Looking for suggestions for "Basic EDS standards"

[Nicholas Ritchie]

I think it would be very helpful for promoting standards-based EDS analysis if the community were to provide a recommended set of basic EDS standards for each accessible element in the periodic table.  These materials could be build into standard blocks which commercial vendors or laboratories could build in volume (relatively speaking).  Such blocks would lower the barrier to standards-based analyses.

The basic EDS standards would be either pure elements or simple stoichiometric compounds.  This way anyone could reproduce them and know that there composition was what it was claimed to be.

Ideally, the samples would be robust and not susceptible to environmental degradation or degradation under the beam. Ideally, the samples will be available in "bulk" form. (At least 100 µm on edge) Ideally, they would not require any maintenance that requires expensive equipment.  They could be polished by hand following provided instructions and then recoated with carbon as necessary.

The choice of standards need not be optimal for all analyses.  Sometimes perfect is the enemy of good enough.  But because they were shared and available to all, they could encourage people who would like better analyses but find the task too daunting to perform standards-based analyses.

Furthermore, these basic standards could be the foundation for a set of "digital standards" that allow users to trade k-ratios rather than materials to optimize their measurements using "matrix matched standards."  This would bring the best of both worlds - easier standards-based analyses and more accurate analyses.

One requirement for these basic EDS standards that is distinct from WDS standards is that the standards should avoid peak interferences.  Pure elements are easy but many stoichiometric compounds have interferences that are not an issue for WDS but make unsuitable for EDS.  It isn't that EDS can't handle interferences; it is simply that ensuring the basic EDS standards don't have interferences allows them to double as peak shape references.  This greatly simplifies quantification for novices.

I have some ideas based on my experience what I'd suggest for most elements.  But I also have many questions that the WDS community could help with.  Frequently, I wonder whether it would be better to choose a robust material like "FeS2" over "Pure Fe", "Al2O3" over "Al" because they are likely to age better.  Some elements are really hard.  Some elements seem to lack obvious robust alternatives. Is there something better than "NaCl" or "KCl" for chlorine?  "RbI" for rubidium?

My "starter list"

H – N/A
He – N/A
Li – N/A
Be – Be
B – BN
C – C
N – BN
F – CaF2
Ne – N/A
Na – NaAlSi3O8
Mg – MgO
Al – Al2O3
Si – SiO2
P – GaP
S – FeS2
Cl – NaCl, KCl
Ar – N/A
K – KBr
Ca – CaF2
Sc - Sc
Ti - Ti
V - V
Cr - Cr
Mn -Mn
Fe – Fe
Co - Co
Ni - Ni
Cu - Cu
Zn - Zn
Ga - GaP
Ge - Ge
As - ?
Se  - Se
Br - KBr
Kr – N/A
Rb - RbI
Sr – SrF2
Y - Y
Zr - Zr
Nb - Nb
Mo - Mo
Tc – N/A
Ru - Ru
Rh - Rh
Pd - Pd
Ag - Ag
Cd - Cd
In - In
Sn - Sn
Sb - Sb
Te - Te
I – RbI
Xe – N/A
Cs – CsCl
Ba – Sanbornite (BaSiO5) and
La – LaB6, LaP5O14
Ce – CeAl2
Pr – PrP5O14
Nd – NdP5O14
Pm – PmP5O14
Sm – SmP5O14
Eu – EuP5O14
Gd – GdP5O14
Tb – TbP5O14
Dy – Dy, DyF3
Ho – HoF3
Er – Er, ErF3
Tm – TmF3
Yb – YbF3
Lu – Lu, LuF3
Hf - Hf
Ta - Ta
W - W
Re - Re
Os - Os
Ir - Ir
Pt - Pt
Au - Au
Hg - HgTe
Tl - Tl
Pb - PbO
Bi - Bi
Po – N/A
At - N/A
Rn - N/A
Fr – N/A
Ra – RaCO3
Ac - ?
Th – ThO2
Pa – N/A
U – UO2
Np - N/A
Pu - N/A

What do you think?  Do you have some favorite standards that I've overlooked?  Have I made some poor choices?

[Probeman]

I think this is a very worthy cause, though my impression is that the instrument vendor supplied mounts already typically contain pure metals and simple oxides (purity?).  At least my Cameca instrument came with such a mount though I almost never used it. As for SEM/EDS vendors do any manufacturers provide similar standard mounts?

I suspect that the much bigger problem is getting SEM users to use any standards at all! To wit:

https://smf.probesoftware.com/index.php?topic=1761.0

Does anyone have any polling data on what percentage of analyses are performed on SEMs using standards versus standardless?  Would the EDS vendors have any inkling of these percentages?

In my past discussions with EDS vendors, they have told me that almost none of their customers use standards, but what do we actually know?

[Nicholas Ritchie]

I view this as an attempt to pull a fraction (albeit a small fraction) over to standards-based analysis.  Make it as easy as possible with as few decisions as possible and maybe a few more people will give it a try.

Because EDS detectors are so stable, you really only need to collect new standards occasionally.  If you are willing to investment a day collecting standards, standards-based analysis can be as simple as standardless.

Sure some vendors provide a handful of pure metals and oxides but are they selected to be optimal for EDS? Do they cover enough of the periodic table?  How many people have attempted to perform standards-based EDS only to find they missed a suitable standard for one element?  Let's solve this problem by covering the entire accessible periodic table with EDS suitable standards.

Combine "virtual standards" with 2 or 3 comprehensive standard blocks and we could have a future in which EDS provides the accuracy of "matrix matched standards" with "close to the convenience" of standardless.

Part of the reason no-one uses standards because no-one has made it easy to use standards.  That is what I'm trying to do with this project.

[Probeman]

As I said, this is a very worthy cause.

And yes, I agree, if you want people to do something correctly, you need to make it easy to do.

Right now, my experience with Bruker/Thermo EDS software and standards is that it's much easier to do standardless, so anything to make it easier would be great.

Does it make any sense for DTSA2 to start interfacing with various EDS vendor hardware?  That is why I went to the effort in Probe for EPMA to interface to both JEOL and Cameca, then one can control the standardization method and make it easier.

In PFE, the standardization data is acquired exactly similar to the unknown data so there is really nothing more to learn.  And users tend to want to focus on acquiring unknown data more than standard data.

But as you said, with EDS it should be even easier as the frequency of re-calibration should be less.

[Nicholas Ritchie]

Any overarching suggestions/thoughts on selecting durable standards?  When would you consider using a binary when the pure element is available?

[Nicholas Ritchie]

Owen Neill suggested InAs for As.  The In M-lines are below the As L and the In L-lines are below the As K. It is a semiconductor and could serve for both In and As.

[Probeman]

For minor/trace element analyses, I would always use a pure metal or pure oxide if available.

As you well know, for minor/trace elements the background characterization is the most important parameter and a high concentration (that is count rate per concentration) in the primary standard gives the best sensitivity. That statement pretty much covers everything except the alkali metals, liquids/gases.

For major elements the standard accuracy (and matrix correction accuracy) are most important, and the chemical effects might also be worth considering (at least for low energy emission lines), so here is my list divided two ways into elemental/alloy and oxides/silicates:

Elemental                          Oxides/silicates

H - N/A                            H - N/A
He - N/A                           He - N/A
Li - N/A                           Li - N/A
Be - Be                            Be - BeAl2O4 (synthetic)
B - BN                             B - LaB6 or BN or HfB2
C - C                              C - CaCO3 (water clear calcite)
N - BN                             N - Si3N4 or AlN or GaN
O - ?                              O - MgO or SiO2
F - CaF2                           F - Na5Al3F14 (synthetic chiolite)
Ne - N/A                           Ne - N/A
Na - Na5Al3F14 (chiolite)          Na - NaAlSi3O8 (albite)
Mg - MgO                           Mg - Mg2SiO4 (synthetic forsterite)
Al - Al                            Al - Al2O3
Si - Si                            Si - SiO2 or Mg2SiO4 (synthetic forsterite)
P - GaP or InP                     P - GaP or NBS K-496 glass or ScPO4
S - FeS2 or CdS (synthetic)        S - CaSO4 (anhydrite)
Cl - RbCl                          Cl - RbCl 
Ar - N/A                           Ar - N/A
K - KBr                            K - KTaO3 or KAlSi3O8 (orthoclase)
Ca - CaF2                          Ca - CaSiO3 (synthetic wollastonite)
Sc - Sc                            Sc - ScPO4
Ti - Ti                            Ti - TiO2
V - V                              V - V2O3
Cr - Cr                            Cr - Cr2O3
Mn - Mn                            Mn - MnO
Fe - Fe                            Fe - Fe2O3 or Fe3O4
Co - Co                            Co - CoO
Ni - Ni                            Ni - NiO or Ni2O4
Cu - Cu or Cu2S (chalcocite)       Cu - Cu2O (synthetic)
Zn - Zn or ZnSe                    Zn - ZnO
Ga - GaP                           Ga - Gd3Ga5O12 (GGG)
Ge - Ge                            Ge - Ge
As - GaAs or InAs                  As - GaAs or InAs
Se - Se or ZnSe                    Se - Se or ZnSe
Br - KBr or CsBr                   Br - KBr or CsBr
Kr - N/A                           Kr - N/A
Rb - RbTiOPO4                      Rb - RbTiOPO4
Sr - SrF2                          Sr - SrTiO3
Y - YAG or YIG                     Y - YAG or YIG or YPO4
Zr - Zr                            Zr - ZrSiO4 (synthetic zircon)
Nb - Nb                            Nb - LiNbO4 LiNbO3
Mo - Mo                            Mo - PbMO4
Tc - N/A                           Tc - N/A
Ru - Ru                            Ru - Ru
Rh - Rh                            Rh - Rh
Pd - Pd                            Pd - Pd
Ag - Ag or Ag2S                    Ag - Ag
Cd - Cd or CdS                     Cd - Cd or CdS
In - In or InP                     In - In or InP
Sn - Sn                            Sn - Sn or SnO2
Sb - Sb or GaSb (synthetic)        Sb - Sb or GaSb (synthetic)
Te - Te or NiTe (synthetic)        Te - Te or NiTe (synthetic)
I - RbI or CsI                     I - RbI or Cu(IO3)(OH) (salesite)
Xe - N/A                           Xe - N/A
Cs - CsCl or CsBr                  Cs - CsCl or CsBr
Ba - BaF2                          Ba - BaSiO5 (sanbornite) or BaSO3 (barite)
La - LaB6 or LaF3                  La - LaPO4 (synthetic)
Ce - CeAl2 or CeF3                 Ce - PrPO4 (synthetic)
Pr - PrP5O14                       Pr - PrPO4 (synthetic)
Nd - NdP5O14                       Nd - NdPO4 (synthetic)
Pm - PmP5O14                       Pm - PmPO4 (synthetic)
Sm - SmP5O14                       Sm - SmPO4 (synthetic)
Eu - EuP5O14                       Eu - EuPO4 (synthetic)
Gd - GdP5O14                       Gd - GdPO4 (synthetic)
Tb - TbP5O14                       Tb - TbPO4 (synthetic)
Dy - Dy, DyF3                      Dy - DyPO4 (synthetic)
Ho - HoF3                          Ho - HoPO4 (synthetic)
Er - Er, ErF3                      Er - ErPO4 (synthetic)
Tm - TmF3                          Tm - TmPO4 (synthetic)
Yb - YbF3                          Yb - YbPO4 (synthetic)
Lu - Lu, LuF3                      Lu - LuPO4 (synthetic)
Hf - Hf or HfC                     Hf - HfSiO4 (synthetic)
Ta - Ta                            Ta - KTaO3 (synthetic) or CrTa2O6 (synthetic)
W - W                              W - PbWO4 (synthetic)
Re - Re                            Re - Re
Os - Os                            Os - Os
Ir - Ir                            Ir - Ir
Pt - Pt                            Pt - Pt
Au - Au                            Au - Au
Hg - HgTe                          Hg - HgTe
Tl - Tl                            Tl - Tl
Pb - PbS (synthetic galena)        Pb - PbSiO3 (alamosite) or PbMoO4 (wulfenite)
Bi - Bi                            Bi - Bi2O3
Po - N/A                           Po - N/A
At - N/A                           At - N/A
Rn - N/A                           Rn - N/A
Fr - N/A                           Fr - N/A
Ra - RaCO3                         Ra - RaCO3
Ac - ?                             Ac - ?
Th - ThC                           Th - ThO2 or ThSiO4 (synthetic)
Pa - N/A                           Pa - N/A
U - UO2                            U - UO2 or USiO4 (synthetic)
Np - N/A                           Np - N/A
Pu - N/A                           Pu - N/A

Also, oxides/silicates tend to oxidize less than pure elements or alloys, so there is some additional advantage in using oxides/silicates in standard mounts.

BTW, I just noticed that one of my standard names in the UofO standard database has a typo.  Standard 259 should be BeAl2O4, not Be4Al2O4!  However, the composition was correct, so just edit the standard name and all is good.

Also, I've always referred to the Smithsonian (Boatner) REE phosphates as PO4, but I see you have the REE phosphates as P5O14. Are those from another source?

[Nicholas Ritchie]

Thanks, this is really helpful. I'm coming around to the idea of using oxides like Al2O3 and SiO2 rather than pure elements although you do have to coat and even then the Duane-Hunt is often a little (100 eV-ish) surpressed.  Does anyone else have some suggestions?

The phosphates are "SPI Rare Earth Phosphates" . I recall that they are of Chineses origin via Astimex and are XP5O14. Seee https://astimex.com/com/catalog/reep.html

[crystalgrower]

I can offer a  shortlist of forms that should NOT be used for standards.  Based on years of purchasing for Astimex (check their website for details which still appears to be online) which have been in use for EDS since 1969.  FYI SPI Supplies does not maintain their online catalog any longer.  I still have the EDS spectra archived...my apologies for many updates.

Use of binary compounds such as GaAs and InP and Sb2S3  take care of many nonmetallic elements on EDS.  Astimex catalog only offered the beam- stable ones.   LiF , BN, also useful and no overlaps. 

Avoid graphite electrodes as a source of C (too porous)

Try to find an older source of Zr metal rod as the currently available wire has a hole down the middle.  This is partly due to Zr being classified as "reactor material".  Check both Zr and Hf for purity from the other element (price will be higher) because these have perfect overlap for EDS.   And ZrO2 NOT good because it will have too much of both Y and Hf  with perfect overlap on EDS.

We had success with Os, Ru, and other refractory metals from Metallium that they melted by electrical current in vacuum.  They tracked the original certificates of analysis faithfully and did custom cuts. We never found any other acceptable form of Os metal 

Optical grade Tl(Br,I) from a reliable source is the only really stable material for all three elements. Sold as KRS-5 for infrared optics, it is 6N and the eutectic composition.  Both TlBr and TlI are difficult to polish and rarely available as crystals. 

Th wire is good and available but must be kept dessicated.

I posted instructions on this forum for easy synthesis of K2SiF6, Rb2SnF6 and Cs2SiF6 which are far more homogeneous than any minerals.  Avoid pollucite  unless you can find a clean museum specimen.

Black "cassiterite" is closer to FeSn2O3  than SnO2.  A museum specimen of white SnO2 would not be better than Sn metal. 

For Cu use OFHC wire NOT any oxide.

Do not use Mg foil.  FAR too fragile.

Be careful that Be metal has no Al or flux present. 

RaSO4 more beam stable than RaCO3 and not hard to grow crystals.

UO2 notorious for being nonstoichiometric.  DU if you can find it would be the only real choice.

Use Y and V unless you have access to hydrothermally recrystallized YVO4.    ALL Y-P compounds have perfect overlap. 

When considering sulfides, ZnS is preferable to FeS2 because Zn has no redox complications.  ZnS is another optical material available 6N pure, it's translucent yellow.

Boatner RPO4 only good for Tb-Lu and Y.  Contact me about La-Gd as PO4 offline.
Please contact me offline for rare earth pentaphosphates.  I was the last supplier to Astimex.

Other options for REE are metals for Sm and higher and fluorides for La-Eu.

[Nicholas Ritchie]

Thanks for all the suggestions crystalgrower. I'm going to have to go through your note line-by-line.

[Probeman]

We had success with Os, Ru, and other refractory metals from Metallium that they melted by electrical current in vacuum.  They tracked the original certificates of analysis faithfully and did custom cuts. We never found any other acceptable form of Os metal 

We obtained some 99.997% Ru powder from Aesar and vacuum arc melted it ourselves.

[Ben Buse]

Interestingly oxford instruments seems to sell standard blocks, not cheap at £3-4K.
https://nano.oxinst.com/assets/uploads/products/nanoanalysis/documents/Brochures/Technical%20Datasheets/AZtecWave%20AZtec%20Standards.pdf

Not sure who makes them. Obviously more generally with EPMA there seems to be a lack of commercial standard blocks of well characterised standards.

[crystalgrower]

I would like to know if NIST is planning to front the funds to either purchase wire anjd solids to be sold as small cuts, or is NIST considering preparing mounts.

FYI Astimex never patented or registered or copyrighted any mount layouts so they can be produced by anybody.  The holes were 2mm drill bit drilled 2mm deep using a radial  stepping microlathe setup.

It is very time efficient to make 10 mounts at a time.  For Astimex/SPI the wire was cut and dropped into each well according to the printed layout, and then Buehler EpoHeat  was added to bond each round.  Polishing can be expedited by clearing off nubs of wire with a diamond wafer saw (which also means much less polishing time and media).  The cost of materials and all prep for 48 wells would be about $2000 CDN at current "expert" wages.  Resellers always add their own markup, so NIIST can offer the best price by only selling direct.

[Nicholas Ritchie]

NIST would never attempt to produce mounts like this for the commercial market.  My intent is to determine a "reasonably good" set of "EDS standards" that any vendor (or individual) could produce and sell.  They would be carefully selected to make EDS standards-based quant relatively easy as well as facilitating the creation and sharing of "virtual standards."

[Nicholas Ritchie]

I've summarized the conversation so far in this table.

The optimal column is my current thinking about the best choice to use as "transfer standards" for EDS analysis - robust, available, no peak interferences, etc.