set of points. If direct and backward trends overlap, experiment is fine.

Points make up line in T - lg fO2 space. We got a nice pack of lines intercepting at 700C and pretty big divergence in high-T (1200C) area. My  boss is interested to estimate the conditions during the end of crystallization so his interpretation is focused on high-T area. My proposal (it is also discussed in paper) is more paranoid. I resemble that ~700C is pretty about the lower boundary of significant diffusion in olivine expect that this interception point if the only very correct one. These grains carry the information about the last equilibrium, but this is deeply in subsolidus.

What did other people find

There was only one another group of people from RAS made an investigation of oxygen fugacity. They made no experiments, but used an equation to calculate log fO2 from mineral chemical compositions with know temperature (which is also calculated using chemical compositions). This approach is based on big experimental datasets, so it is a classical one and used worldwide nowadays.

They used this method in a strange way. Direct application resulted in the same low T values and low lg fO2 estimates. Thus they proposed some interesting corrections, and one of them presume magmatic chromite to have no Ti in it's composition. Basically they countdown all the titanium to obtain high temperatures and ultra high fO2. This idea resulted in my Musculus corrugator supercilii and handface muscles cramp.

Some conclusive remarks

We used a method between qualitative and quantitative. It agrees with all the general suggestions and ideas that mineral compositions in long-living magmatic systems correspond not to mineral formation, but later subsolidus reequilibration. Numerical values fit those expectations.

It is plausible that some of our data might reflect original magmatic conditions. All lines are below QFM which is something fine for ultramafic chambers. These data is much better than high values (QFM+1 to NNO) shown in aforementioned alternative data.

It might seems that two paragraphs written above contradict each other. Equilibrium thermodynamics says that system can't remember the way passed before the equilibrium. The delicate thing is that nature might be far from equilibrium. The key parameter to reach the equilibrium even under high temperatures is the presence of water (even trace amounts are fine). If the rock is deformed (so that it contains a lot of cracks to transport water and fluid) or it has a lot of water-rich minerals, then it might reequilibrate. But if conditions were dry, it can stay unchanged. Huge layered intrusions have a diversity of conditions with different degree of deformation and different proximity from water-rich surrounding rocks. In rocks we've studied even far sufficiently (100-150 m) from bedrock one can find phlogopite and other water-bearing micas.