Ideal sodalite contains roughly 24 wt% Na2O in a low-density zeolite-like structure, so presents extreme challenges for probe analysis. With a normal 1-micron fixed beam (15 kV, 20 nA) you can actually see the surface of the mineral bubbling. Our solution was to go with a 5-micron defocussed (not rastered) beam at 2 nA. We used Amelia Albite as our standard (not ideal, and I would now probably use something like jadeite) but it turned out well based on an oddly fortuitous circumstance. We set up a task on the probe to record Na K-alpha count rates at 0.1 sec intervals for 60 seconds to monitor loss of Na from the excitation volume. It turned out that Na counts dropped at almost exactly the same rates over that interval for a 2-micron 20 nA spot on Amelia compared to a 5-micron 2nA spot on end-member sodalite. The resulting analyses using this protocol ended up with excellent analytical totals and great stoichiometry.

For those who might be interested in this topic and who read French, there is a PhD thesis by Francis Autefage (1980) titled "Etude de la migration du sodium et du potassium dans les mineraux au cours de leur analyse a la microsonde electronique" [Study of the migration of sodium and potassium in minerals in the course of their analysis by electron microprobe] - Universite Paul Sabatier de Toulouse. It deals in detail with the process of sodium counts loss and I have found it a useful resource. It was published in part in the Bulletin de Mineralogie, 103, 48-53 (1980) "Variations de la teneur en sodium et en potassium dans des mineraux au cours de leur analyse a la microsonde electronique" (also in French) but I have not found any evidence of publication of this research elsewhere so far.

Dr. Robert Tracy

Professor of Geosciences

Associate Department Head

Director, Museum of Geosciences

Virginia Tech

Blacksburg VA 24061-0420

540-231-5980

540-231-3386 (F)

On Jan 26, 2016, at 6:37 AM, Linda Campbell <[log in to unmask]> wrote:

Dear Howard,

   We have a paper in press (Mineralogical Magazine) that examines microprobe analytical issues, including Na loss, for the 'severe cases' of zeolites. There might be something useful in there that is applicable to omphacite. But for any analytical method, I would recommend trying to find closely matching reference minerals, in line with what Alex was saying about Na-silicate standards.

   A pre-proof full-text of the above paper, Campbell et al (2016) "Determination of zeolite-group mineral compositions by EPMA", is available at
http://www.researchgate.net/profile/Linda_Campbell10/

   I hope this helps a little.

Regards,
Linda Campbell.

Dr. Linda S. Campbell,
Academic Visitor,
Sch. Earth, Atmospheric and Environmental Sciences, The University of Manchester, U.K.

----------------------------------------------------------------------

Date:    Mon, 25 Jan 2016 19:41:51 +0200
From:    "Proyer, Alexander (Prof)" <[log in to unmask]>
Subject: Re: Na loss in pyroxene

Dear Howard,

as nobody has come up yet with a published systematic study of the matter, let me contribute some of my experience at this point.

Analysing feldspars and omphacites routinely for years on a SEM equipped with an EDX system at 15kV, 5nA beam current (EMPAs usually use higher beam current), we were of course concerned about potential Na loss
and made a series of tests on our albite and jadeite standards: between 5 and 10 measurements subsequently on the same spot on the mineral, to see if later analyses show a trend in the Na/Si ratio that might indicate Na-loss.
Measurement times were 20 sec on peak, 10 sec on each background.
The jadeit worked perfect under these conditions, and even the albite which was considered a sure candidate for Na-loss did not show significant changes (given the EDX precision) after 5 measurements.
The setup was not perfect, e.g. I don't think we had freshly polished surfaces of the standards, etc., but I stopped being worried about Na-loss at that point.

As mentioned by others, non-stoichiometry in high-temperture/high-pressure pyroxenes due to significant amounts of Ca-Escola endmember might be an issue.
I have summarized some of the references around this in a Contrib. Mineral. Petrol. paper in 2004 (Pitfalls in geothermobarometry of eclogites... - CMP 147:305-318) - that study was done, hoever, with a high precision microprobe at 15 kV, 20nA and used micro-Moessbauer analysis to determine ferrous and ferric iron.

What never worked was trying to analyse Na in pyroxenes with an albite standard or Na in a feldspar with the jedeite standard.

Hope that adds something you did not know yet.
Alex

Alexander Proyer
Professor of Petrology
Department of Geology
University of Botswana
________________________________________
From: Metamorphic Studies Group [[log in to unmask]] On Behalf Of day [[log in to unmask]]
Sent: Sunday, January 24, 2016 5:07 AM
To: [log in to unmask]
Subject: [geo-metamorphism] Na loss in pyroxene

Colleagues

Does anyone know of accessible studies of sodium loss (or not) from omphacitic pyroxene during microprobe analysis?

Thanks

Howard

Howard W. Day, Professor Emeritus
The Dept. Formerly Known as Geology
University of California Davis
One Shields Ave.
Davis CA 95616