Monazite end member calculation

Hi everyone, bear with me on this little issue, my background in chemistry is not fantastic, and my knowledge of calculating formulas is rather limited.

I have microprobe data on several monazite grains that I would like to plot up as a monazite-huttonite-cheralite ternary diagram. Unfortunately, I can't find many resources on how one divides up the elements between the end members. I have the recent paper by Linthout, which offers a few clues, but I suspect it isn't the whole story, with much of the process being assumed information for people reading the paper. It seems much more complex than the standard amphibole calculations, etc.

I have tried to work it out the following way. It's certainly not very good, but it seems to give answers that plot with monazites from similar settings:

Divided Wt% elements by atomic number, giving relative mol amounts (I doubt this is quite right, can't be this easy)

REE, Al, Y into Monazite
Ca, Pb, Sr, Fe, U into Cheralite (don't know the valence of Fe, but amount is quite small)

All Si into Huttonite, use this value to figure out the required amount of O and Th
Use (Ca, Pb, Sr, Fe, U, remaining Th) to figure out P and O needed for Cheralite component

The left over P and O needed to balance out the Monazite components is pretty close, though there is a small amount left over. What is the right way to do this? I imagine this method might make a few people laugh, but it's the best I could think of

Here's an analysis of one of the monazites in weight percent:

Si 0.04
Al 0.01
K 0.03
Fe 0.01
Ca 2.93
Sr -0.06
Pb 0.64
Y 1.39
Th 8.15
U 7.97
La 6.48
Ce 14.40
Pr 1.77
Nd 6.18
Sm 4.49
Eu -0.17
Gd 3.48
Dy 0.95
Yb 0.05
Er -0.01
P 12.80
S 0.00
As -0.02
O 26.65
Total 98.41

I get a composition of 38.3% Monazite, 61.3% Cheralite, .3% Huttonite from this one. Someone come and tell me how wrong I am please

Hi Ryan,
it is MUCH MORE simple to calculate end members contents not from analzis itself, but from mineral formula calculated from this analyze.

Besides that you should to have in view at least 5 end members - monazite-(LREE), xenotyme-(Y&HREE), huttonite, coffinite, cheralite and apparently some other, if you had obtain residue after your recalculation.

What about your results, there is suspiciously many of cheralite end member for such composition.

Hi Ryan,

I'll run it through my spreadsheet (in the next few days) and let you know what I get.

Thanks Stuart! That would be fantastic. I expect fairly low huttonite/xenotime substitution due to the very low Si. Pavel is right, the high cheralite I've been getting is a bit suspect, but I can't really discount it until the correct calculation is done. The monazites are coming from a likely environment for cheralite, being a highly fractionated LCT pegmatite. High Th/U levels, however a consistently low Th/U ratio amongst my data is interesting. Reading up on these, it seems that a variable, but generally high Th/U ratio is normal.

Don't forget Boga has plenty of cheralite.

why do you have negative figures on Sr, Eu and As?

Johan, thats a really good question, I'm sure a lot of people here are more up to speed with microprobe theory than I am of course, but I don't know the answer completely myself. I'll have a crack at trying to explain anyway. As I understand it, it is a limitation of the method that occurs when a spectral peak corresponding to two elements overlap. The algorithim that tries to deconvolute the overlapping peaks sometimes will give an element that is near the detection limit a negative value. Or it could be that after the spectrum is corrected for background, there is literally nothing there but none the less the software tries to make something out of nothing.

I've seen through my own mucking about that (large) negative figures are disastrous when normalising to 100%, but is it neccesary to correct for them in this case? Or can you simply just replace them with 'not detected'?

I suppose this is the point where the old saying ' you don't need to know how an engine works to drive a car' starts to break down. I've stopped going for sunday drives and started rallying.

40.2 monazite, 44.9 cheralite, 10.7 xenotime, 0.04 huttonite from my quick play at it. So it sits basically where cheralite used to be before it turned into brabantite. If you cf LB the high U are always in the old cheralitic field.

Thats an interesting result Stuart, I didn't expect such a high xenotime component. Am I right in assuming that under the current classification system that it would still be called cheralite? Forgive me if I'm wrong but I was under the impression that brabantite as a name was done away with.

Any chance I could get a copy of that spreadsheet? You actually probably have a copy of my data that came back from UTAS...

I'll have to follow up on these monazites in the future I think. A quick read of the Monazite-(Sm) paper by Masau et al in the Canadian Mineralogist notes that monazite is meant to be pretty scarce in Li,P rich pegmatites, which is exactly where this one came from. Odd locality for lepidolite-amblygonite pegmatites too for sure.

Hi Ryan,

"Divided Wt% elements by atomic number, giving relative mol amounts (I doubt this is quite right, can't be this easy)"

It's almost that easy. You must divide wt% by atomic mass, not atomic number.

Haha yes, I saw what I had written some time afterwards, and hoped everyone would glaze over it. That's exactly what I've been working in, atomic mass. My formulas are still rubbish though. I found a paper apparently correct formulas in it eventually, but even after trying to replicate them in Excel, the results were not satisfactory. For the same analysis that Stuart ran, it gave me something around 60% monazite, 19% cheralite, 9% huttonite 12% xenotime. I fiddled with the Ca calculations in the cheralite part in every concievable way, but that only made it worse. Those formulas did not take into account Si, P and O though, so thats probably where the wierd numbers are coming from. I guess it's assumed knowledge how to incorporate those into the calculations, but it's beyond my current level of understanding. I love getting down into the nitty gritty of geological problems, but not being much of a natural at the mathematical side of chemistry, my own ignorance is letting me down.

I forgot about this one, I don't think I have enough information to solve this one at the moment, but I think it's interesting enough to show everyone.

This grain occurs in metapelite directly adjacent to one of the pegmatite dykes I am studying, and I would have called it uranothorite were it not for the presence of a little P, Ca and LREE (though rather close to detection), which indicates a monazite group affinity to me. It has rimmed a diagenetic pyrite, and is stacked full of little galena inclusions. I was running out of time when I found it and straight out forgot to put yttrium in the EDS analysis, which pretty much wrecks my chances of figuring out it's true composition from what I've learned in this thread so far, but I don't seem to remember it showing up very strongly. The brighnesses in the element map I have attached aren't really relative to each other either, and the Fe/Pb channels are overlapping. The composition is a little variable across the grain, so there may be domains of different composition.

Not really after a clear cut determination at this stage, I can always revisit the thin section later next year. Just chasing some opinions, it kind of looks, walks and quacks like a huttonite with minor cheralite, etc with the data i have...

Normalised EDS:

Compound %
SiO2 15.98
P2O5 4.83
CaO 1.29
FeO 2.19
La2O3 0.07*
Ce2O3 0.85*
Pr2O3 0.28*
Nd2O3 1.04
Sm2O3 0.66*
Gd2O3 0.32*
Tb2O3 0.23*
Dy2O3 0.55*
PbO 1.09
ThO2 65.55
UO3 5.07

Huttonite should have about 18% Si, so that's what this is with a little cheralite component.

Neat! I think the groundwater from Lake Boga must flow towards my way, instilling me with the unconsious drive of 'find similar monazite group minerals'.

Just remember to reference me

Absolutely, I already have you for your help with the other monazites, and the Lake Boga paper is in the thesis a fair bit, as one of the only other victorian references on similar environments.

You can get something from Whycheproof, but it isn't as exciting as LB. See this also (if you havent already): Barium contents of granites: key to understanding crustal architecture in the southern Lachlan Fold Belt? Rossiter A G, Gray C M, Australian Journal of Earth Sciences, 2008, 55, 433-448. ISSN: 0812-0099.