Looking for suggestions for "Basic EDS standards"

[crystalgrower]

Here is the best for EDS list that was placed on the commercial market in 2010.

Li-LiF
Be
B-BN
C-pyrolytic plate
N-BN
O-MgO
F-LiF
Na-Na3AlF6
Mg-MgO
Al, Si element
P from GaP or InP
S from ZnS
Cl from K2PtCl6
K from K2PtCl6
Ca-CaF2
Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn metal
Ga-GaAs or GaP
Se element
As from GaAs
Br from Tl (Br,I)
Rb from Rb2SnF6
Sr from SrF2
Y,Zr, Nb, Mo, Ru, Rh, Pg, Ag, Cd from metal
InP as In difficult to polish
Sn, Sb, Te as element
I from Tl(Br,I)
Cs-Cs2SiF6
Ba-BaF2
La, Ce, Pr, Nd, Sm, Eu as RF3
Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Hf, Ta, W, Re, Os, Ir, Pt, Au as metal
Hg as Hg2Te
Tl from Tl(Br,i)
Bi metal

The constraints which are required in addition to no overlap are INSOLUBLE, SYNTHETIC, and MELTING OVER 100C.

These materials were all checked for overlap in both EDS and WDS. Any trivial EDS overlap in minor peaks was accepted if no other substance could be mounted and remain stable.  AND a ;license required to purchase RaSO4 and DU.  You might get lucky with Th.

FYI peak shape is a function of detector.  Overload any detector and the peak energy shifts to higher channels AND spreads out. 

[JonF]

Nb - LiNbO4

Out of curiosity, why the preference for LiNbO4 over a simple oxide like Nb2O5? Ted Pella are listing Nb2O5 with a 3N5 purity rating. Is there a stability issue or something?

LiNbO4 seems an odd choice for a reference material where most people can't analyse one of the major constituents (Li).

[Probeman]

Nb - LiNbO4

Out of curiosity, why the preference for LiNbO4 over a simple oxide like Nb2O5? Ted Pella are listing Nb2O5 with a 3N5 purity rating. Is there a stability issue or something?

LiNbO4 seems an odd choice for a reference material where most people can't analyse one of the major constituents (Li).

Interesting. I did not know this material is available. Indeed, Nb2O5 would be a great primary standard for Nb.

The fact that there are two high purity synthetic materials available, both containing Nb, is also a good thing for our consensus k-ratio project:

https://smf.probesoftware.com/index.php?topic=1442.0

That Li cannot normally be analyzed is no problem for its use as a standard, since all elements are declared in standard compositional databases.

[John Donovan]

Another point. When looking at the claimed purity of synthetic materials, be aware that sometimes they will claim something like "99.9% pure (metals basis)":

https://smf.probesoftware.com/index.php?topic=1504.msg11584#msg11584

[sem-geologist]

LiNbO4? Personally I don't know this kind of material. There is indeed LiNbO3, Also there is LiTaO3. Both of these are synthesized in industrial quantities in very pure form as they are one of materials used in many optical devices. It is much easier to source pure synthetically grown crystals of these, than Nb2O5 which purity can be questionable (the purity of bulk quantities of Nb2O5 is not so industrially important compared to LiNbO3).
 

[Nicholas Ritchie]

Right.  LiNbO3.

[JonF]

Slightly off tangent, but when a standard material as the above LiNbO3 is listed as "high purity", is it simply the absence of significant concentrations of contaminants, or is it inferring that a material is (also?) stoichiometric?

Doing a (very) quick literature search suggests LiNbO3, Li3NbO4 and LiNb3O8 may all be made simultaneously, and all would pass a test for contaminants, but the resulting material would obviously not make a great standard material.

[Probeman]

LiNbO4? Personally I don't know this kind of material. There is indeed LiNbO3, Also there is LiTaO3. Both of these are synthesized in industrial quantities in very pure form as they are one of materials used in many optical devices. It is much easier to source pure synthetically grown crystals of these, than Nb2O5 which purity can be questionable (the purity of bulk quantities of Nb2O5 is not so industrially important compared to LiNbO3).

Yes.

I edited this post to reflect that:

https://smf.probesoftware.com/index.php?topic=1771.msg13570#msg13570

[sem-geologist]

Doing a (very) quick literature search suggests LiNbO3, Li3NbO4 and LiNb3O8 may all be made simultaneously, and all would pass a test for contaminants, but the resulting material would obviously not make a great standard material.

Could you share that quick literature search?

I mean LiNbO3 grown using Czochralski process, that is a single uniform crystal with no boundaries or defects within. Can we go any further to define more "homogeneous" and more "pure" – purity of these are know to have contaminants at level below ppb's. Not some LiNbO3, LiNbO4, LiNb3O8 grown in some "kitchen sink" simultaneously from some Li,Nb rich "soup".

[sem-geologist]

Doing a (very) quick literature search suggests LiNbO3, Li3NbO4 and LiNb3O8 may all be made simultaneously, and all would pass a test for contaminants, but the resulting material would obviously not make a great standard material.

Could you share that quick literature search?

I mean LiNbO3 grown using Czochralski process, that is a single uniform crystal with no boundaries or defects within. Can we go any further to define more "homogeneous" and more "pure" – purity of these are know to have contaminants at level below ppb's. Not some LiNbO3, LiNbO4, LiNb3O8 grown in some "kitchen sink" simultaneously from some Li,Nb rich "soup".

Indeed Your pointing out made me worried. Seems that LiNbO3 can have defects of Li deficiency (vacancies) even then grown with Czochralski's process. My own LiNbO3 is a small cubic crystal and that makes me to suspect it rather was not grown using Czochralski process, but "kitchen sink" with Li,Nb rich "soup". While WDS scan at high current (1000nA) showed no detectible contaminants, about the Li:Nb:O ratio now I am on doubt.

[JonF]

Indeed Your pointing out made me worried. Seems that LiNbO3 can have defects of Li deficiency (vacancies) even then grown with Czochralski's process. My own LiNbO3 is a small cubic crystal and that makes me to suspect it rather was not grown using Czochralski process, but "kitchen sink" with Li,Nb rich "soup". While WDS scan at high current (1000nA) showed no detectible contaminants, about the Li:Nb:O ratio now I am on doubt.

Sorry, I wasn't intending to worry anyone!

This was the stand-out paper that got me thinking:
Identification of LiNbO3, LiNb3O8 and Li3NbO4 phases in thin films synthesized with different deposition techniques by means of XRD and Raman spectroscopy, Bartasyte et al, 2013, DOI 10.1088/0953-8984/25/20/205901

Seems the different phases can be distinguished quite easily by Raman.
The paper was dealing with Li-Nb thin films deposited by various means - I've no idea how applicable it is to EPMA standards. I was just trying to see whether variations in Li:Nb stoichiometry could occur.


Going back to the topic, my point was simply that I'm not sure how much we can trust a standard when we can't verify it against our other standards.

For example, potassium niobate (KNbO3)*, we would be able to measure the K and the Nb (and the O if you wanted!) and compare them against our other K and Nb (and O) standards. With lithium niobate, we're putting a lot of faith in the standard manufacturer to get what we asked for, although we could measure Nb and O.

*I've no idea whether KNbO3 would make a good standard, mostly likely not. It also looks to be toxic.

[sem-geologist]

The major distinction is how these crystals are grown. Thin films for sure are not grown with Czochralski's process, thus it practically is not comparable. I think one of thing I could do with my own LiNbO3 standard is to try looking at it with EBSD, if it is monocrystal - there is higher chance it is a small regular shard from the larger crystal grown with Czochralski's process. Albeit these regular shapes are rather characteristic for small synthetic hydrothermal growth. Yes, for sure due to unmeasurable Li, we are at mercy of manufacturer and distributor, we have no easy way to verify its stochiometry.

[Probeman]

The major distinction is how these crystals are grown. Thin films for sure are not grown with Czochralski's process, thus it practically is not comparable. I think one of thing I could do with my own LiNbO3 standard is to try looking at it with EBSD, if it is monocrystal - there is higher chance it is a small regular shard from the larger crystal grown with Czochralski's process. Albeit these regular shapes are rather characteristic for small synthetic hydrothermal growth. Yes, for sure due to unmeasurable Li, we are at mercy of manufacturer and distributor, we have no easy way to verify its stochiometry.

We had a similar problem with the RbTiOPO4 that Marc Schirer grew for me back in the day at UC Berkeley (now at The CalChemist).  That is, how to know its stoichiometry given that we needed it as a Rb standard precisely because we didn't have any other Rb standard!

But we did measure the Ti using TiO2 and P using a YPO4 that we got from the Oak Ridge/Smithsonian, and of course checked it for impurities and homogeneity:

https://smf.probesoftware.com/index.php?topic=701.0

This summer at M&M I raised the issue of determining stoichiometric accuracy with a few of our colleagues and Ian Anderson mentioned that if one assumed that the stoichiometry was related to the degree of lattice defects, one might measure defects using a technique such as positron annihilation spectroscopy:

https://en.wikipedia.org/wiki/Positron_annihilation_spectroscopy

Not exactly a bench top characterization method, but maybe someone here has access to such a technique?  I note that Washington University has such a device:

https://materialsresearch.wsu.edu/positronannihilationspectrometer/