Synthetic Quartz Standard?

[UT E-Beam Lab]

Can anyone recommend a manufacturer where I can purchase a synthetic quartz standard in order to do Donovan's "blank correction" method for measuring Ti concentations in quartz grains?

Many thanks

[UofO EPMA Lab]

I got my original SiO2 synthetic material at ESPI, now (amazingly enough) located in Ashland, Oregon!

http://www.espimetals.com/index.php/msds/245-silicon-oxide-sio2

They might already have characterized the trace Ti and/or Al. The lot I bought 20 years ago had 1.42 PPM Ti by ICP-MS (A. Konig).

The SiO2 they currently have is probably different material but here is what I have on the original material from 20 years ago:

Specimen from ESPI, 99.99%, EPMA (UCB): Al2O3 ~ 0.01%
Catalog #K4699M
ICP-MS (Alan Konig, PPM) Ti=1.42, Li=3.18, Cu=1.26, Mn=2.93, Pb=0.11, Zn=2.09
Flame AA (Chris Lewis, PPM):
Al=15 +/- 5
Fe=6 +/- 3
Mn=1.5 +/- 0.3
Na=5 +/- 3
Li= 2.3 +/- 0.2

[UT E-Beam Lab]

Great! Thank you so much

[Karsten Goemann]

We use Saint Gobain Spectrosil

[Probeman]

We use Saint Gobain Spectrosil

Could you post the trace element characterization and how it was performed on this material?

[Karsten Goemann]

I've attached an info sheet with the specs. <10 ppb typical for most elements, but not sure how that was determined. Our LA-ICP-MS people are using it too and it seems to be clean, certainly by EPMA standards.

[Probeman]

This is a very interesting idea to utilize an ultra pure SiO2 glass as a "blank" standard for quartz analyses.

In fact, one advantage of using SiO2 glass is that it is actually *less* sensitive to the electron beam than SiO2 crystal. Here is a reflected light image of a scanned electron beam on SiO2 crystal:



and here is SiO2 glass under the same scan conditions:



But maybe this is not a good thing in that the SiO2 glass might no longer be a similar enough matrix for calibration of the "blank" concentration?

I suppose we'll need to do some high precision measurements and compare traces in SiO2 glass and SiO2 crystal...

In the meantime, I wanted ask everyone: is there a similar ultra pure titania (TiO2) glass that could also be utilized as a blank standard for traces in rutile?

[Karsten Goemann]

I don't know - I would expect the beam sensitivity to vary a bit between quartz samples as well?

It does seem to work for us - We've put our Hot Springs Quartz chip under the LA-ICP-MS and got 1.2 ppm Ti. With the probe our long term average is 3+/-3 ppm using the blank correction, which is +16+/-2ppm.

[BenjaminWade]

Late reply...

A lot of XRD uses Spectrosil qtz glass tubes to put their samples in. If you have a friendly XRD tech they might be happy to smash one for you. I obtained a few tubes and analysed fragments from each one for Ti (+Al and REE) via LA-ICP-MS. The resulting average Ti concs. from each tube were well within error of each other (~4ppm). So although it sounds like my glass is a little less pure than some others Spectrosil, it doesn't really matter what it is as long as its well characterised.

At the microprobe workshop Karsten just ran here in Adelaide, I gave Dave Adam's a small chip of synthetic TiO2. This material was given to me on request to our local museum (and apparently was quite expensive for them to obtain), but I would be willing to send you a small chip. However....there was no information on its trace composition, AND I haven't had a chance to analyse it myself. So it could be poor material, I just don't know. If you are willing to take that chance I could send you some.

Cheers

[Probeman]

A lot of XRD uses Spectrosil qtz glass tubes to put their samples in. If you have a friendly XRD tech they might be happy to smash one for you. I obtained a few tubes and analysed fragments from each one for Ti (+Al and REE) via LA-ICP-MS. The resulting average Ti concs. from each tube were well within error of each other (~4ppm). So although it sounds like my glass is a little less pure than some others Spectrosil, it doesn't really matter what it is as long as its well characterised.

Yes, that is correct.

Because the blank correction is fully quantitative it can handle the situation where one's "blank" standard contains a non-zero concentration of the element of interest, as long as that non-zero value is accurate. The reason it's called a "blank correction" is simply because it's usually easier to find a pure material with essentially a zero concentration as opposed to a standard containing a known non-zero concentration.  But both concentrations (zero and non-zero) are equally good at ascertaining accuracy!

Here is a poster and PPT with the blank correction equation attached. Note that this correction must be incorporated into the matrix iteration to achieve the necessary accuracy for trace element characterization:

[Probeman]

Here's a plot I made a while ago comparing quartz glass and crystal for Si Ka and O ka.  Note the absorbed current changes.



Note that over time the oxygen intensity goes up in the crystal, while it goes down in the glass, though more slowly. I have no idea what this means... but thank goodness for the TDI correction.