Probe Software Users Forum

General EPMA => Discussion of General EPMA Issues => Topic started by: EricKelly on September 02, 2014, 02:42:23 PM

Title: Quantitative Apatite Analysis
Post by: EricKelly on September 02, 2014, 02:42:23 PM
Hi Everyone,

I plan to analyze many (~50) apatite crystals spanning the full range of F and Cl proportions, and I am hoping for advice on an efficient quantitative protocol that includes several trace elements as well (e.g., La, Ce, Mn).

I have read Goldoff et al (2012, Am Min v97 p1103-1115) and Henderson (2011, U Michigan MS thesis), and I have spoken with several people on the topic.  Most agree on these:
However, it sounds like there are two camps on the topic of counting time:
I am mostly interested in advice on counting time (do you agree with 1 or 2?) but other thoughts are certainly welcome.  Thanks!
(Instrument: JEOL 8200)
Title: Re: Quantitative Apatite Analysis
Post by: Probeman on September 02, 2014, 09:48:55 PM
As long as you use a Time Dependent Intensity Correction (TDI), and perform that correction on *both* your unknowns and standards, you shouldn't be required to orient the crystals for quantification.

http://smf.probesoftware.com/index.php?topic=11.0
Title: Re: Quantitative Apatite Analysis
Post by: Anette von der Handt on September 02, 2014, 10:14:00 PM
Hi,

I just finished a big batch of apatite analyses, the following is my routine for apatites in carbonatites and phonolites (so you may want/need other elements) on a JEOL 8900. Overall, I go with approach 2.

Analytical conditions: 15kV, 10 nA and a beam diameter of 15 µm

Element setup:
LDE1: F
PET: Ca, P, S, La, Ce, Cl
TAP: Si, Mg, Na, Sr
LIF: Fe, Mn

Counting times: 60/30 (Pk/bkgr) for F and Cl, 40/20 for Fe, all other 30/15

PFE specific settings:
Measurement options: TDI (but I only use it for F in the end; TDI also on the standards), Asynchronous mode, pick background position and models from a good wavescan
Analytical options: Halogen correction

I usually re-calibrate every six hours at least the apatite standard (Durango) to track any drift. I also carbon coat the standards together with the samples if possible.

Another good paper to read is also: Marks et al. 2012: The volatile inventory (F, Cl, Br, S, C) of magmatic apatite: An integrated analytical approach. Chemical Geology 291, 241 - 255.

I hope this helps. Enjoy your apatites!
Title: Re: Quantitative Apatite Analysis
Post by: EricKelly on September 05, 2014, 08:01:34 AM
Thanks for both of your replies.  I'll let the forum know if I find anything else to add.
Title: Re: Quantitative Apatite Analysis
Post by: camebax on September 05, 2014, 12:17:30 PM
1) Durango apatite is heterogeneous and is thus not recommended as a standard.  Instead, use F-topaz or MgF2 for the F standard.  These two materials are much more robust than any fluorapatite and show very limited F decay with beam exposure.  Synthetic MgF2 should be easy to find.

2) Performing TDI corrections on both the standard and unknown does not assure good results, since the shape of the TDI curve changes with sample crystallographic orientation.  The TDI curve shapes are also complex and not conducive to fitting.  Eliminate at least one source of error by using a robust standard (see above).

3) As Cl concentration increases, you have a better chance of getting good analyses, but I am skeptical of any results on fluorapatites, unless I know that the sample was analyzed with the beam perpendicular to the c-axis.  If the sample is hexagonal, move on. 

4) In general, there will be a systematic over-estimation of F with TDI back-calculation.  The literature is full of EPMA apatite analyses with higher F concentrations than are possible in end-member fluorapatite.

Regards,
Carl




Title: Re: Quantitative Apatite Analysis
Post by: John Donovan on September 05, 2014, 01:36:07 PM
Quote from: chenderson on September 05, 2014, 12:17:30 PM
2) Performing TDI corrections on both the standard and unknown does not assure good results, since the shape of the TDI curve changes with sample crystallographic orientation.  The TDI curve shapes are also complex and not conducive to fitting.  Eliminate at least one source of error by using a robust standard (see above).

Hi Carl,
The reason for performing TDI curves on both the unknown and standard is *exactly* because the crystallographic orientation will generally be different between the unknown and standard!

Now the TDI extrapolation may not always be "conducive" to fitting, but compared to trying to orient one's unknowns to the same crystallographic orientation as the standard, it is quite easy. Also, PFE has an enormous number of TDI fitting options so one can usually do a very nice TDI extrapolation to zero time and Eric is using PFE.

Quote from: chenderson on September 05, 2014, 12:17:30 PM4) In general, there will be a systematic over-estimation of F with TDI back-calculation.  The literature is full of EPMA apatite analyses with higher F concentrations than are possible in end-member fluorapatite.

Yes, true. But this is generally *not* due to the TDI overfitting, instead it is due to the matrix correction being miscalculated by assuming stoichiometric oxygen when the fluorine content is replacing some of that stoichiometric oxygen in the mineral.

How could this matter? Well F Ka is highly absorbed by oxygen, so if the matrix totals say, 106% due to excess stoichiometric oxygen, then the F concentration is overestimated by around 15%.

See here for details on this interesting (and unpublished!) method to automatically subtract the oxygen equivalence of all measured (or specified!) halogens during the matrix iteration:

http://smf.probesoftware.com/index.php?topic=81.msg292#msg292

For the above post, scroll all the way to the bottom for the halogen example.

http://smf.probesoftware.com/index.php?topic=8.msg1127#msg1127
Title: Re: Quantitative Apatite Analysis
Post by: EricKelly on September 05, 2014, 02:38:05 PM
Quote from: chenderson on September 05, 2014, 12:17:30 PM
1) Durango apatite is heterogeneous and is thus not recommended as a standard.  Instead, use F-topaz or MgF2 for the F standard.  These two materials are much more robust than any fluorapatite and show very limited F decay with beam exposure.  Synthetic MgF2 should be easy to find.

Thanks for the input, Carl.  I have both Durango Apatite and F-topaz.  A previous user in our lab said that F in our F-topaz standard migrates much faster than Durango Apatite.  I did not see the data myself, so I will do my own testing to be sure of the better standard.  The heterogeneity in Durango Apatite is certainly a problem.  Perhaps I will map our grains with hopes of finding a good portion of the standard (and then repolish).

Quote from: chenderson on September 05, 2014, 12:17:30 PM
... I am skeptical of any results on fluorapatites, unless I know that the sample was analyzed with the beam perpendicular to the c-axis.  If the sample is hexagonal, move on. 

The orientation debate is difficult to assess given that several studies seem to report different conclusions.  However, many of my samples are already oriented – they come from a previous collection – so I don't have much choice anyway.
Title: Re: Quantitative Apatite Analysis
Post by: EricKelly on September 08, 2014, 12:02:37 PM
In Anette's post, she suggests asynchronous mode but in Probewin.pdf (e.g., p. 216, 218, 252) synchronous is the suggested mode when using the TDI correction.  Has anyone noticed a significant difference between modes?  I would like to keep the analysis fast so asynchronous is my preference, but not if it degrades the TDI correction significantly.

Thanks
Title: Re: Quantitative Apatite Analysis
Post by: Probeman on September 08, 2014, 12:16:16 PM
Quote from: EricKelly on September 08, 2014, 12:02:37 PM
In Anette's post, she suggests asynchronous mode but in Probewin.pdf (e.g., p. 216, 218, 252) synchronous is the suggested mode when using the TDI correction.  Has anyone noticed a significant difference between modes?  I would like to keep the analysis fast so asynchronous is my preference, but not if it degrades the TDI correction significantly.
My word! Someone is actually reading the documentation!  ;)

Actually I have to apologize because the documentation is slightly out of date on that point. It used to be that to get the interval between removing the faraday cup and starting the counting as short as possible, it was necessary to utilize synchronous mode. But not anymore.

It might also be true that this interval is now still slightly shorter using synchronous mode, but I believe that the asynchronous mode is almost as fast if not as fast.

Try turning on the "Debug Mode" and "Time Stamp Mode" in the Output menu and see exactly what the timing differences are- would be worth as post.

I should also note that some investigators such as Stuart Kearns rightfully warn of a non-linearity in the first few seconds of counting in some alkali glasses because the Na and K ion migration doesn't really get starting until the sample heats up from beam exposure. See here for more discussion:

http://smf.probesoftware.com/index.php?topic=116.msg454#msg454

Not sure if that applies to F in apatites though...

Title: Re: Quantitative Apatite Analysis
Post by: BenjaminWade on September 24, 2014, 06:14:19 PM
Hi all
Quick question with regards to MgF2. Where have people sourced this from? I can easily find MgF2 on chemical companies like Sigma Aldrich, but should I be concerned about the purity? Should I be trying to source a single crystal of MgF2?
In addition, has anyone had any success using the Uni of Edin REEFluoride glasses as F standards for apatite analyses?

Cheers
Title: Re: Quantitative Apatite Analysis
Post by: Jeremy Wykes on September 24, 2014, 08:49:22 PM
MTI corp has substrates that would be good for standard material:

http://www.mtixtl.com/crystalssubstratesa-z.aspx

Single crystal, oriented, high purity (99.99%).

Unfortunately, many of the substrates must be purchased in multiples of 4 or more. Are any other labs interested in this material? Our standard collection is so dire that I am thinking of getting most of the simple oxide substrates.
Title: Re: Quantitative Apatite Analysis
Post by: BenjaminWade on September 24, 2014, 09:41:30 PM
Interesting link. Thanks Jeremy, I will have a peruse of that website.

Also to all, with regard to my earlier question about the Uni of Edin REE Fluoride glasses....ignore that, I had a brain fade, they are silicate glasses. However I know Astimex has some LREE Fluoride glasses (which we have). Apart from purity/stoichiometry issues, does anyone have a feeling for F migration in REE fluorides?

Cheers
Title: Re: Quantitative Apatite Analysis
Post by: BenjaminWade on September 25, 2014, 05:17:01 PM
Thanks Owen. Unfortunately we don't have MgF2 on our Taylor block, but we do have PbF2, SrF2, and BaF2 on it. Sounds like I will have to do some TDI tests on these and the Astimex REE F2 to see if any of them are suitable.
Title: Re: Quantitative Apatite Analysis
Post by: BenjaminWade on September 25, 2014, 09:29:48 PM
Ah yeah, didn't think of that. Looking at the MACS it lookes like BaF2 and possibly the REEF2 might be the best of a bad bunch out of that lot. But yes, if you are listening Jeremy, I would be interested in getting some MgF2 from that company if you wanted to go in together. I will have a proper look at the website and see if there is anything else we might need.
Thanks all.
Title: Re: Quantitative Apatite Analysis
Post by: Jeremy Wykes on September 25, 2014, 09:38:37 PM
I have contacted them about getting raw material that is not wafered, oriented, polished or clean room packed.

Let me know what phases you are interested in and I will add that to my list.
Title: Re: Quantitative Apatite Analysis
Post by: BenjaminWade on September 25, 2014, 10:47:48 PM
Hi Jeremy
I had a good look and made a list, looked at the prices and cut some off the list...so I am left with MgF2, TiO2, SiO2, and YVO4. That would be great if you could add them to the list.

Cheers
Title: Re: Quantitative Apatite Analysis
Post by: AndrewLocock on October 08, 2014, 07:54:59 AM
Hello,
An alternative source of high-purity MgF2 is Sigma-Aldrich, catalog number 378836: "Magnesium fluoride, random crystals, optical grade, ≥99.99% trace metals basis, Synonym: Sellaite". A quantity of 5 grams is available for $128 (Cdn), before an academic discount, if any.
Best regards,
Andrew
Title: Re: Quantitative Apatite Analysis
Post by: EricKelly on May 04, 2015, 07:56:51 PM
I am revisiting this topic to post our new protocol for apatite, which might be useful to others.  The most notable part of this is that our biggest improvement came from obtaining new standards.

After getting poor results from Utah Topaz, Fluorite, or Durango Apatite as primary standards for F, we obtained MgF2 (Sigma-Aldrich, 378836 - Thanks for the tip Andrew) and saw big improvements.  We also obtained several of the Schettler et al. (2011, American Mineralogist v96 p138-152) synthetic apatite crystals for Ca, P, and Cl.  Our APS-26 grain works well as a primary standard.  We tried several others for use as secondary standards and found that APS-27 worked well.  Durango apatite now serves as a secondary standard in our lab (This seems to be another example of Durango's variability, given that it works as a primary standard in some labs and not others).

Our goal was to develop a fast protocol to handle a large number of grains.  For those grains that require better analyses, we will return to them with longer counting times and fewer trace elements.  Starting with Anette's protocol and other comments from the forum, we now use the following on our JEOL 8200:

Analytical conditions:
15 kV
10 nA
10 micron spot diameter

Crystal assignments:
LDE (with expanded slit): F
TAP: Na, Si
PETH/J: Ca, P, Cl, Sr, Ce, Nd
LIFH: Fe, Mn

Peak/background counting times:
F: 50/25
Cl, Na, and Si: 40/20
All others: 20/10 (or 30/15 for better analyses)

Frequency of background measurements:
F, Cl, Ca, P: Every analysis
All others: First analysis of a grain only (Nth backgrounds)

Other options:
We use an exponential fit for F backgrounds (and Ce)
We include TDI measurements, but for our samples typically Ca and F are the only elements that need a TDI fit, and Ca rarely needs it.
We use PFE's halogen correction (or apply our own correction).
We recalibrate every 12 hours or so using different spot locations on standards that contain migrating elements, but so far have not seen significant drift.

Thanks again for everyone's help.
Title: Re: Quantitative Apatite Analysis
Post by: Brian Joy on May 05, 2015, 07:32:55 AM
I would urge extreme caution in using MgF2 as a standard for analysis of F in apatite.  In this case the F Ka absorption correction is astronomically large, and hence the accuracy of the correction is not likely to be good.  For instance, using PAP in conjunction with MAC values suggested by Pouchou and Pichoir (in Electron Probe Quantitation) and using 15 kV potential and 40 deg. takeoff angle, f(chi) for F Ka in MgF2 is 0.6090, while in Ca5(PO4)3F it is 0.1355.  This produces an absorption correction factor of 4.494, i.e. a correction of ~350%.  Any small error in MACs or in the phi(rho*z) model could affect calculated wt% F significantly.  Not only is F Ka strongly absorbed by oxygen, but it is roughly equally strongly absorbed by Ca, as it is energetic enough to ionize Ca L1,2,3.  For instance, comparison of F Ka f(chi) in MgF2 and CaF2 gives 0.6090 versus 0.2180.  Other problems notwithstanding (TDI, for instance), if there ever were a case in which a "matrix match" were needed between standard and unknown, the analysis of F in apatite is it.

Also, when analyzing for F using LDE1, are you taking into account the overlap from P Ka(3)?  If you are using pulse amplitude discrimination, have you verified that you have completely eliminated the contribution from P Ka?  For instance, have you done a wavelength scan for F Ka in differential mode on a nominally F-free material such as Ca2P2O7 (calcium pyrophosphate), which has roughly similar wt% CaO and P2O5 as apatite?
Title: Re: Quantitative Apatite Analysis
Post by: Probeman on May 05, 2015, 11:48:41 AM
Quote from: EricKelly on May 04, 2015, 07:56:51 PM
Other options:
We use an exponential fit for F backgrounds (and Ce)

We include TDI measurements, but for our samples typically Ca and F are the only elements that need a TDI fit, and Ca rarely needs it.

We use PFE's halogen correction (or apply our own correction).

Hi Eric,
Very nice.  A couple of questions...

For F ka did you use a 40 or 60 angstrom LDE?

Can you provide some example output from PFE?  I'm especially interested in the difference between the results with and without the halogen - oxygen equivalent correction.

Also maybe could you post a couple graphs of the TDI effects?  I've seen some strange behavior there and wondered how your curves look.
Title: Re: Quantitative Apatite Analysis
Post by: Probeman on May 05, 2015, 12:54:57 PM
Quote from: Brian Joy on May 05, 2015, 07:32:55 AM
I would urge extreme caution in using MgF2 as a standard for analysis of F in apatite.  In this case the F Ka absorption correction is astronomically large, and hence the accuracy of the correction is not likely to be good.  For instance, using PAP in conjunction with MAC values suggested by Pouchou and Pichoir (in Electron Probe Quantitation) and using 15 kV potential and 40 deg. takeoff angle, f(chi) for F Ka in MgF2 is 0.6090, while in Ca5(PO4)3F it is 0.1355.  This produces an absorption correction factor of 4.494, i.e. a correction of ~350%.  Any small error in MACs or in the phi(rho*z) model could affect calculated wt% F significantly.  Not only is F Ka strongly absorbed by oxygen, but it is roughly equally strongly absorbed by Ca, as it is energetic enough to ionize Ca L1,2,3.  For instance, comparison of F Ka f(chi) in MgF2 and CaF2 gives 0.6090 versus 0.2180.  Other problems notwithstanding (TDI, for instance), if there ever were a case in which a "matrix match" were needed between standard and unknown, the analysis of F in apatite is it.

Also, when analyzing for F using LDE1, are you taking into account the overlap from P Ka(3)?  If you are using pulse amplitude discrimination, have you verified that you have completely eliminated the contribution from P Ka?  For instance, have you done a wavelength scan for F Ka in differential mode on a nominally F-free material such as Ca2P2O7 (calcium pyrophosphate), which has roughly similar wt% CaO and P2O5 as apatite?

Hi Brian,
I think you make some good points.

On the standard for fluorine I suspect that a robust standard may be more import than the matrix match.  Here are some comparisons between the Heinrich, Henke and FFAST MACs for these matrices:

MAC value for F ka in O =   12439.63  (LINEMU   Henke (LBL, 1985) < 10KeV / CITZMU > 10KeV)
MAC value for F ka in O =   12390.00  (CITZMU   Heinrich (1966) and Henke and Ebisu (1974))
MAC value for F ka in O =   11927.75  (MAC30    Heinrich (Fit to Goldstein tables, 1987))
MAC value for F ka in O =   11863.62  (FFAST    Chantler (NIST v 2.1, 2005))

MAC value for F ka in Ca =   12415.20  (LINEMU   Henke (LBL, 1985) < 10KeV / CITZMU > 10KeV)
MAC value for F ka in Ca =   12370.00  (CITZMU   Heinrich (1966) and Henke and Ebisu (1974))
MAC value for F ka in Ca =   12623.93  (MAC30    Heinrich (Fit to Goldstein tables, 1987))
MAC value for F ka in Ca =   12132.31  (FFAST    Chantler (NIST v 2.1, 2005))

MAC value for F ka in P =    5550.00  (LINEMU   Henke (LBL, 1985) < 10KeV / CITZMU > 10KeV)
MAC value for F ka in P =    5526.00  (CITZMU   Heinrich (1966) and Henke and Ebisu (1974))
MAC value for F ka in P =    5219.60  (MAC30    Heinrich (Fit to Goldstein tables, 1987))
MAC value for F ka in P =    4905.57  (FFAST    Chantler (NIST v 2.1, 2005))


The MAC for F ka by P is fairly small and the others are larger but in relative agreement, but I do think having a matrix match is ideal as you suggest.

I wonder if we should talk to Marc Schrier about the possibility of synthesizing an end member fluor-apatite?  Wouldn't that be nice to have for everyone?

I already have a synthetic chlor-apatite (just a few grains left, but hydrothermally grown from the University of Nice, see Argiolas and Baumer, Can. Min., v. 16, pp 285-290, 1978), which is wonderful, though a little beam sensitive. Fortunately the Cl intensity doesn't seem to change much over time with mild beam conditions.  I should also mention to Marc him synthesizing this in 100 - 200 gram quantities as well...

By the way, I received about 200 grams of RbTiOPO4 and after we analyze it for traces I will be sending it to Marc Schrier for distribution, so I'll let you know when that is ready.

Back to fluor-apatite though- the question isn't so much the size of the matrix correction relative to the standard, though it is worth considering, but the relative accuracy error on that correction is important I agree. Here is a comparison of two different matrix corrections on a fluor-apatite material (all using the same MACs) done by a student at the USGS so please ignore the accuracy!  I also cut out most of the elements because she analyzed 36 elements in this standard!

First the JTA matrix correction (the default in my software):

ELEM:        W      Ta      Fe      Ti      Nb      Mn      Na       F   SUM 
   100    .000    .000    .079    .000    .011    .037    .157   4.342 101.390
   101    .000    .000    .000    .306    .000    .007    .130   3.655 100.569
   102    .000    .000    .000    .000    .020    .038    .139   4.484 101.617

AVER:     .000    .000    .026    .102    .011    .028    .142   4.160 101.192
SDEV:     .000    .000    .045    .176    .010    .017    .014    .444    .551
SERR:     .000    .000    .026    .102    .006    .010    .008    .256
%RSD:      .16     .19  173.20  173.21   96.92   63.20    9.61   10.66

PUBL:     n.a.    n.a.    n.a.    n.a.    n.a.    n.a.    .178   3.700  99.662
%VAR:      ---     ---     ---     ---     ---     ---  -20.44   12.44
DIFF:      ---     ---     ---     ---     ---     ---   -.036    .460
STDS:      468     467    7852    7840     442    7845    7815    8811

STKF:   1.0000  1.0000   .4985   .5548  1.0000   .1418   .0505   .1537
STCT:   602.94  622.84   44.65   43.13  135.92   13.23   52.31   16.58

UNKF:    .0000   .0000   .0002   .0008   .0001   .0002   .0007   .0083
UNCT:    -1.53    -.86     .01     .05     .00     .02     .68     .89
UNBG:     4.20    3.11     .29     .03     .27     .29     .80     .15

ZCOR:   1.1917  1.2114  1.1812  1.2234  1.2559  1.2066  2.1646  5.0168
KRAW:   -.0025  -.0014   .0002   .0011   .0000   .0016   .0129   .0540
PKBG:      .64     .72    1.04     .52     .99    1.08    1.85    7.21
INT%:     5.62     .30    ----    ----    ----    ----    ----  -18.21


And here is the original PAP (XPP) correction:

ELEM:        W      Ta      Fe      Ti      Nb      Mn      Na       F   SUM 
   100    .000    .000    .079    .000    .012    .037    .153   4.386 101.453
   101    .000    .000    .000    .307    .000    .007    .127   3.715 100.627
   102    .000    .000    .000    .000    .022    .038    .136   4.549 101.685

AVER:     .000    .000    .026    .102    .012    .028    .139   4.217 101.255
SDEV:     .000    .000    .046    .177    .011    .017    .013    .442    .556
SERR:     .000    .000    .026    .102    .006    .010    .008    .255
%RSD:      .13     .15  173.20  173.21   96.94   63.20    9.43   10.48

PUBL:     n.a.    n.a.    n.a.    n.a.    n.a.    n.a.    .178   3.700  99.662
%VAR:      ---     ---     ---     ---     ---     ---  -22.07   13.96
DIFF:      ---     ---     ---     ---     ---     ---   -.039    .517
STDS:      468     467    7852    7840     442    7845    7815    8811

STKF:   1.0000  1.0000   .4892   .5435  1.0000   .1383   .0485   .1498
STCT:   602.94  622.84   44.65   43.13  135.92   13.23   52.31   16.58

UNKF:    .0000   .0000   .0002   .0008   .0001   .0002   .0006   .0081
UNCT:    -1.53    -.86     .01     .05     .00     .02     .68     .90
UNBG:     4.20    3.11     .29     .03     .27     .29     .80     .15

ZCOR:   1.3693  1.3918  1.2122  1.2560  1.3782  1.2383  2.2085  5.2191
KRAW:   -.0025  -.0014   .0002   .0011   .0000   .0016   .0129   .0540
PKBG:      .64     .72    1.04     .52     .99    1.08    1.85    7.21
INT%:     5.62     .30    ----    ----    ----    ----    ----  -18.19


The matrix correction for F Ka goes from 5.0168 to 5.2191 so about 4% which is less than the relative standard deviation. 

Also, the interference from P Ka III is important and I observe that on my Sx100, and it should be a slightly larger overlap on a JEOL instrument. For that of course we'd like to have a material with known P and no F, for instance the chlor-apatite standard I mentioned above. By the way, the above analyses are with the P III interference on F Ka corrected (see the line labeled INT% which is the interference correction percent which is ~18% so that is quite important as you mentioned.

Lots more to discuss here...
Title: Re: Quantitative Apatite Analysis
Post by: Marc Schrier on May 05, 2015, 07:05:42 PM
John, I found the paper at http://rruff.info/uploads/CM16_285.pdf  Unfortunately I took Spanish in school, so I'm not doing really well with this French paper.  Copying and pasting sections into a translator is giving me a sense for the paper, but not to enough detail yet.  I do have a Leco Tem-Pres Reactor, so the 1-3000 bar and 200-850°C (I cannot go beyond 750°C) are possibilities.  But... the reactor volume is just 28.2 cc, and if I use gold tubes like they did, it's dramatically less, so I  would not get very much material.  I did a search to see if I could find any other papers (hopefully one in English too), and I found several where chlorapatite was prepared from a molten flux, CaCl2.  That's much more amenable to bulk syntheses!  Two papers I saw were: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3638812/ and http://journal.chemistrycentral.com/content/7/1/56  Are you guys familiar with chlorapatite prepared in this fashion?     -Marc
Title: Re: Quantitative Apatite Analysis
Post by: John Donovan on May 06, 2015, 08:43:11 AM
Quote from: Marc Schrier on May 05, 2015, 07:05:42 PM
John, I found the paper at http://rruff.info/uploads/CM16_285.pdf  Unfortunately I took Spanish in school, so I'm not doing really well with this French paper.

I'm pretty sure I found the paper in The Canadian Mineralogist...  in English.
Title: Re: Quantitative Apatite Analysis
Post by: EricKelly on May 06, 2015, 10:42:08 AM
Quote from: Brian Joy on May 05, 2015, 07:32:55 AM
I would urge extreme caution in using MgF2 as a standard for analysis of F in apatite.  In this case the F Ka absorption correction is astronomically large, and hence the accuracy of the correction is not likely to be good.  For instance, using PAP in conjunction with MAC values suggested by Pouchou and Pichoir (in Electron Probe Quantitation) and using 15 kV potential and 40 deg. takeoff angle, f(chi) for F Ka in MgF2 is 0.6090, while in Ca5(PO4)3F it is 0.1355.  This produces an absorption correction factor of 4.494, i.e. a correction of ~350%.  Any small error in MACs or in the phi(rho*z) model could affect calculated wt% F significantly.  Not only is F Ka strongly absorbed by oxygen, but it is roughly equally strongly absorbed by Ca, as it is energetic enough to ionize Ca L1,2,3.  For instance, comparison of F Ka f(chi) in MgF2 and CaF2 gives 0.6090 versus 0.2180.  Other problems notwithstanding (TDI, for instance), if there ever were a case in which a "matrix match" were needed between standard and unknown, the analysis of F in apatite is it.

Also, when analyzing for F using LDE1, are you taking into account the overlap from P Ka(3)?  If you are using pulse amplitude discrimination, have you verified that you have completely eliminated the contribution from P Ka?  For instance, have you done a wavelength scan for F Ka in differential mode on a nominally F-free material such as Ca2P2O7 (calcium pyrophosphate), which has roughly similar wt% CaO and P2O5 as apatite?

Hi Brian,

How appropriate that you responded – you were my TA at Davis!

I also prefer to matrix-match the standard to the unknown, but all of the other standards I've tried gave impossible F compositions (>4 wt%).  Of course there may be something else wrong and the MgF2 standard fortuitously brings the F counts into alignment, but I haven't found anything else to blame.  I am happy to hear more suggestions though.

Thanks for bringing up the P III peak issue.  I think I've excluded it in my runs so far.  Here are a couple of screenshots from a wavescan done on a nearly pure chlorapatite.  The first image shows the location of the F peak, and the second shows the P peak.

(https://dl.dropboxusercontent.com/s/aalej7qjvwtlwf4/Wavescan1F.jpg?dl=0)

(https://dl.dropboxusercontent.com/s/stqv63s3vjqkn06/Wavescan2P.jpg?dl=0)
Title: Re: Quantitative Apatite Analysis
Post by: EricKelly on May 06, 2015, 10:59:53 AM
Quote from: Probeman on May 05, 2015, 11:48:41 AM
Quote from: EricKelly on May 04, 2015, 07:56:51 PM
Other options:
We use an exponential fit for F backgrounds (and Ce)

We include TDI measurements, but for our samples typically Ca and F are the only elements that need a TDI fit, and Ca rarely needs it.

We use PFE's halogen correction (or apply our own correction).

Hi Eric,
Very nice.  A couple of questions...

For F ka did you use a 40 or 60 angstrom LDE?

Can you provide some example output from PFE?  I'm especially interested in the difference between the results with and without the halogen - oxygen equivalent correction.

Also maybe could you post a couple graphs of the TDI effects?  I've seen some strange behavior there and wondered how your curves look.

I don't remember off-hand which LDE crystal we have.  I think the spacing is 60 but I'll have to check when I'm back on campus tomorrow.

Here are results from one of the synthetic standards (Schettler et al).  The halogen correction gives slightly better results.  I attached a longer print out in case anyone wants to see more.

Halogen corrected
TDI log-linear fit to F


St  209 Set   1 Apatite APS-27, Results in Oxide Weight Percents

ELEM:     P2O5       F      Cl     MnO     FeO     CaO       O   SUM 
    30  41.505   2.512   2.135    .043    .008  54.989  -3.039  98.151
    31  41.652   2.442   2.192   -.076    .036  54.656  -3.023  97.879
    32  42.261   2.350   2.158    .017   -.004  54.783  -2.976  98.588
    33  42.090   2.350   2.158   -.008   -.009  54.980  -2.976  98.585
    34  41.784   2.509   2.213    .046    .040  54.826  -3.056  98.362
    35  42.034   2.321   2.164   -.039    .019  54.820  -2.965  98.353
    36  42.011   2.518   2.129    .034    .028  55.012  -3.041  98.691
    37  41.605   2.601   2.173    .014   -.010  55.114  -3.086  98.411
    38  42.069   2.413   2.168   -.021    .034  55.166  -3.005  98.824

AVER:   41.890   2.446   2.166    .001    .016  54.927  -3.019  98.427
SDEV:     .260    .095    .026    .041    .020    .166    .041    .288
SERR:     .087    .032    .009    .014    .007    .055    .014
%RSD:      .62    3.90    1.20 4609.42  127.17     .30   -1.35

PUBL:   41.863   2.472   2.290    n.a.    n.a.  54.874  -1.500 100.000
%VAR:      .06   -1.06   -5.42     ---     ---     .10  101.26
DIFF:     .027   -.026   -.124     ---     ---    .053  -1.519
STDS:      211     206     211      42      39     211       0

(https://dl.dropboxusercontent.com/s/4tjagz4hb374j74/TDI%20APS-27.jpg?dl=0)


Without halogen correction

St  209 Set   1 Apatite APS-27, Results in Oxide Weight Percents

ELEM:     P2O5       F      Cl     MnO     FeO     CaO       O   SUM 
    30  41.556   2.521   2.135    .043    .008  55.016  -1.500  99.778
    31  41.703   2.451   2.193   -.076    .036  54.682  -1.500  99.489
    32  42.310   2.359   2.158    .017   -.004  54.808  -1.500 100.148
    33  42.139   2.358   2.159   -.008   -.009  55.005  -1.500 100.145
    34  41.836   2.518   2.213    .046    .040  54.853  -1.500 100.006
    35  42.082   2.329   2.164   -.039    .019  54.845  -1.500  99.901
    36  42.062   2.527   2.130    .034    .028  55.038  -1.500 100.319
    37  41.657   2.611   2.174    .014   -.010  55.141  -1.500 100.087
    38  42.119   2.421   2.168   -.021    .035  55.192  -1.500 100.414

AVER:   41.940   2.455   2.166    .001    .016  54.953  -1.500 100.032
SDEV:     .259    .096    .026    .041    .020    .167    .000    .282
SERR:     .086    .032    .009    .014    .007    .056    .000
%RSD:      .62    3.91    1.20 4608.25  127.17     .30     .00

PUBL:   41.863   2.472   2.290    n.a.    n.a.  54.874  -1.500 100.000
%VAR:      .18    -.70   -5.40     ---     ---     .14     .00
DIFF:     .077   -.017   -.124     ---     ---    .079    .000
STDS:      211     206     211      42      39     211       0


Without the TDI fit (shown next), F matches the published value a bit more closely, but I think the slope is real so I use the TDI fit.

Halogen corrected
No TDI fit


St  209 Set   1 Apatite APS-27, Results in Oxide Weight Percents

ELEM:     P2O5       F      Cl     MnO     FeO     CaO       O   SUM 
    30  41.506   2.560   2.135    .043    .008  54.988  -3.059  98.179
    31  41.653   2.498   2.192   -.076    .036  54.655  -3.046  97.911
    32  42.263   2.490   2.158    .017   -.004  54.780  -3.035  98.668
    33  42.092   2.462   2.158   -.008   -.009  54.978  -3.024  98.650
    34  41.784   2.477   2.213    .046    .040  54.827  -3.042  98.343
    35  42.035   2.384   2.164   -.039    .019  54.819  -2.992  98.389
    36  42.010   2.475   2.129    .034    .028  55.012  -3.023  98.666
    37  41.604   2.534   2.173    .014   -.010  55.115  -3.057  98.372
    38  42.070   2.464   2.168   -.021    .034  55.165  -3.027  98.853

AVER:   41.891   2.483   2.166    .001    .016  54.926  -3.034  98.448
SDEV:     .261    .049    .026    .041    .020    .167    .021    .292
SERR:     .087    .016    .009    .014    .007    .056    .007
%RSD:      .62    1.99    1.20 4607.13  127.17     .30    -.69

PUBL:   41.863   2.472   2.290    n.a.    n.a.  54.874  -1.500 100.000
%VAR:      .07     .41   -5.42     ---     ---     .09  102.28
DIFF:     .027    .010   -.124     ---     ---    .052  -1.534
STDS:      211     206     211      42      39     211       0


Here is an example from our Durango standard.  The TDI fit uses all but one obviously bad analysis.  Note that the published value shown for F is probably wrong as the halogen site is overfilled.  Other reports show about 3.3-3.4 wt% F.

St   60 Set   1 Apatite (Fluor) Durango, Results in Oxide Weight Percents

ELEM:     P2O5       F      Cl     MnO     FeO     CaO    Na2O    SiO2     SO3   As2O3   La2O3   Ce2O3   Pr2O3   Nd2O3       O   SUM 
    21  41.098   3.102    .441    .017    .042  53.983    .187    .351    .370    .092    .468    .591    .060    .180  -1.400  99.582
    22  40.498   3.544    .421    .029    .022  53.808    .187    .351    .370    .092    .468    .591    .060    .180  -1.582  99.039
    23  40.862   3.298    .429    .023    .012  54.027    .187    .351    .370    .092    .468    .591    .060    .180  -1.480  99.469
    25  40.764   3.894    .414   -.028    .004  54.163    .187    .351    .370    .092    .468    .591    .060    .180  -1.728  99.782
    26  41.244   3.805    .418   -.040    .022  54.051    .187    .351    .370    .092    .468    .591    .060    .180  -1.691 100.108
    27  41.182   3.232    .421    .006   -.005  53.968    .187    .351    .370    .092    .468    .591    .060    .180  -1.451  99.653
    28  40.897   3.409    .430   -.032    .070  53.897    .187    .351    .370    .092    .468    .591    .060    .180  -1.527  99.443
    29  40.300   3.403    .423    .033   -.006  53.974    .187    .351    .370    .092    .468    .591    .060    .180  -1.523  98.904

AVER:   40.856   3.461    .425    .001    .020  53.984    .187    .351    .370    .092    .468    .591    .060    .180  -1.548  99.497
SDEV:     .330    .274    .008    .030    .026    .105    .000    .000    .000    .000    .000    .000    .000    .000    .114    .388
SERR:     .117    .097    .003    .010    .009    .037    .000    .000    .000    .000    .000    .000    .000    .000    .040
%RSD:      .81    7.92    2.00 2774.99  128.99     .19     .00     .00     .00     .00     .00     .00     .00     .00   -7.36

PUBL:   40.881   3.530    .373    n.a.    .054  54.020    .187    .351    .370    .092    .468    .591    .060    .180    .005 101.259
%VAR:     -.06   -1.96   13.87     ---  -62.98    -.07     .00     .00     .00     .00     .00     .00     .00     .00-28754.95
DIFF:    -.025   -.069    .052     ---   -.034   -.036    .000    .000    .000    .000    .000    .000    .000    .000  -1.553
STDS:      211     206     211      42      39     211       0       0       0       0       0       0       0       0       0

(https://dl.dropboxusercontent.com/s/ll3hi53sg9uvf4w/TDI%20DurAp%208%20Lines.jpg?dl=0)


The scatter is difficult to work with so I tend to get 7-10 analyses and disable several of them.   Here is another try with more analyses removed.  The F composition is more reasonable.  If I also remove lines 21 and 27, F comes out to 3.35 wt%.

St   60 Set   1 Apatite (Fluor) Durango, Results in Oxide Weight Percents

ELEM:     P2O5       F      Cl     MnO     FeO     CaO    Na2O    SiO2     SO3   As2O3   La2O3   Ce2O3   Pr2O3   Nd2O3       O   SUM 
    21  41.098   3.102    .441    .017    .042  53.983    .187    .351    .370    .092    .468    .591    .060    .180  -1.400  99.582
    23  40.862   3.298    .429    .023    .012  54.027    .187    .351    .370    .092    .468    .591    .060    .180  -1.480  99.469
    27  41.182   3.232    .421    .006   -.005  53.968    .187    .351    .370    .092    .468    .591    .060    .180  -1.451  99.653
    28  40.897   3.409    .430   -.032    .070  53.897    .187    .351    .370    .092    .468    .591    .060    .180  -1.527  99.443

AVER:   41.010   3.260    .430    .003    .030  53.969    .187    .351    .370    .092    .468    .591    .060    .180  -1.465  99.537
SDEV:     .155    .128    .008    .025    .033    .054    .000    .000    .000    .000    .000    .000    .000    .000    .053    .098
SERR:     .078    .064    .004    .012    .017    .027    .000    .000    .000    .000    .000    .000    .000    .000    .027
%RSD:      .38    3.93    1.92  722.05  112.26     .10     .00     .00     .00     .00     .00     .00     .00     .00   -3.62

PUBL:   40.881   3.530    .373    n.a.    .054  54.020    .187    .351    .370    .092    .468    .591    .060    .180    .005 101.259
%VAR:      .32   -7.64   15.38     ---  -45.07    -.10     .00     .00     .00     .00     .00     .00     .00     .00-27215.08
DIFF:     .129   -.270    .057     ---   -.024   -.052    .000    .000    .000    .000    .000    .000    .000    .000  -1.470
STDS:      211     206     211      42      39     211       0       0       0       0       0       0       0       0       0

(https://dl.dropboxusercontent.com/s/ordr7dyrtdm9z2z/TDI%20DurAp%204%20Lines.jpg?dl=0)


Here is the log-linear fit.

St   60 Set   1 Apatite (Fluor) Durango, Results in Oxide Weight Percents

ELEM:     P2O5       F      Cl     MnO     FeO     CaO    Na2O    SiO2     SO3   As2O3   La2O3   Ce2O3   Pr2O3   Nd2O3       O   SUM 
    21  41.103   3.445    .441    .017    .042  53.977    .187    .351    .370    .092    .468    .591    .060    .180  -1.545  99.779
    23  40.863   3.404    .429    .023    .012  54.025    .187    .351    .370    .092    .468    .591    .060    .180  -1.525  99.530
    27  41.186   3.504    .421    .006   -.005  53.962    .187    .351    .370    .092    .468    .591    .060    .180  -1.565  99.808
    28  40.899   3.509    .430   -.032    .070  53.895    .187    .351    .370    .092    .468    .591    .060    .180  -1.569  99.501

AVER:   41.013   3.466    .430    .003    .030  53.965    .187    .351    .370    .092    .468    .591    .060    .180  -1.551  99.654
SDEV:     .157    .050    .008    .025    .033    .054    .000    .000    .000    .000    .000    .000    .000    .000    .020    .162
SERR:     .078    .025    .004    .012    .017    .027    .000    .000    .000    .000    .000    .000    .000    .000    .010
%RSD:      .38    1.45    1.92  722.17  112.27     .10     .00     .00     .00     .00     .00     .00     .00     .00   -1.32

PUBL:   40.881  3.530    .373    n.a.    .054  54.020    .187    .351    .370    .092    .468    .591    .060    .180    .005 101.259
%VAR:      .32   -1.83   15.38     ---  -45.07    -.10     .00     .00     .00     .00     .00     .00     .00     .00-28814.16
DIFF:     .132   -.064    .057     ---   -.024   -.056    .000    .000    .000    .000    .000    .000    .000    .000  -1.556
STDS:      211     206     211      42      39     211       0       0       0       0       0       0       0       0       0

(https://dl.dropboxusercontent.com/s/y7rcbjlmaqn6dpu/TDI%20DurAp%204%20Lines%20Linear.jpg?dl=0)

As I stated in my previous post, we now use our Durango as a secondary standard.  Ca, P, and Cl tend to come out well most of the time, but F is variable, partly due to TDI fits, but heterogeneity is probably a factor, and our standard could use a new polish too.

Edited by John to make output results use a fixed width font (see the Tt button)
Title: Re: Quantitative Apatite Analysis
Post by: Brian Joy on May 06, 2015, 01:45:01 PM
Quote from: EricKelly on May 06, 2015, 10:42:08 AM
Hi Brian,

How appropriate that you responded – you were my TA at Davis!

Hi Eric,

I was wondering if you were the same Eric Kelly.  Good to hear from you.

By the way, since your F Ka peak is near 85 mm, then you are using LDE1, and the 2d is roughly 60 angstroms.

Brian
Title: Re: Quantitative Apatite Analysis
Post by: Jeremy Wykes on May 07, 2015, 02:58:21 AM
Quote from: Marc Schrier on May 05, 2015, 07:05:42 PM
Two papers I saw were: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3638812/ and http://journal.chemistrycentral.com/content/7/1/56  Are you guys familiar with chlorapatite prepared in this fashion?     -Marc

I have made an attempt using Klemme et al method, and you can readily produce mm sized crystals. Obviously, the crystal in the paper is the absolute best one they made, so ours did not look quite so nice.