At 9:41 AM +0200 7/28/03, Dr H.MOURI wrote:
>Dear friends,
>Is there any body who could tell me if we can analyse accessory phases (such
>as monazite) using JEOL JXA-8200 for in situ dating and get reliable
>results? I know that we can do that with CAMECA SX50/100, but I have no idea
>about JEOL. Which type of machine is the best for this kind of work?
>Any information about this subject will be highly appreciated and thanks a
>million in advance.
>With kind regards
>Hassina
Dear Hassina (and others) -
At RPI, myself and Frank Spear have been using a JEOL 733
Superprobe (ca. 1986) to analyze monazite, and we are pretty happy
with the results from a 15+ year old machine. I think that with a
JEOL 8200 you will do fine, provided you are careful in setting up
the analytical protocol.
Both CAMECA and JEOL make EMPs that are adequate for the
chemical dating of Pb-rich phases, such as monazite, xenotime,
uraninite, thorite, etc. What is critical in determining whether the
machine will work to the user's desired level of precision is the
following:
1) Detector gas: Ar (or P10) generates escape peaks from second order
LREE lines (notably Ce and La) that are unfilterable using Pulse
Height discrimination un-filterable. The Ce escape peak is
particularly bad, as it overlaps Pb Mb (for Ce-rich minerals such as
monazite). Xe generates LREE escape peaks that, due to the energy
difference between Ar and Xe, may be filtered with Pulse Height
discrimination. Additionally, there is the problem of the Ar
absorption edge near the U Ma and Mb lines.
2) Rowland Circle Diameter: For large (160 mm) Rowland circles, the
above interference (Ce escape peak on Pb Mb) may not be problematic
(or as problematic), as the 160 mm circle may provide enough
wavelength resolution to resolve Pb Mb and Ce La (escape). For 140 mm
Rowland circles, the wavelength resolution is inadequate to remove
this interference. Conversely, the count intensity increases for a
140 mm Rowland circle, and even more so for a 100 mm Rowland circle
(so called "high intensity" spectrometers), but again, the wavelength
resolution decreases even more for the 100 mm circle, and both
avoiding interferences and placing background collection positions
for Pb and U can be quite difficult.
3) diffraction crystals: the more PET crystals, the better. If you
can analyze lead simultaneously on two PET crystals, your average
analytical precision increases by a factor of the square root of 2
(1.4). Likewise, three simultaneous analyses of lead using PET
crystals increases the precision of your mean analysis by a factor
of 1.7. At RPI, we have 4 PET crystals, and we analyze Th and U on
one PET crystal, Ce and Y on another PET crystal (these elements are
measured for interference corrections and as monitor elements), and
Pb simultaneously with the other 2 PET crystals.
4) X-ray collimation: Wide open settings produce the highest
intensity and highest analytical precision, therefore, but with the
lowest peak to background ratio. Narrow slit settings increase P/B
ratio, but at expense of intensity. Since this analysis is
essentially about maximizing Pb precision, you should open the slits
up.
There are a lot of other things to consider in producing chemical age
analyses on an EMP, but these are the major machine-related issues.
I've attached a preprint of an article we have written which
discusses estimation of precision and accuracy of EMP chemical ages
of monazite - maybe you will find this useful. Hope this all helps.
Regards,
Joe Pyle
PS apologies for the large file size - if you (or any other
interested parties) do not receive the file, let me know, and I can
post the article to my web site
