Hi All,
I'm trying to analyze some serpentines and am wondering if anyone had any suggestions on how to improve my totals, or if this is just what it is for serpentines. I'm consistently getting totals that are 3-7% low after I specify as much water as I think I can reasonably add in. I'm at low current (tested even lower currents and if anything got worse totals), I'm using TDI (really only see a drift in the Mg), and I don't see any missing elements in EDX (further tested this by tuning up the whole geologic suite and running a wavescan... only saw a couple of tiny peaks that added in a fraction of a percent to my totals). I get good totals on my biotite standard run as an unknown (the best option that I really have against serpentines). I've never run serpentines before so I'm wondering if there's any tricks I'm not thinking of or if this is pretty much what to expect from a high water content sheet mineral with a funky structure?
-J
Hi Jeb,
What's the polish like, serpentine is tricky to polish. Does it contain any ferric iron? Are you using a defocused beam?
Hi Ben,
It's a polished thin section (not prepped by me), and the polish looks good. I do think that I'm analyzing from the top of a sheet down and not the side of the book (that's just how it's exposed in the sample).
I am using a defocused beam.
And it contains minor amounts of Fe, ~0.5-1.5wt%, I don't know if it's ferrous vs ferric. Oxygen is by stoichiometry so I do see some difference in the totals depending on what oxidation state I assign, but the totals are low regardless.
-J
Quote from: jeb on April 29, 2025, 11:00:42 AMI do think that I'm analyzing from the top of a sheet down and not the side of the book (that's just how it's exposed in the sample).
The ion migration should be minimal in this case I think. And you tried both lower beam currents and a defocused beam?
Can you show us some results and how you're calculating water by stoichiometry?
Yes, my beam is defocused and the attempts that I've got the best results for have been at 15nA. I tried a round at 5nA and the totals were 1-2% lower. Here's an example, this area had the best totals. The areas with lower totals have basically the same Si content, but the Mg is a couple percent lower and the Fe is ~1.2wt%. I specified the water concentration to be 14% (the highest that it seemed like I could realistically include from looking at DHZ). I turned the TDI off for everything but the Mg as it just looks like noise for the others. Would actually love to hear if there's something just obviously wrong with how I'm calculating things!
Un 5 Area 4, Results in Elemental Weight Percents
ELEM: Si Al Mg Ca Fe Cl S H O
TYPE: ANAL ANAL ANAL ANAL ANAL ANAL ANAL SPEC CALC
BGDS: MAN MAN MAN MAN MAN LIN LIN
TIME: 60.00 180.00 180.00 60.00 180.00 60.00 60.00 --- ---
BEAM: 14.85 14.85 14.85 14.85 14.85 14.85 14.85 --- ---
ELEM: Si Al Mg Ca Fe Cl S H O SUM
78 19.281 .016 24.939 .011 .666 .134 .015 1.567 51.124 97.753
79 19.253 .006 25.213 .011 .478 .056 .023 1.567 51.182 97.788
80 19.391 .010 25.254 .012 .646 .152 .012 1.567 51.443 98.487
81 19.518 .004 24.773 .017 .813 .092 .014 1.567 51.339 98.136
AVER: 19.361 .009 25.044 .013 .651 .108 .016 1.567 51.272 98.041
SDEV: .120 .005 .229 .003 .137 .043 .005 .000 .146 .344
SERR: .060 .003 .114 .002 .069 .021 .003 .000 .073
%RSD: .62 57.51 .91 25.24 21.07 39.65 32.26 .00 .28
STDS: 112 112 112 358 112 285 730 --- ---
STKF: .1368 .0383 .0528 .1693 .1205 .0602 .5062 --- ---
STCT: 53.04 168.15 76.65 310.88 37.42 60.96 179.98 --- ---
UNKF: .1411 .0001 .1776 .0001 .0053 .0009 .0001 --- ---
UNCT: 54.71 .25 257.73 .21 1.64 .87 .04 --- ---
UNBG: .11 2.14 .47 1.34 .31 .34 .10 --- ---
ZCOR: 1.3724 1.6207 1.4106 1.1294 1.2296 1.2651 1.2873 --- ---
KRAW: 1.0314 .0015 3.3624 .0007 .0439 .0142 .0002 --- ---
PKBG: 514.99 1.12 552.45 1.15 6.29 3.52 1.49 --- ---
INT%: ---- ---- ---- ---- ---- ---- ---- --- ---
TDI%: .000 .000 -1.386 .000 .000 .000 .000 --- ---
DEV%: .0 .0 .1 .0 .0 .0 .0 --- ---
TDIF: ---- ---- LOG-LIN ---- ---- ---- ---- --- ---
TDIT: .00 .00 217.25 .00 .00 .00 .00 --- ---
TDII: ---- ---- 258. ---- ---- ---- ---- --- ---
TDIL: ---- ---- 5.55 ---- ---- ---- ---- --- ---
Un 5 Area 4, Results in Oxide Weight Percents
ELEM: SiO2 Al2O3 MgO CaO Fe2O3 Cl S H2O O SUM
78 41.250 .031 41.356 .016 .953 .134 .015 14.000 .000 97.753
79 41.189 .012 41.810 .015 .683 .056 .023 14.000 .000 97.788
80 41.485 .020 41.878 .016 .924 .152 .012 14.000 .000 98.487
81 41.755 .008 41.081 .024 1.162 .092 .014 14.000 .000 98.136
AVER: 41.420 .017 41.531 .018 .931 .108 .016 14.000 .000 98.041
SDEV: .258 .010 .379 .004 .196 .043 .005 .000 .000 .344
SERR: .129 .005 .190 .002 .098 .021 .003 .000 .000
%RSD: .62 57.51 .91 25.24 21.07 39.65 32.26 .00 86.07
STDS: 112 112 112 358 112 285 730 --- ---
Hello,
The trioctahedral serpentine minerals (chrysotile, lizardite, antigorite) can be considered to have a formula analogous to the trioctahedral chlorite minerals (e.g., chamosite, clinochlore):
2 * Mg3Si2O5(OH)4 = Mg6Si4O10(OH)8; similar to Mg5Al(Si3Al)O10(OH)8
I have an Excel spreadsheet at: https://www.eas.ualberta.ca/eml/files/Chlorite_spreadsheet2015.xlsx
It calculates both hydroxyl content and estimates ferric- and ferrous-iron proportions (but see below).
However, its assumptions for ferric iron calculation are only truly correct for trioctahedral phases.
This spreadsheet does not deal with dehydrogenation (oxygen in place of hydroxyl).
In principle, it can be used for both trioctahedral chlorite and trioctahedral serpentine minerals.
Like, Jeb above, I analyze serpentine and chlorite group minerals at 15 kV, 15 to 20 nA, a defocused beam of 5 to 10 microns, and a count time on peak (= backgrounds) of 20 or 30 seconds.
Time-dependent instensity correction for Mg, Fe, and Si may help.
I tend to use olivine as the standard for both Mg and Si for analysis of Mg-rich serpentine.
In Probe-for-EPMA software, I usually assume 8 hydrogen for 18 oxygen (H:O = 0.44444) in the Analyze! | Calculation Options dialogue box.
I don't use a fixed amount of H2O.
The polish is often not great on serpentine (or chlorite), and so totals tend toward 98 wt% or so.
I rarely bother to try and calculate ferric iron in chlorite or serpentine in Probe-for-EPMA because of the large uncertainty in charge balance that results from error propagation of the measured elements.
Finally, it is conceivable that the serpentine has a significant amount of absorbed/included water.
Best of luck with your analyses.
Andrew