Standards Which Should Be Developed For EPMA Next

[crystalgrower]

I agree completely.

In addition, please note that XRF-WDS is a fast and sensitive tool for the trace impurities in both starting materials and final crystals.  Loose grains give a very reliable result if they are anywhere near heat stable. If you are lucky then the peak and bgd settings are close to those used for EPMA.  I think that unless you are using LA, then ICP-MS is good for some starting materials but by no means all products. 

I have been working on monazite synthesis on and off.  Have some good starting materials and a process to recrystallize into mm size crystals.  Process will restart in the fall after I finish moving.

Right now a wishlist for monazite would be helpful.  Th-free?  High-Th (12%) ?  Mid-Th (8%)? Low-Th(4%)?  La-Ce-Nd synthetic  mix to correct assays of other trace REE?

And oh yes single element RPO4 for La-Gd would also be part of the project.  Free of any flux traces, of course.

[Mike Jercinovic]

Yes, this is a very interesting approach John.  The zero test tells you a lot, and, as you suggest, the analysis of inhomogeneous materials can be quite useful indeed, and matrix matching is a critical part of this.  Of course, synthetic monazite would be useful as suggested by crystalgrower, maybe even as a primary standard if it could be synthesized homogeneously.  Any such reference material for blank testing, however, would need a range of Th concentrations, and a few wt.% Y, and at least a typical La, Ce, Nd concentration.  The range of Th in natural monazite can be very large (less than 1 wt.% to almost 20 wt.% but is most typically in the 1-8 wt.% range.  The measurements of Pb and U are complicated by interferences by Th and Y mostly, and in the case of UMb, by the Th M4 and M5 absorption edges.  When Philippe Goncalves synthesized some monazite here years ago, he could get the REE phosphates, and a mixed LREE phosphate, but adding Th became interesting as the products were very heterogeneous.  But this is still useful as you suggest!

[Probeman]

Yes, this is a very interesting approach John.  The zero test tells you a lot, and, as you suggest, the analysis of inhomogeneous materials can be quite useful indeed, and matrix matching is a critical part of this.  Of course, synthetic monazite would be useful as suggested by crystalgrower, maybe even as a primary standard if it could be synthesized homogeneously.  Any such reference material for blank testing, however, would need a range of Th concentrations, and a few wt.% Y, and at least a typical La, Ce, Nd concentration.  The range of Th in natural monazite can be very large (less than 1 wt.% to almost 20 wt.% but is most typically in the 1-8 wt.% range.  The measurements of Pb and U are complicated by interferences by Th and Y mostly, and in the case of UMb, by the Th M4 and M5 absorption edges.  When Philippe Goncalves synthesized some monazite here years ago, he could get the REE phosphates, and a mixed LREE phosphate, but adding Th became interesting as the products were very heterogeneous.  But this is still useful as you suggest!

Cool.

So crystalgrower should look at synthesizing something along the lines of the chemical ranges you suggest, but she should just be sure to obtain U and Pb free starting materials.

Of course the final products can be checked for U and Pb to be below EPMA sensitivities using mass spec techniques in any case.

[Mike Jercinovic]

Yes, U and Pb free would be great.  But after that, some U bearing, with and without Th, would be awesome as these actinides are so intertwined.  Now and then, we see a monazite with more U than Th in nature, but this is important just to sort out actinides at high spatial resolution in general - that is, lower kV so not using the U L line.  We want everything of course, whether is it reasonable or not!  hashtag delusional...

[Probeman]

Yes, U and Pb free would be great.  But after that, some U bearing, with and without Th, would be awesome as these actinides are so intertwined.  Now and then, we see a monazite with more U than Th in nature, but this is important just to sort out actinides at high spatial resolution in general - that is, lower kV so not using the U L line.  We want everything of course, whether is it reasonable or not!  hashtag delusional...

Yes, but since it's difficult to produce non-zero values (and maybe even impossible to dope trace levels homogeneously in some materials), but easy to produce (homogeneous) zero values. Let's start with easy!   

[crystalgrower]

Sometime in the next few weeks I will post a table of ideas.

I made one big mistake last time I made mixed cation phosphates--I did not ask enough people for ideal compositions of mixes.  At that time I had a set of single element EDS spectra to test overlaps by plotting, but no WDS.  Got those right after I finished a 2010 AGU poster...

I found a way to separate black crystals  (I suspect UO2) from natural monazite.  That means the U can be crystallized into something like Ca(Bi,U)(PO4)2.  The Pb would be the natural "old" mineral amount--please advise if that is acceptable.

The "natural" monazite is a lower priority--I was thinking of recvrystallizing it with added Ca to improve homogeneity without losing too much of the Th.  Only the natural accumulated Pb would be present.  It would be Si-free.

Synthetics are in my world made from "clean" materials.  That means using at least 99.99% Nd and Pr.  NO Pb in any reagent, and NO U in the Th. 

There is one exception--any Th would have some Ra present since it ingrows quickly to equilibrium.  I could remove natural decay product Ra and Pb if that is considered part of the Pb-free requirement.  But the final material would have Ra present.

 

[Probeman]

Hi CG,
That sounds great if you can produce synthetic monazite with no Pb or U.

Again, for the purposes of testing our zero measurements in monazite, we would like highly zoned crystals that include the elements that Mike mentioned above, and also that cover the compositional ranges he discussed. Do you think this is possible?  Or would it require more than one synthesis?

I'll bet you never thought you'd be asked to grow *zoned* crystals!  But as long as they are U and Pb free, they would be perfect for testing our backgrounds and interference corrections. This material could add a whole new level of accuracy to our trace monazite measurements.
john

[crystalgrower]

Welll...if you insist...

I actually grew a layer of pure CePO4 on top of a natural monazite recently.  Which probe experts want to play with this?  I don't know what the gradient looks like. I have never been allowed to use the local EPMA or SEM. 

[Probeman]

Welll...if you insist...

It would be totally awesome if you could develop such a material.  So can that range of monazite compositions, that Mike described above, be created in a single synthesis?  This might even be worth a short paper if you are interested.

I actually grew a layer of pure CePO4 on top of a natural monazite recently.  Which probe experts want to play with this?  I don't know what the gradient looks like. I have never been allowed to use the local EPMA or SEM.

I defer to Mike Jercinovic on this question, but I suspect that anything involving natural compositions won't be all that useful, as we need to really need to know "a priori" that there is zero U and Pb in the blank material for testing zero measurement accuracy.

But maybe this type of material would be interesting for a different experimental measurement?  Not being a geologist I can't really say, but I will give it some thought.
john

[crystalgrower]

I suggested a material which is already made to see if any improvements are needed.  I understand that it is not ideal.  But it will make a useful first trial.  I just cannot assess it myself. One of the two layers is certainly free of Pb and U. The other is from Petaca  NM which has been dated and assayed before.  There is also some of the "clean" CePO4 to test separately for absence of U and Pb.

[Probeman]

I suggested a material which is already made to see if any improvements are needed.  I understand that it is not ideal.  But it will make a useful first trial.  I just cannot assess it myself. One of the two layers is certainly free of Pb and U. The other is from Petaca  NM which has been dated and assayed before.  There is also some f the "clean" CePO4 to test separately for absence of U and Pb.

Hi CG,
I appreciate that.  Thanks.

The thing is, in order to test our zero measurements we need a synthetic monazite that has the all the elements that Mike mentioned and in the ranges that he mentioned.  Basically a zoned synthetic monazite that we *know* does not contain *any* Pb or U. 

This will allow us to test the absolute accuracy of the background positions for off-peak interferences and also test the absolute accuracy of the interference corrections for on-peak interferences.   Unfortunately a pure CePO4 doesn't really test either of those issues. And in the case of the natural composition seed crystal, knowing that the natural monazite core been analyzed previously doesn't help us either, because we can only truly trust a value of "absolute" zero (or at least a value less than the best possible EPMA sensitivity). 

As Mike said when I discussed this with him a few days ago, this sort of accuracy test is a very different way of approaching trace element standards and therefore requires a very different type of standard material than we traditionally utilize for EPMA trace element measurements.

If you want to call me some time please feel free and I can probably explain it better by skype or phone.

[Mike Jercinovic]

Right, the CePO4 is certainly not that good a material for the blank test, but is, of course excellent as a primary standard (same with all the REE phosphates).  We also have a piece of the Petaca NM monazite, which is both heterogeneous compositionally as well as having lots of inclusions...  looks a lot like some of the other large pegmatite monazites from northern NM (like the Wards, Elk Mt. monazite).  Not a super useful one for us compared to some of the cleaner ones from Brazil (Minas Gerais).  Natural monazite is very useful as a secondary reference material.  For the blank testing, a synthetic monazite with at least LREEs, Th, and Y is certainly most useful.  As John says, heterogeneity is fine (even preferable), or at least of range of Th/LREE.

[crystalgrower]

This is exactly the kind of detailed info that I need.  I need to use the pure CePO4 as seeds.  You can ignore its presence. 

So to confirm:  I should start with a layer of pure La-Ce-Nd-Y mix (free of U and Pb) and then add Th+Ca.  Do you care if the Ce+3 goes to Ce+4 in that Th layer? (means slightly higher Ca). 

Extra base mix can be run by XRF to show traces present. 

[Mike Jercinovic]

Yes, awesome.  The Ce+3 to Ce+4 should not matter.  Only a few wt.% Y at most.  I don't think the structure will accommodate a much larger amount anyway and still stay monoclinic.