Standards for EPMA-Alkalis

[crystalgrower]

Here is what I know about making “insoluble” alkali metal salts to be used as  standards for microanalysis  (SfM hereafter).

First of all these are not perfectly insoluble in water.  These salts are all truly  insoluble in 100% lab (not pharmacy) isopropanol from a glass bottle.  Not too toxic, compatible with epoxy and plastics, and was found to be safe for metal mounts. 

Second critical point is the matter of mounting perfect crystals.  Most crystal growing from solution yields excellent octahedral or cubic crystals and then some residue of hoppered garbage.  The hoppers are worse than powder because they can retain microlayers of saturated solution that outgas for far too long.   

Buyer Beware:  Inspect loose grains that you buy as “SfM” and return any with hoppers for a full refund.  Inspect any natural SfM for inclusions by immersing in a few drops of isopropanol, and rotate  to see any second phase under magnification.

Lithium Commercial LiF and natural LiAlSi2O6 spodumene.  Spodumene has a good cubic cleavage and should be transparent and free of any Fe/Mn discoloration. It can turn light green during XRF analysis.

Sodium  If you want an O-free material there is NaMgF3 prepared by mixing stoichiometric amounts of NaF and MgF2 and melting in an alumina crucible.  Details most recently A. Chakhmouradian et al, Physics and Chemistry of Minerals (2001) volume 28 pp 277-293 which also details the KMgF3 analogue.

Na migration is known to occur in some minerals.  I consider that if any ion migration can occur then the material is not SfM.  Aegirine and true jadeite are known to be stable under the electron beam. 

Potassium It was simple to prepare K2SiF6 from concentrated HF solution. 
WARNING YOU MUST BE EXPERIENCED IN WORKING WITH HF AND USE AN APPROVED FUME HOOD AT ALL TIMES.
I added a weighed amount (less than 5g) of ultrapure SiO2 to 300mL of fresh conc HF in a 500mL bottle with stir bar.  I put this in a boiled-water bath (superheated in microwave, not on hotplate) and ran the mag stirplate until the water bath cooled down to room temp.  I added the stoichiometric amount of K2CO3 and stirred that without any additional heating. 
I poured a layer of this solution no more than 1cm deep into a PTFE evaporating dish and let that  dry overnight on lowest hotplate setting.  NO you do not want this in an HF still because you lose about half of the Si as SiF4.  Residual KF washes off, leaving good clear crystals with one perfect octahedral face and clean edges. 
I redid this synthesis using 4K:Si and the yield is about half good crystals and half hoppers.  Then wasted time picking clean.  Better to start with the 2:1 K:Si and wash off excess K.

Rubidium Major overlaps in both EDS and WDS for Rb2SiF6 but the batch I made should still be good for SIMS. Same method as K.

Rb2SnF6 could be made using SnF4 in place of SiO2.  No overlaps. Use 10% excess of Rb (added as Rb2CO3) to Sn to ensure that hoppers wash off.

I made Rb2PtCl6 from  both Pt-aqua regia and Alfa H2PtCl6, adding RbCl.   Nice raft of octahedral crystals grew at the top of a 1 litre Erlenmeyer flask over the space of 2 months using water with 5mL HCl and 1mL HNO3 as solvent.  Cover with watch glass.  Lowest setting on hotplate, airflow at  front corner of fume hood made an excellent thermal gradient.  No stirring, no hoppers formed.  You want a good supply of yellow solid at the bottom of the flask to feed the crystals at the top.  Green solid means you need more HCl and HNO3.

Cesium  Cs2SiF6 easy to make exactly as for K2SiF6. 

Please see separate comments for Cs-Al-Si and phosphates.